Journal articles on the topic 'Clouds Classification'
To see the other types of publications on this topic, follow the link: Clouds Classification.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 journal articles for your research on the topic 'Clouds Classification.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse journal articles on a wide variety of disciplines and organise your bibliography correctly.
1
Hutchison, Keith D., Barbara D. Iisager, Thomas J. Kopp, and John M. Jackson. "Distinguishing Aerosols from Clouds in Global, Multispectral Satellite Data with Automated Cloud Classification Algorithms." Journal of Atmospheric and Oceanic Technology 25, no. 4 (April 1, 2008): 501–18. http://dx.doi.org/10.1175/2007jtecha1004.1.
Full textAbstract:
Abstract A new approach is presented to distinguish between clouds and heavy aerosols with automated cloud classification algorithms developed for the National Polar-orbiting Operational Environmental Satellite System (NPOESS) program. These new procedures exploit differences in both spectral and textural signatures between clouds and aerosols to isolate pixels originally classified as cloudy by the Visible/Infrared Imager/Radiometer Suite (VIIRS) cloud mask algorithm that in reality contains heavy aerosols. The procedures have been tested and found to accurately distinguish clouds from dust, smoke, volcanic ash, and industrial pollution over both land and ocean backgrounds in global datasets collected by NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) sensor. This new methodology relies strongly upon data collected in the 0.412-μm bandpass, where smoke has a maximum reflectance in the VIIRS bands while dust simultaneously has a minimum reflectance. The procedures benefit from the VIIRS design, which is dual gain in this band, to avoid saturation in cloudy conditions. These new procedures also exploit other information available from the VIIRS cloud mask algorithm in addition to cloud confidence, including the phase of each cloudy pixel, which is critical to identify water clouds and restrict the use of spectral tests that would misclassify ice clouds as heavy aerosols. Comparisons between results from these new procedures, automated cloud analyses from VIIRS heritage algorithms, manually generated analyses, and MODIS imagery show the effectiveness of the new procedures and suggest that it is feasible to identify and distinguish between clouds and heavy aerosols in a single cloud mask algorithm.
2
Urbanek, Benedikt, Silke Groß, Andreas Schäfler, and Martin Wirth. "Determining stages of cirrus evolution: a cloud classification scheme." Atmospheric Measurement Techniques 10, no. 5 (May 3, 2017): 1653–64. http://dx.doi.org/10.5194/amt-10-1653-2017.
Full textAbstract:
Abstract. Cirrus clouds impose high uncertainties on climate prediction, as knowledge on important processes is still incomplete. For instance it remains unclear how cloud microphysical and radiative properties change as the cirrus evolves. Recent studies classify cirrus clouds into categories including in situ, orographic, convective and liquid origin clouds and investigate their specific impact. Following this line, we present a novel scheme for the classification of cirrus clouds that addresses the need to determine specific stages of cirrus evolution. Our classification scheme is based on airborne Differential Absorption and High Spectral Resolution Lidar measurements of atmospheric water vapor, aerosol depolarization, and backscatter, together with model temperature fields and simplified parameterizations of freezing onset conditions. It identifies regions of supersaturation with respect to ice (ice-supersaturated regions, ISSRs), heterogeneous and homogeneous nucleation, depositional growth, and ice sublimation and sedimentation with high spatial resolution. Thus, all relevant stages of cirrus evolution can be classified and characterized. In a case study of a gravity lee-wave-influenced cirrus cloud, encountered during the ML-CIRRUS flight campaign, the applicability of our classification is demonstrated. Revealing the structure of cirrus clouds, this valuable tool might help to examine the influence of evolution stages on the cloud's net radiative effect and to investigate the specific variability of optical and microphysical cloud properties in upcoming research.
3
Wang, Y., M. Penning de Vries, P. H. Xie, S. Beirle, S. Dörner, J. Remmers, A. Li, and T. Wagner. "Cloud and aerosol classification for 2 1/2 years of MAX-DOAS observations in Wuxi (China) and comparison to independent data sets." Atmospheric Measurement Techniques Discussions 8, no. 5 (May 6, 2015): 4653–709. http://dx.doi.org/10.5194/amtd-8-4653-2015.
Full textAbstract:
Abstract. Multi-Axis-Differential Optical Absorption Spectroscopy (MAX-DOAS) observations of trace gases can be strongly influenced by clouds and aerosols. Thus it is important to identify clouds and characterise their properties. In a recent study Wagner et al. (2014) developed a cloud classification scheme based on the MAX-DOAS measurements themselves with which different "sky conditions" (e.g. clear sky, continuous clouds, broken clouds) can be distinguished. Here we apply this scheme to long term MAX-DOAS measurements from 2011 to 2013 in Wuxi, China (31.57° N, 120.31° E). The original algorithm has been modified, in particular in order to account for smaller solar zenith angles (SZA). Instrumental degradation is accounted for to avoid artificial trends of the cloud classification. We compared the results of the MAX-DOAS cloud classification scheme to several independent measurements: aerosol optical depth from a nearby AERONET station and from MODIS, visibility derived from a visibility meter; and various cloud parameters from different satellite instruments (MODIS, OMI, and GOME-2). The most important findings from these comparisons are: (1) most cases characterized as clear sky with low or high aerosol load were associated with the respective AOD ranges obtained by AERONET and MODIS, (2) the observed dependences of MAX-DOAS results on cloud optical thickness and effective cloud fraction from satellite indicate that the cloud classification scheme is sensitive to cloud (optical) properties, (3) separation of cloudy scenes by cloud pressure shows that the MAX-DOAS cloud classification scheme is also capable of detecting high clouds, (4) some clear sky conditions, especially with high aerosol load, classified from MAX-DOAS observations corresponding to the optically thin and low clouds derived by satellite observations probably indicate that the satellite cloud products contain valuable information on aerosols.
4
Marchant, Benjamin, Steven Platnick, Kerry Meyer, G. Thomas Arnold, and Jérôme Riedi. "MODIS Collection 6 shortwave-derived cloud phase classification algorithm and comparisons with CALIOP." Atmospheric Measurement Techniques 9, no. 4 (April 11, 2016): 1587–99. http://dx.doi.org/10.5194/amt-9-1587-2016.
Full textAbstract:
Abstract. Cloud thermodynamic phase (ice, liquid, undetermined) classification is an important first step for cloud retrievals from passive sensors such as MODIS (Moderate Resolution Imaging Spectroradiometer). Because ice and liquid phase clouds have very different scattering and absorbing properties, an incorrect cloud phase decision can lead to substantial errors in the cloud optical and microphysical property products such as cloud optical thickness or effective particle radius. Furthermore, it is well established that ice and liquid clouds have different impacts on the Earth's energy budget and hydrological cycle, thus accurately monitoring the spatial and temporal distribution of these clouds is of continued importance. For MODIS Collection 6 (C6), the shortwave-derived cloud thermodynamic phase algorithm used by the optical and microphysical property retrievals has been completely rewritten to improve the phase discrimination skill for a variety of cloudy scenes (e.g., thin/thick clouds, over ocean/land/desert/snow/ice surface, etc). To evaluate the performance of the C6 cloud phase algorithm, extensive granule-level and global comparisons have been conducted against the heritage C5 algorithm and CALIOP. A wholesale improvement is seen for C6 compared to C5.
5
Marchant, B., S. Platnick, K. Meyer, G. T. Arnold, and J. Riedi. "MODIS Collection 6 shortwave-derived cloud phase classification algorithm and comparisons with CALIOP." Atmospheric Measurement Techniques Discussions 8, no. 11 (November 16, 2015): 11893–924. http://dx.doi.org/10.5194/amtd-8-11893-2015.
Full textAbstract:
Abstract. Cloud thermodynamic phase (ice, liquid, undetermined) classification is an important first step for cloud retrievals from passive sensors such as MODIS (Moderate-Resolution Imaging Spectroradiometer). Because ice and liquid phase clouds have very different scattering and absorbing properties, an incorrect cloud phase decision can lead to substantial errors in the cloud optical and microphysical property products such as cloud optical thickness or effective particle radius. Furthermore, it is well established that ice and liquid clouds have different impacts on the Earth's energy budget and hydrological cycle, thus accurately monitoring the spatial and temporal distribution of these clouds is of continued importance. For MODIS Collection 6 (C6), the shortwave-derived cloud thermodynamic phase algorithm used by the optical and microphysical property retrievals has been completely rewritten to improve the phase discrimination skill for a variety of cloudy scenes (e.g., thin/thick clouds, over ocean/land/desert/snow/ice surface, etc). To evaluate the performance of the C6 cloud phase algorithm, extensive granule-level and global comparisons have been conducted against the heritage C5 algorithm and CALIOP. A wholesale improvement is seen for C6 compared to C5.
6
Gryspeerdt, Edward, Johannes Quaas, Tom Goren, Daniel Klocke, and Matthias Brueck. "An automated cirrus classification." Atmospheric Chemistry and Physics 18, no. 9 (May 3, 2018): 6157–69. http://dx.doi.org/10.5194/acp-18-6157-2018.
Full textAbstract:
Abstract. Cirrus clouds play an important role in determining the radiation budget of the earth, but many of their properties remain uncertain, particularly their response to aerosol variations and to warming. Part of the reason for this uncertainty is the dependence of cirrus cloud properties on the cloud formation mechanism, which itself is strongly dependent on the local meteorological conditions. In this work, a classification system (Identification and Classification of Cirrus or IC-CIR) is introduced to identify cirrus clouds by the cloud formation mechanism. Using reanalysis and satellite data, cirrus clouds are separated into four main types: orographic, frontal, convective and synoptic. Through a comparison to convection-permitting model simulations and back-trajectory-based analysis, it is shown that these observation-based regimes can provide extra information on the cloud-scale updraughts and the frequency of occurrence of liquid-origin ice, with the convective regime having higher updraughts and a greater occurrence of liquid-origin ice compared to the synoptic regimes. Despite having different cloud formation mechanisms, the radiative properties of the regimes are not distinct, indicating that retrieved cloud properties alone are insufficient to completely describe them. This classification is designed to be easily implemented in GCMs, helping improve future model–observation comparisons and leading to improved parametrisations of cirrus cloud processes.
7
Chen, Xidong, Liangyun Liu, Yuan Gao, Xiao Zhang, and Shuai Xie. "A Novel Classification Extension-Based Cloud Detection Method for Medium-Resolution Optical Images." Remote Sensing 12, no. 15 (July 23, 2020): 2365. http://dx.doi.org/10.3390/rs12152365.
Full textAbstract:
Accurate cloud detection using medium-resolution multispectral satellite imagery (such as Landsat and Sentinel data) is always difficult due to the complex land surfaces, diverse cloud types, and limited number of available spectral bands, especially in the case of images without thermal bands. In this paper, a novel classification extension-based cloud detection (CECD) method was proposed for masking clouds in the medium-resolution images. The new method does not rely on thermal bands and can be used for masking clouds in different types of medium-resolution satellite imagery. First, with the support of low-resolution satellite imagery with short revisit periods, cloud and non-cloud pixels were identified in the resampled low-resolution version of the medium-resolution cloudy image. Then, based on the identified cloud and non-cloud pixels and the resampled cloudy image, training samples were automatically collected to develop a random forest (RF) classifier. Finally, the developed RF classifier was extended to the corresponding medium-resolution cloudy image to generate an accurate cloud mask. The CECD method was applied to Landsat-8 and Sentinel-2 imagery to test the performance for different satellite images, and the well-known function of mask (FMASK) method was employed for comparison with our method. The results indicate that CECD is more accurate at detecting clouds in Landsat-8 and Sentinel-2 imagery, giving an average F-measure value of 97.65% and 97.11% for Landsat-8 and Sentinel-2 imagery, respectively, as against corresponding results of 90.80% and 88.47% for FMASK. It is concluded, therefore, that the proposed CECD algorithm is an effective cloud-classification algorithm that can be applied to the medium-resolution optical satellite imagery.
8
Wang, Y., M. Penning de Vries, P. H. Xie, S. Beirle, S. Dörner, J. Remmers, A. Li, and T. Wagner. "Cloud and aerosol classification for 2.5 years of MAX-DOAS observations in Wuxi (China) and comparison to independent data sets." Atmospheric Measurement Techniques 8, no. 12 (December 10, 2015): 5133–56. http://dx.doi.org/10.5194/amt-8-5133-2015.
Full textAbstract:
Abstract. Multi-axis differential optical absorption spectroscopy (MAX-DOAS) observations of trace gases can be strongly influenced by clouds and aerosols. Thus it is important to identify clouds and characterize their properties. In a recent study Wagner et al. (2014) developed a cloud classification scheme based on the MAX-DOAS measurements themselves with which different "sky conditions" (e.g., clear sky, continuous clouds, broken clouds) can be distinguished. Here we apply this scheme to long-term MAX-DOAS measurements from 2011 to 2013 in Wuxi, China (31.57° N, 120.31° E). The original algorithm has been adapted to the characteristics of the Wuxi instrument, and extended towards smaller solar zenith angles (SZA). Moreover, a method for the determination and correction of instrumental degradation is developed to avoid artificial trends of the cloud classification results. We compared the results of the MAX-DOAS cloud classification scheme to several independent measurements: aerosol optical depth from a nearby Aerosol Robotic Network (AERONET) station and from two Moderate Resolution Imaging Spectroradiometer (MODIS) instruments, visibility derived from a visibility meter and various cloud parameters from different satellite instruments (MODIS, the Ozone Monitoring Instrument (OMI) and the Global Ozone Monitoring Experiment (GOME-2)). Here it should be noted that no quantitative comparison between the MAX-DOAS results and the independent data sets is possible, because (a) not exactly the same quantities are measured, and (b) the spatial and temporal sampling is quite different. Thus our comparison is performed in a semi-quantitative way: the MAX-DOAS cloud classification results are studied as a function of the external quantities. The most important findings from these comparisons are as follows: (1) most cases characterized as clear sky with low or high aerosol load were associated with the respective aerosol optical depth (AOD) ranges obtained by AERONET and MODIS; (2) the observed dependences of MAX-DOAS results on cloud optical thickness and effective cloud fraction from satellite confirm that the MAX-DOAS cloud classification scheme is sensitive to cloud (optical) properties; (3) the separation of cloudy scenes by cloud pressure shows that the MAX-DOAS cloud classification scheme is also capable of detecting high clouds; (4) for some cloud-free conditions, especially with high aerosol load, the coincident satellite observations indicated optically thin and low clouds. This finding indicates that the satellite cloud products contain valuable information on aerosols.
9
Wagner, T., S. Beirle, S. Dörner, U. Friess, J. Remmers, and R. Shaiganfar. "Cloud detection and classification based on MAX-DOAS observations." Atmospheric Measurement Techniques Discussions 6, no. 6 (December 3, 2013): 10297–360. http://dx.doi.org/10.5194/amtd-6-10297-2013.
Full textAbstract:
Abstract. Multi-AXis-Differential Optical Absorption Spectroscopy (MAX-DOAS) observations of aerosols and trace gases can be strongly influenced by clouds. Thus it is important to identify clouds and characterise their properties. In this study we investigate the effects of clouds on several quantities which can be derived from MAX-DOAS observations, like the radiance, the colour index (radiance ratio at two selected wavelengths), the absorption of the oxygen dimer O4 and the fraction of inelastically scattered light (Ring effect). To identify clouds, these quantities can be either compared to their corresponding clear sky reference values, or their dependencies on time or viewing direction can be analysed. From the investigation of the temporal variability the influence of clouds can be identified even for individual measurements. Based on our investigations we developed a cloud classification scheme, which can be applied in a flexible way to MAX-DOAS or zenith DOAS observations: in its simplest version, zenith observations of the colour index are used to identify the presence of clouds (or high aerosol load). In more sophisticated versions, also other quantities and viewing directions are considered, which allows sub-classifications like e.g. thin or thick clouds, or fog. We applied our cloud classification scheme to MAX-DOAS observations during the CINDI campaign in the Netherlands in Summer 2009 and found very good agreement with sky images taken from ground.
10
Behrangi, Ali, Terry Kubar, and Bjorn Lambrigtsen. "Phenomenological Description of Tropical Clouds Using CloudSat Cloud Classification." Monthly Weather Review 140, no. 10 (October 1, 2012): 3235–49. http://dx.doi.org/10.1175/mwr-d-11-00247.1.
Full textAbstract:
Abstract Two years of tropical oceanic cloud observations are analyzed using the operational CloudSat cloud classification product and Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) lidar. Relationships are examined between cloud types, sea surface temperature (SST), and location during the CloudSat early morning and afternoon overpasses. Based on CloudSat and combined lidar–radar products, the maximum and minimum cloud fractions occur at SSTs near 303 and 299 K, respectively, corresponding to deep convective/detrained cloud populations and the transition from shallow to deep convection. For SSTs below approximately 301 K, low clouds (stratiform and stratocumulus) are dominant (cloud fraction between 0.15 and 0.37) whereas high clouds are dominant for SSTs greater than about 301 K (cloud fraction between 0.18 and 0.28). Consistent with previous studies, most tropical low clouds are associated with lower SSTs, with a strong inverse linear relationship between low cloud frequency and SST. For all cloud types except nimbostratus, stratus, and stratocumulus, a sharp increase in frequency of occurrence is observed for SSTs between 299 and 300.5 K, deduced as the onset of deeper convection. Peak fractions of high, deep convective, altostratus, and altocumulus clouds occur at SSTs close to 303 K, while cumulus clouds, which have lower tops, show a smooth cloud fractional peak about 2° cooler. Deep convective and other high cloud types decrease sharply above SSTs of 303 K, in accordance with previous work suggesting a narrow window of tropical deep convection. Finally, significant cloud frequency differences exist between CloudSat early morning and afternoon overpasses, suggesting a diurnal cycle of some cloud types, particularly stratocumulus, high, and deep convective clouds.
11
Unglaub, Claudia, Karoline Block, Johannes Mülmenstädt, Odran Sourdeval, and Johannes Quaas. "A new classification of satellite-derived liquid water cloud regimes at cloud scale." Atmospheric Chemistry and Physics 20, no. 4 (February 28, 2020): 2407–18. http://dx.doi.org/10.5194/acp-20-2407-2020.
Full textAbstract:
Abstract. Clouds are highly variable in time and space, affecting climate sensitivity and climate change. To study and distinguish the different influences of clouds on the climate system, it is useful to separate clouds into individual cloud regimes. In this work we present a new cloud classification for liquid water clouds at cloud scale defined using cloud parameters retrieved from combined satellite measurements from CloudSat and CALIPSO. The idea is that cloud heterogeneity is a measure that allows us to distinguish cumuliform and stratiform clouds, and cloud-base height is a measure to distinguish cloud altitude. The approach makes use of a newly developed cloud-base height retrieval. Using three cloud-base height intervals and two intervals of cloud-top variability as an inhomogeneity parameter provides six new liquid cloud classes. The results show a smooth transition between marine and continental clouds as well as between stratiform and cumuliform clouds in different latitudes at the high spatial resolution of about 20 km. Analysing the micro- and macrophysical cloud parameters from collocated combined MODIS, CloudSat and CALIPSO retrievals shows distinct characteristics for each cloud regime that are in agreement with expectation and literature. This demonstrates the usefulness of the classification.
12
Heinle, A., A. Macke, and A. Srivastav. "Automatic cloud classification of whole sky images." Atmospheric Measurement Techniques Discussions 3, no. 1 (January 27, 2010): 269–99. http://dx.doi.org/10.5194/amtd-3-269-2010.
Full textAbstract:
Abstract. The recently increasing development of whole sky imagers enables temporal and spatial high-resolution sky observations. One application already performed in most cases is the estimation of fractional sky cover. A distinction between different cloud types, however, is still in progress. Here, an automatic cloud classification algorithm is presented, based on a set of mainly statistical features describing the color as well as the texture of an image. The k-nearest-neighbour classifier is used due to its high performance in solving complex issues, simplicity of implementation and low computational complexity. Seven different sky conditions are distinguished: high thin clouds (cirrus and cirrostratus), high patched cumuliform clouds (cirrocumulus and altocumulus), stratocumulus clouds, low cumuliform clouds, thick clouds (cumulonimbus and nimbostratus), stratiform clouds and clear sky. Based on the Leave-One-Out Cross-Validation the algorithm achieves an accuracy of about 97%, outperforming previous algorithms with accuracies of at most 62%. An additional test run of random images is presented, still yielding a success rate of about 75%, or up to 88% if only "serious" errors with respect to radiation impact are considered. Reasons for the decrement in accuracy are discussed, and ideas to further improve the classification results, especially in problematic cases, are investigated.
13
Heinle, A., A. Macke, and A. Srivastav. "Automatic cloud classification of whole sky images." Atmospheric Measurement Techniques 3, no. 3 (May 6, 2010): 557–67. http://dx.doi.org/10.5194/amt-3-557-2010.
Full textAbstract:
Abstract. The recently increasing development of whole sky imagers enables temporal and spatial high-resolution sky observations. One application already performed in most cases is the estimation of fractional sky cover. A distinction between different cloud types, however, is still in progress. Here, an automatic cloud classification algorithm is presented, based on a set of mainly statistical features describing the color as well as the texture of an image. The k-nearest-neighbour classifier is used due to its high performance in solving complex issues, simplicity of implementation and low computational complexity. Seven different sky conditions are distinguished: high thin clouds (cirrus and cirrostratus), high patched cumuliform clouds (cirrocumulus and altocumulus), stratocumulus clouds, low cumuliform clouds, thick clouds (cumulonimbus and nimbostratus), stratiform clouds and clear sky. Based on the Leave-One-Out Cross-Validation the algorithm achieves an accuracy of about 97%. In addition, a test run of random images is presented, still outperforming previous algorithms by yielding a success rate of about 75%, or up to 88% if only "serious" errors with respect to radiation impact are considered. Reasons for the decrement in accuracy are discussed, and ideas to further improve the classification results, especially in problematic cases, are investigated.
14
Costa, Anja, Jessica Meyer, Armin Afchine, Anna Luebke, Gebhard Günther, James R. Dorsey, Martin W. Gallagher, et al. "Classification of Arctic, midlatitude and tropical clouds in the mixed-phase temperature regime." Atmospheric Chemistry and Physics 17, no. 19 (October 13, 2017): 12219–38. http://dx.doi.org/10.5194/acp-17-12219-2017.
Full textAbstract:
Abstract. The degree of glaciation of mixed-phase clouds constitutes one of the largest uncertainties in climate prediction. In order to better understand cloud glaciation, cloud spectrometer observations are presented in this paper, which were made in the mixed-phase temperature regime between 0 and −38 °C (273 to 235 K), where cloud particles can either be frozen or liquid. The extensive data set covers four airborne field campaigns providing a total of 139 000 1 Hz data points (38.6 h within clouds) over Arctic, midlatitude and tropical regions. We develop algorithms, combining the information on number concentration, size and asphericity of the observed cloud particles to classify four cloud types: liquid clouds, clouds in which liquid droplets and ice crystals coexist, fully glaciated clouds after the Wegener–Bergeron–Findeisen process and clouds where secondary ice formation occurred. We quantify the occurrence of these cloud groups depending on the geographical region and temperature and find that liquid clouds dominate our measurements during the Arctic spring, while clouds dominated by the Wegener–Bergeron–Findeisen process are most common in midlatitude spring. The coexistence of liquid water and ice crystals is found over the whole mixed-phase temperature range in tropical convective towers in the dry season. Secondary ice is found at midlatitudes at −5 to −10 °C (268 to 263 K) and at higher altitudes, i.e. lower temperatures in the tropics. The distribution of the cloud types with decreasing temperature is shown to be consistent with the theory of evolution of mixed-phase clouds. With this study, we aim to contribute to a large statistical database on cloud types in the mixed-phase temperature regime.
15
Calbó, Josep, and Jeff Sabburg. "Feature Extraction from Whole-Sky Ground-Based Images for Cloud-Type Recognition." Journal of Atmospheric and Oceanic Technology 25, no. 1 (January 1, 2008): 3–14. http://dx.doi.org/10.1175/2007jtecha959.1.
Full textAbstract:
Abstract Several features that can be extracted from digital images of the sky and that can be useful for cloud-type classification of such images are presented. Some features are statistical measurements of image texture, some are based on the Fourier transform of the image and, finally, others are computed from the image where cloudy pixels are distinguished from clear-sky pixels. The use of the most suitable features in an automatic classification algorithm is also shown and discussed. Both the features and the classifier are developed over images taken by two different camera devices, namely, a total sky imager (TSI) and a whole sky imager (WSC), which are placed in two different areas of the world (Toowoomba, Australia; and Girona, Spain, respectively). The performance of the classifier is assessed by comparing its image classification with an a priori classification carried out by visual inspection of more than 200 images from each camera. The index of agreement is 76% when five different sky conditions are considered: clear, low cumuliform clouds, stratiform clouds (overcast), cirriform clouds, and mottled clouds (altocumulus, cirrocumulus). Discussion on the future directions of this research is also presented, regarding both the use of other features and the use of other classification techniques.
16
Wang, Lei, Zhiyong Zhang, Xiaonan Li, and Yueshun He. "MFOC-CliqueNet: A CliqueNet-Based Optimal Combination of Multidimensional Features Classification Method for Large-Scale Laser Point Clouds." Computational Intelligence and Neuroscience 2022 (August 11, 2022): 1–11. http://dx.doi.org/10.1155/2022/2446212.
Full textAbstract:
As large-scale laser 3D point clouds data contains massive and complex data, it faces great challenges in the automatic intelligent processing and classification of large-scale 3D point clouds. Aiming at the problem that 3D point clouds in complex scenes are self-occluded or occluded, which could reduce the object classification accuracy, we propose a multidimension feature optimal combination classification method named MFOC-CliqueNet based on CliqueNet for large-scale laser point clouds. The optimal combination matrix of multidimension features is constructed by extracting the three-dimensional features and multidirectional two-dimension features of 3D point cloud. This is the first time that multidimensional optimal combination features are introduced into cyclic convolutional networks CliqueNet. It is important for large-scale 3D point cloud classification. The experimental results show that the MFOC-CliqueNet framework can realize the latest level with fewer parameters. The experiments on the Large-Scale Scene Point Cloud Oakland dataset show that the classification accuracy of our method is 98.9%, which is better than other classification algorithms mentioned in this paper.
17
Barnefske, E., and H. Sternberg. "PCCT: A POINT CLOUD CLASSIFICATION TOOL TO CREATE 3D TRAINING DATA TO ADJUST AND DEVELOP 3D CONVNET." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-2/W16 (September 17, 2019): 35–40. http://dx.doi.org/10.5194/isprs-archives-xlii-2-w16-35-2019.
Full textAbstract:
<p><strong>Abstract.</strong> Point clouds give a very detailed and sometimes very accurate representation of the geometry of captured objects. In surveying, point clouds captured with laser scanners or camera systems are an intermediate result that must be processed further. Often the point cloud has to be divided into regions of similar types (object classes) for the next process steps. These classifications are very time-consuming and cost-intensive compared to acquisition. In order to automate this process step, conventional neural networks (ConvNet), which take over the classification task, are investigated in detail. In addition to the network architecture, the classification performance of a ConvNet depends on the training data with which the task is learned. This paper presents and evaluates the point clould classification tool (PCCT) developed at HCU Hamburg. With the PCCT, large point cloud collections can be semi-automatically classified. Furthermore, the influence of erroneous points in three-dimensional point clouds is investigated. The network architecture PointNet is used for this investigation.</p>
18
Nachamkin, Jason E., Adam Bienkowski, Rich Bankert, Krishna Pattipati, David Sidoti, Melinda Surratt, Jacob Gull, and Chuyen Nguyen. "Classification and Evaluation of Stable and Unstable Cloud Forecasts." Monthly Weather Review 150, no. 1 (January 2021): 81–98. http://dx.doi.org/10.1175/mwr-d-21-0056.1.
Full textAbstract:
Abstract A physics-based cloud identification scheme, originally developed for a machine-learning forecast system, was applied to verify cloud location and coverage bias errors from two years of 6-h forecasts. The routine identifies stable and unstable environments by assessing the potential for buoyant versus stable cloud formation. The efficacy of the scheme is documented by investigating its ability to identify cloud patterns and systematic forecast errors. Results showed that stable cloud forecasts contained widespread, persistent negative cloud cover biases most likely associated with turbulent, radiative, and microphysical feedback processes. In contrast, unstable clouds were better predicted despite being poorly resolved. This suggests that scale aliasing, while energetically problematic, results in less-severe short-term cloud cover errors. This study also evaluated Geostationary Operational Environmental Satellite (GOES) cloud-base retrievals for their effectiveness at identifying regions of lower-tropospheric cloud cover. Retrieved cloud-base heights were sometimes too high with respect to their actual values in regions of deep-layered clouds, resulting in underestimates of the extent of low cloud cover in these areas. Sensitivity experiments indicate that the most accurate cloud-base estimates existed in regions with cloud tops at or below 8 km. Significance Statement Cloud forecasts are difficult to verify because the height, depth, and type of the clouds are just as important as the spatial location. Satellite imagery and retrievals are good for verifying location, but these measurements are sometimes uncertain because of obscuration from above. Despite these uncertainties, we can learn a lot about specific forecast errors by tracking general areas of clouds based on their physical forcing mechanisms. We chose to sort by atmospheric stability because buoyant and stable processes are physically very distinct. Studies of this nature exist, but they typically assess mean cloud frequencies without considering spatial and temporal displacements. Here, we address displacement error by assessing the direct overlap between the observed and predicted clouds.
19
Politz, F., and M. Sester. "JOINT CLASSIFICATION OF ALS AND DIM POINT CLOUDS." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-2/W13 (June 5, 2019): 1113–20. http://dx.doi.org/10.5194/isprs-archives-xlii-2-w13-1113-2019.
Full textAbstract:
<p><strong>Abstract.</strong> National mapping agencies (NMAs) have to acquire nation-wide Digital Terrain Models on a regular basis as part of their obligations to provide up-to-date data. Point clouds from Airborne Laser Scanning (ALS) are an important data source for this task; recently, NMAs also started deriving Dense Image Matching (DIM) point clouds from aerial images. As a result, NMAs have both point cloud data sources available, which they can exploit for their purposes. In this study, we investigate the potential of transfer learning from ALS to DIM data, so the time consuming step of data labelling can be reduced. Due to their specific individual measurement techniques, both point clouds have various distinct properties such as RGB or intensity values, which are often exploited for classification of either ALS or DIM point clouds. However, those features also hinder transfer learning between these two point cloud types, since they do not exist in the other point cloud type. As the mere 3D point is available in both point cloud types, we focus on transfer learning from an ALS to a DIM point cloud using exclusively the point coordinates. We are tackling the issue of different point densities by rasterizing the point cloud into a 2D grid and take important height features as input for classification. We train an encoder-decoder convolutional neural network with labelled ALS data as a baseline and then fine-tune this baseline with an increasing amount of labelled DIM data. We also train the same network exclusively on all available DIM data as reference to compare our results. We show that only 10% of labelled DIM data increase the classification results notably, which is especially relevant for practical applications.</p>
20
Wagner, T., A. Apituley, S. Beirle, S. Dörner, U. Friess, J. Remmers, and R. Shaiganfar. "Cloud detection and classification based on MAX-DOAS observations." Atmospheric Measurement Techniques 7, no. 5 (May 19, 2014): 1289–320. http://dx.doi.org/10.5194/amt-7-1289-2014.
Full textAbstract:
Abstract. Multi-axis differential optical absorption spectroscopy (MAX-DOAS) observations of aerosols and trace gases can be strongly influenced by clouds. Thus, it is important to identify clouds and characterise their properties. In this study we investigate the effects of clouds on several quantities which can be derived from MAX-DOAS observations, like radiance, the colour index (radiance ratio at two selected wavelengths), the absorption of the oxygen dimer O4 and the fraction of inelastically scattered light (Ring effect). To identify clouds, these quantities can be either compared to their corresponding clear-sky reference values, or their dependencies on time or viewing direction can be analysed. From the investigation of the temporal variability the influence of clouds can be identified even for individual measurements. Based on our investigations we developed a cloud classification scheme, which can be applied in a flexible way to MAX-DOAS or zenith DOAS observations: in its simplest version, zenith observations of the colour index are used to identify the presence of clouds (or high aerosol load). In more sophisticated versions, other quantities and viewing directions are also considered, which allows subclassifications like, e.g., thin or thick clouds, or fog. We applied our cloud classification scheme to MAX-DOAS observations during the Cabauw intercomparison campaign of Nitrogen Dioxide measuring instruments (CINDI) campaign in the Netherlands in summer 2009 and found very good agreement with sky images taken from the ground and backscatter profiles from a lidar.
21
Özdemir, E., and F. Remondino. "CLASSIFICATION OF AERIAL POINT CLOUDS WITH DEEP LEARNING." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-2/W13 (June 4, 2019): 103–10. http://dx.doi.org/10.5194/isprs-archives-xlii-2-w13-103-2019.
Full textAbstract:
<p><strong>Abstract.</strong> Due to their usefulness in various implementations, such as energy evaluation, visibility analysis, emergency response, 3D cadastre, urban planning, change detection, navigation, etc., 3D city models have gained importance over the last decades. Point clouds are one of the primary data sources for the generation of realistic city models. Beside model-driven approaches, 3D building models can be directly produced from classified aerial point clouds. This paper presents an ongoing research for 3D building reconstruction based on the classification of aerial point clouds without given ancillary data (e.g. footprints, etc.). The work includes a deep learning approach based on specific geometric features extracted from the point cloud. The methodology was tested on the ISPRS 3D Semantic Labeling Contest (Vaihingen and Toronto point clouds) showing promising results, although partly affected by the low density and lack of points on the building facades for the available clouds.</p>
22
Pîrloagă, Răzvan, Dragoş Ene, Mihai Boldeanu, Bogdan Antonescu, Ewan J. O’Connor, and Sabina Ştefan. "Ground-Based Measurements of Cloud Properties at the Bucharest–Măgurele Cloudnet Station: First Results." Atmosphere 13, no. 9 (September 6, 2022): 1445. http://dx.doi.org/10.3390/atmos13091445.
Full textAbstract:
Data collected over a period of 18 months (December 2019–May 2021) at the Bucharest–Măgurele Cloudnet station were analysed for the first time to determine the macrophysical and microphysical cloud properties over this site. A total number of 1,327,680 vertical profiles containing the target classification based on the Cloudnet algorithm were analysed, of which 1,077,858 profiles contained hydrometeors. The highest number of profiles with hydrometeors (>60%) was recorded in December 2020, with hydrometeors being observed mainly below 5 km. Above 5 km, the frequency of occurrence of hydrometeors was less than <20%. Based on the initial Cloudnet target classification, a cloud classification scheme was implemented. Clouds were more frequently observed during winter compared with other seasons (45% of all profiles). Ice clouds were the most frequent type of cloud (468,463 profiles) during the study period, followed by mixed phases (220,280 profiles) and mixed phased precipitable clouds (164,868 profiles). The geometrical thickness varied from a median value of 244 m for liquid clouds during summer to 3362 m for mix phased precipitable clouds during spring.
23
Tang, Qingyun, Letan Zhang, Guiwen Lan, Xiaoyong Shi, Xinghui Duanmu, and Kan Chen. "A Classification Method of Point Clouds of Transmission Line Corridor Based on Improved Random Forest and Multi-Scale Features." Sensors 23, no. 3 (January 24, 2023): 1320. http://dx.doi.org/10.3390/s23031320.
Full textAbstract:
Classification of airborne laser scanning (ALS) point clouds of power lines is of great importance to their reconstruction. However, it is still a difficult task to efficiently and accurately classify the ground, vegetation, power lines and power pylons from ALS point clouds. Therefore, in this paper, a method is proposed to improve the accuracy and efficiency of the classification of point clouds of transmission lines, which is based on improved Random Forest and multi-scale features. The point clouds are filtered by the optimized progressive TIN densification filtering algorithm, then the elevations of the filtered point cloud are normalized. The features of the point cloud at different scales are calculated according to the basic features of the point cloud and the characteristics of transmission lines. The Relief F and Sequential Backward Selection algorithm are used to select the best subset of features to estimate the parameters of the learning model, then an Improved Random Forest classification model is built to classify the point clouds. The proposed method is verified by using three different samples from the study area and the results show that, compared with the methods based on Support Vector Machines, AdaBoost or Random Forest, our method can reduce feature redundancy and has higher classification accuracy and efficiency.
24
Hao, Qi, Wenguang Zheng, and Yingyuan Xiao. "Fusion Information Multi-View Classification Method for Remote Sensing Cloud Detection." Applied Sciences 12, no. 14 (July 20, 2022): 7295. http://dx.doi.org/10.3390/app12147295.
Full textAbstract:
In recent years, many studies have been carried out to detect clouds on remote sensing images. Due to the complex terrain, the variety of clouds, the density, and content of clouds are various, and the current model has difficulty accurately detecting the cloud in the image. In our strategy, a multi-view data training set based on super pixel is constructed. View A uses multi-level network to extract the boundary, texture, and deep abstract feature of super pixels. View B is the statistical feature of the three channels of the image. Privilege information View P contains the cloud content of super pixels and the tag status of adjacent super pixels. Finally, we propose a cloud detection method for remote sensing image classification based on multi-view support vector machine (SVM). The proposed method is tested on images of different terrain and cloud distribution in GF-1_WHU and Cloud-38 remote sensing datasets. Visual performance and quantitative analysis show that the method has excellent cloud detection performance.
25
Keckhut, P., F. Borchi, S. Bekki, A. Hauchecorne, and M. SiLaouina. "Cirrus Classification at Midlatitude from Systematic Lidar Observations." Journal of Applied Meteorology and Climatology 45, no. 2 (February 1, 2006): 249–58. http://dx.doi.org/10.1175/jam2348.1.
Full textAbstract:
Abstract Systematic cirrus lidar measurements performed in the south of France during 2000 are analyzed statistically to search for cloud classes. The classes are based on cloud characteristics (cloud thickness, light backscattering efficiency, and its variance), cloud absolute geometric height, cloud height relative to the tropopause, and the temperature at the cloud level. The successive use of principal component analysis, cluster methods, and linear discriminant analysis allows the identification of four cirrus classes. Almost all the cirrus detections correspond to three classes with similar proportion of the total cirrus detected (around 30%). The absolute geometric height and the thickness are found to be the main discriminant variables. The first cirrus class corresponds to thin clouds above the local tropopause (absolute geometric height: 11.5 km), or at least around the tropopause, while another class corresponds also to thin clouds but at a lower altitude range in the troposphere (absolute geometric height: 8.6 km). The third class corresponds to thick clouds (thickness of 3.2 km) located below the tropopause, in an altitude range between the two first classes (absolute geometric height: 9.8 km). As expected, the high-altitude cirrus class is characterized with the lowest mean temperature. It is noted that the temperature is closely related to the altitude and so the role of temperature in the cirrus classes cannot be disentangled from the role of the altitude.
26
Ling, Jing, Hongsheng Zhang, and Yinyi Lin. "Improving Urban Land Cover Classification in Cloud-Prone Areas with Polarimetric SAR Images." Remote Sensing 13, no. 22 (November 21, 2021): 4708. http://dx.doi.org/10.3390/rs13224708.
Full textAbstract:
Urban land cover (ULC) serves as fundamental environmental information for urban studies, while accurate and timely ULC mapping remains challenging due to cloud contamination in tropical and subtropical areas. Synthetic aperture radar (SAR) has excellent all-weather working capability to overcome the challenge, while optical SAR data fusion is often required due to the limited land surface information provided by SAR. However, the mechanism by which SAR can compensate optical images, given the occurrence of clouds, in order to improve the ULC mapping, remains unexplored. To address the issue, this study proposes a framework, through various sampling strategies and three typical supervised classification methods, to quantify the ULC classification accuracy using optical and SAR data with various cloud levels. The land cover confusions were investigated in detail to understand the role of SAR in distinguishing land cover under different types of cloud coverage. Several interesting experimental results were found. First, 50% cloud coverage over the optical images decreased the overall accuracy by 10–20%, while the incorporation of SAR images was able to improve the overall accuracy by approximately 4%, by increasing the recognition of cloud-covered ULC information, particularly the water bodies. Second, if all the training samples were not contaminated by clouds, the cloud coverage had a higher impact with a reduction of 35% in the overall accuracy, whereas the incorporation of SAR data contributed to an increase of approximately 5%. Third, the thickness of clouds also brought about different impacts on the results, with an approximately 10% higher reduction from thick clouds compared with that from thin clouds, indicating that certain spectral information might still be available in the areas covered by thin clouds. These findings provide useful references for the accurate monitoring of ULC over cloud-prone areas, such as tropical and subtropical cities, where cloud contamination is often unavoidable.
27
Duran, Zaide, Kubra Ozcan, and Muhammed Enes Atik. "Classification of Photogrammetric and Airborne LiDAR Point Clouds Using Machine Learning Algorithms." Drones 5, no. 4 (September 24, 2021): 104. http://dx.doi.org/10.3390/drones5040104.
Full textAbstract:
With the development of photogrammetry technologies, point clouds have found a wide range of use in academic and commercial areas. This situation has made it essential to extract information from point clouds. In particular, artificial intelligence applications have been used to extract information from point clouds to complex structures. Point cloud classification is also one of the leading areas where these applications are used. In this study, the classification of point clouds obtained by aerial photogrammetry and Light Detection and Ranging (LiDAR) technology belonging to the same region is performed by using machine learning. For this purpose, nine popular machine learning methods have been used. Geometric features obtained from point clouds were used for the feature spaces created for classification. Color information is also added to these in the photogrammetric point cloud. According to the LiDAR point cloud results, the highest overall accuracies were obtained as 0.96 with the Multilayer Perceptron (MLP) method. The lowest overall accuracies were obtained as 0.50 with the AdaBoost method. The method with the highest overall accuracy was achieved with the MLP (0.90) method. The lowest overall accuracy method is the GNB method with 0.25 overall accuracy.
28
Marinou, Eleni, Kalliopi Artemis Voudouri, Ioanna Tsikoudi, Eleni Drakaki, Alexandra Tsekeri, Marco Rosoldi, Dragos Ene, et al. "Geometrical and Microphysical Properties of Clouds Formed in the Presence of Dust above the Eastern Mediterranean." Remote Sensing 13, no. 24 (December 9, 2021): 5001. http://dx.doi.org/10.3390/rs13245001.
Full textAbstract:
In this work, collocated lidar–radar observations are used to retrieve the vertical profiles of cloud properties above the Eastern Mediterranean. Measurements were performed in the framework of the PRE-TECT experiment during April 2017 at the Greek atmospheric observatory of Finokalia, Crete. Cloud geometrical and microphysical properties at different altitudes were derived using the Cloudnet target classification algorithm. We found that the variable atmospheric conditions that prevailed above the region during April 2017 resulted in complex cloud structures. Mid-level clouds were observed in 38% of the cases, high or convective clouds in 58% of the cases, and low-level clouds in 2% of the cases. From the observations of cloudy profiles, pure ice phase occurred in 94% of the cases, mixed-phase clouds were observed in 27% of the cases, and liquid clouds were observed in 8.7% of the cases, while Drizzle or rain occurred in 12% of the cases. The significant presence of Mixed-Phase Clouds was observed in all the clouds formed at the top of a dust layer, with three times higher abundance than the mean conditions (26% abundance at −15 °C). The low-level clouds were formed in the presence of sea salt and continental particles with ice abundance below 30%. The derived statistics on clouds’ high-resolution vertical distributions and thermodynamic phase can be combined with Cloudnet cloud products and lidar-retrieved aerosol properties to study aerosol-cloud interactions in this understudied region and evaluate microphysics parameterizations in numerical weather prediction and global climate models.
29
Aires, Filipe, Francis Marquisseau, Catherine Prigent, and Geneviève Sèze. "A Land and Ocean Microwave Cloud Classification Algorithm Derived from AMSU-A and -B, Trained Using MSG-SEVIRI Infrared and Visible Observations." Monthly Weather Review 139, no. 8 (August 2011): 2347–66. http://dx.doi.org/10.1175/mwr-d-10-05012.1.
Full textAbstract:
AbstractA statistical cloud classification and cloud mask algorithm is developed based on Advanced Microwave Sounding Unit (AMSU-A and -B) microwave (MW) observations. The visible and infrared data from the Meteosat Third Generation-Spinning Enhanced Visible and Infrared Imager (MSG-SEVIRI) are used to train the microwave classifier. The goal of the MW algorithms is not to fully reproduce this MSG-SEVIRI cloud classification, as the MW observations do not have enough information on clouds to reach this level of precision. The objective is instead to obtain a stand-alone MW cloud mask and classification algorithm that can be used efficiently in forthcoming retrieval schemes of surface or atmospheric parameters from microwave satellite observations. This is an important tool over both ocean and land since the assimilation of the MW observations in the operational centers is independent from the other satellite observations.Clear sky and low, medium, and opaque–high clouds can be retrieved over ocean and land at a confidence level of more than 80%. An information content analysis shows that AMSU-B provides significant information over both land and ocean, especially for the classification of medium and high clouds, whereas AMSU-A is more efficient over ocean when discriminating clear situations and low clouds.
30
Brakhasi, F., M. Hajeb, and F. Fouladinejad. "DISCRIMINATION AEROSOL FORM CLOUDS USING CATS-ISS LIDAR OBSERVATIONS BASED ON RANDOM FOREST AND SVM ALGORITHMS OVER THE EASTERN PART OF MIDDLE EAST." ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences XLII-4/W18 (October 18, 2019): 235–40. http://dx.doi.org/10.5194/isprs-archives-xlii-4-w18-235-2019.
Full textAbstract:
Abstract. Aerosols and Clouds play an important role in the Earth's environment, climate change and climate models. The Cloud-Aerosol Transport System (CATS) as a lidar remote sensing instrument, from the International Space Station (ISS), provides range-resolved profile measurements of atmospheric aerosols and clouds. Discrimination aerosols from clouds have always been a challenges task in the classification of space-born lidars. In this study, two algorithms including Random Forest (RF) and Support Vector Machine (SVM) were used to tackle the problem in a nighttime lidar data from 18 October 2016 which passes form the western part of Iran. The procedure includes 3 stages preprocessing (improving the signal to noise, generating features, taking training sample), classification (implementing RF and SVM), and postprocessing (correcting misleading classification). Finally, the result of classifications of the two algorithms (RF-SVM) were compared against ground truth samples and Vertical Feature Mask (VFM) of CATS product indicated 0.96–0.94 and 0.88–0.88 respectively. Also, it should be mentioned that a kappa accuracy 0.88 was acquired when we compared VFM against our ground truth samples. Moreover, a visual comparison with Moderate Resolution Imaging Spectroradiometer (MODIS) AOD and RGB products demonstrating that clouds and aerosol can be well detected and discriminated. The experimental results elucidated that the proposed method for classification of space borne lidar observation leads to higher accuracy compared to PDFs based algorithms.
31
Ghonima, M. S., B. Urquhart, C. W. Chow, J. E. Shields, A. Cazorla, and J. Kleissl. "A method for cloud detection and opacity classification based on ground based sky imagery." Atmospheric Measurement Techniques Discussions 5, no. 4 (July 2, 2012): 4535–69. http://dx.doi.org/10.5194/amtd-5-4535-2012.
Full textAbstract:
Abstract. Digital images of the sky obtained using a total sky imager (TSI) are classified pixel by pixel into clear sky, optically thin and optically thick clouds. A new classification algorithm was developed that compares the pixel red-blue ratio (RBR) to the RBR of a clear sky library (CSL) generated from images captured on clear days. The difference, rather than the ratio, between pixel RBR and CSL RBR resulted in more accurate cloud classification. High correlation between TSI image RBR and aerosol optical depth (AOD) measured by an AERONET photometer was observed and motivated the addition of a haze correction factor (HCF) to the classification model to account for variations in AOD. Thresholds for clear and thick clouds were chosen based on a training image set and validated with set of manually annotated images. Misclassifications of clear and thick clouds into the opposite category were less than 1%. Thin clouds were classified with an accuracy of 60%. Accurate cloud detection and opacity classification techniques will improve the accuracy of short-term solar power forecasting.
32
Ghonima, M. S., B. Urquhart, C. W. Chow, J. E. Shields, A. Cazorla, and J. Kleissl. "A method for cloud detection and opacity classification based on ground based sky imagery." Atmospheric Measurement Techniques 5, no. 11 (November 27, 2012): 2881–92. http://dx.doi.org/10.5194/amt-5-2881-2012.
Full textAbstract:
Abstract. Digital images of the sky obtained using a total sky imager (TSI) are classified pixel by pixel into clear sky, optically thin and optically thick clouds. A new classification algorithm was developed that compares the pixel red-blue ratio (RBR) to the RBR of a clear sky library (CSL) generated from images captured on clear days. The difference, rather than the ratio, between pixel RBR and CSL RBR resulted in more accurate cloud classification. High correlation between TSI image RBR and aerosol optical depth (AOD) measured by an AERONET photometer was observed and motivated the addition of a haze correction factor (HCF) to the classification model to account for variations in AOD. Thresholds for clear and thick clouds were chosen based on a training image set and validated with set of manually annotated images. Misclassifications of clear and thick clouds into the opposite category were less than 1%. Thin clouds were classified with an accuracy of 60%. Accurate cloud detection and opacity classification techniques will improve the accuracy of short-term solar power forecasting.
33
Sirmacek, B., and R. Lindenbergh. "AUTOMATIC CLASSIFICATION OF TREES FROM LASER SCANNING POINT CLOUDS." ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences II-3/W5 (August 19, 2015): 137–44. http://dx.doi.org/10.5194/isprsannals-ii-3-w5-137-2015.
Full textAbstract:
Development of laser scanning technologies has promoted tree monitoring studies to a new level, as the laser scanning point clouds enable accurate 3D measurements in a fast and environmental friendly manner. In this paper, we introduce a probability matrix computation based algorithm for automatically classifying laser scanning point clouds into ’tree’ and ’non-tree’ classes. Our method uses the 3D coordinates of the laser scanning points as input and generates a new point cloud which holds a label for each point indicating if it belongs to the ’tree’ or ’non-tree’ class. To do so, a grid surface is assigned to the lowest height level of the point cloud. The grids are filled with probability values which are calculated by checking the point density above the grid. Since the tree trunk locations appear with very high values in the probability matrix, selecting the local maxima of the grid surface help to detect the tree trunks. Further points are assigned to tree trunks if they appear in the close proximity of trunks. Since heavy mathematical computations (such as point cloud organization, detailed shape 3D detection methods, graph network generation) are not required, the proposed algorithm works very fast compared to the existing methods. The tree classification results are found reliable even on point clouds of cities containing many different objects. As the most significant weakness, false detection of light poles, traffic signs and other objects close to trees cannot be prevented. Nevertheless, the experimental results on mobile and airborne laser scanning point clouds indicate the possible usage of the algorithm as an important step for tree growth observation, tree counting and similar applications. While the laser scanning point cloud is giving opportunity to classify even very small trees, accuracy of the results is reduced in the low point density areas further away than the scanning location. These advantages and disadvantages of two laser scanning point cloud sources are discussed in detail.
34
LAI, CARMEN, DAVID M. J. TAX, ROBERT P. W. DUIN, ELŻBIETA PĘKALSKA, and PAVEL PACLÍK. "A STUDY ON COMBINING IMAGE REPRESENTATIONS FOR IMAGE CLASSIFICATION AND RETRIEVAL." International Journal of Pattern Recognition and Artificial Intelligence 18, no. 05 (August 2004): 867–90. http://dx.doi.org/10.1142/s0218001404003459.
Full textAbstract:
A flexible description of images is offered by a cloud of points in a feature space. In the context of image retrieval such clouds can be represented in a number of ways. Two approaches are here considered. The first approach is based on the assumption of a normal distribution, hence homogeneous clouds, while the second one focuses on the boundary description, which is more suitable for multimodal clouds. The images are then compared either by using the Mahalanobis distance or by the support vector data description (SVDD), respectively. The paper investigates some possibilities of combining the image clouds based on the idea that responses of several cloud descriptions may convey a pattern, specific for semantically similar images. A ranking of image dissimilarities is used as a comparison for two image databases targeting image classification and retrieval problems. We show that combining of the SVDD descriptions improves the retrieval performance with respect to ranking, on the contrary to the Mahalanobis case. Surprisingly, it turns out that the ranking of the Mahalanobis distances works well also for inhomogeneous images.
35
Peterson, Michael, Scott Rudlosky, and Daile Zhang. "Thunderstorm Cloud-Type Classification from Space-Based Lightning Imagers." Monthly Weather Review 148, no. 5 (April 13, 2020): 1891–98. http://dx.doi.org/10.1175/mwr-d-19-0365.1.
Full textAbstract:
Abstract The organization and structure of thunderstorms determines the extent and severity of their hazards to the general public and their consequences for the Earth system. Distinguishing vigorous convective regions that produce heavy rain and hail from adjacent regions of stratiform clouds or overhanging anvil clouds that produce light to no rainfall is valuable in operations and physical research. Cloud-type algorithms that partition convection from stratiform regions have been developed for space-based radar, passive microwave, and now Geostationary Operational Environmental Satellites (GOES) Advanced Baseline Imager (ABI) multispectral products. However, there are limitations for each of these products including temporal availability, spatial coverage, and the degree to which they based on cloud microphysics. We have developed a cloud-type algorithm for GOES Geostationary Lightning Mapper (GLM) observations that identifies convective/nonconvective regions in thunderstorms based on signatures of interactions with nonconvective charge structures in the lightning flash data. The GLM sensor permits a rapid (20 s) update cycle over the combined GOES-16–GOES-17 domain across all hours of the day. Storm regions that do not produce lightning will not be classified by our algorithm, however. The GLM cloud-type product is intended to provide situational awareness of electrified nonconvective clouds and to complement other cloud-type retrievals by providing a contemporary assessment tied to lightning physics. We propose that a future combined ABI–GLM cloud-type algorithm would be a valuable product that could draw from the strengths of each instrument and approach.
36
Sirch, Tobias, Luca Bugliaro, Tobias Zinner, Matthias Möhrlein, and Margarita Vazquez-Navarro. "Cloud and DNI nowcasting with MSG/SEVIRI for the optimized operation of concentrating solar power plants." Atmospheric Measurement Techniques 10, no. 2 (February 2, 2017): 409–29. http://dx.doi.org/10.5194/amt-10-409-2017.
Full textAbstract:
Abstract. A novel approach for the nowcasting of clouds and direct normal irradiance (DNI) based on the Spinning Enhanced Visible and Infrared Imager (SEVIRI) aboard the geostationary Meteosat Second Generation (MSG) satellite is presented for a forecast horizon up to 120 min. The basis of the algorithm is an optical flow method to derive cloud motion vectors for all cloudy pixels. To facilitate forecasts over a relevant time period, a classification of clouds into objects and a weighted triangular interpolation of clear-sky regions are used. Low and high level clouds are forecasted separately because they show different velocities and motion directions. Additionally a distinction in advective and convective clouds together with an intensity correction for quickly thinning convective clouds is integrated. The DNI is calculated from the forecasted optical thickness of the low and high level clouds. In order to quantitatively assess the performance of the algorithm, a forecast validation against MSG/SEVIRI observations is performed for a period of 2 months. Error rates and Hanssen–Kuiper skill scores are derived for forecasted cloud masks. For a forecast of 5 min for most cloud situations more than 95 % of all pixels are predicted correctly cloudy or clear. This number decreases to 80–95 % for a forecast of 2 h depending on cloud type and vertical cloud level. Hanssen–Kuiper skill scores for cloud mask go down to 0.6–0.7 for a 2 h forecast. Compared to persistence an improvement of forecast horizon by a factor of 2 is reached for all forecasts up to 2 h. A comparison of forecasted optical thickness distributions and DNI against observations yields correlation coefficients larger than 0.9 for 15 min forecasts and around 0.65 for 2 h forecasts.
37
Khain, A. P., N. BenMoshe, and A. Pokrovsky. "Factors Determining the Impact of Aerosols on Surface Precipitation from Clouds: An Attempt at Classification." Journal of the Atmospheric Sciences 65, no. 6 (June 1, 2008): 1721–48. http://dx.doi.org/10.1175/2007jas2515.1.
Full textAbstract:
Abstract The simulation of the dynamics and the microphysics of clouds observed during the Large-Scale Biosphere–Atmosphere Experiment in Amazonia—Smoke, Aerosols, Clouds, Rainfall, and Climate (LBA–SMOCC) campaign, as well as extremely continental and extremely maritime clouds, is performed using an updated version of the Hebrew University spectral microphysics cloud model (HUCM). A new scheme of diffusional growth allows the reproduction of in situ–measured droplet size distributions including those formed in extremely polluted air. It was shown that pyroclouds forming over the forest fires can precipitate. Several mechanisms leading to formation of precipitation from pyroclouds are considered. The mechanisms by which aerosols affect the microphysics and precipitation of warm cloud-base clouds have been investigated by analyzing the mass, heat, and moisture budgets. The increase in aerosol concentration increases both the generation and the loss of the condensate mass. In the clouds developing in dry air, the increase in the loss is dominant, which suggests a decrease in the accumulated precipitation with the aerosol concentration increase. On the contrary, an increase in aerosol concentration in deep maritime clouds leads to an increase in precipitation. The precipitation efficiency of clouds in polluted air is found to be several times lower than that of clouds forming in clean air. A classification of the results of aerosol effects on precipitation from clouds of different types developing in the atmosphere with high freezing level (about 4 km) is proposed. The role of air humidity and other factors in precipitation’s response to aerosols is discussed. The analysis shows that many discrepancies between the results reported in different observational and numerical studies can be attributed to the different atmospheric conditions and cloud types analyzed.
38
Zhang, Zheng, Zhiwei Xu, Chang’an Liu, Qing Tian, and Yanping Wang. "Cloudformer: Supplementary Aggregation Feature and Mask-Classification Network for Cloud Detection." Applied Sciences 12, no. 7 (March 22, 2022): 3221. http://dx.doi.org/10.3390/app12073221.
Full textAbstract:
Cloud detection is an important step in the processing of optical satellite remote-sensing data. In recent years, deep learning methods have achieved excellent results in cloud detection tasks. However, most of the current models have difficulties to accurately classify similar objects (e.g., clouds and snow) and to accurately detect clouds that occupy a few pixels in an image. To solve these problems, a cloud-detection framework (Cloudformer) combining CNN and Transformer is being proposed to achieve high-precision cloud detection in optical remote-sensing images. The framework achieves accurate detection of thin and small clouds using a pyramidal structure encoder. It also achieves accurate classification of similar objects using a dual-path decoder structure of CNN and Transformer, reducing the rate of missed detections and false alarms. In addition, since the Transformer model lacks the perception of location information, an asynchronous position-encoding method is being proposed to enhance the position information of the data entering the Transformer module and to optimize the detection results. Cloudformer is experimented on two datasets, AIR-CD and 38-Cloud, and the results show that it has state-of-the-art performance.
39
Liu, Bingjie, Shuxin Chen, Huaguo Huang, and Xin Tian. "Tree Species Classification of Backpack Laser Scanning Data Using the PointNet++ Point Cloud Deep Learning Method." Remote Sensing 14, no. 15 (August 7, 2022): 3809. http://dx.doi.org/10.3390/rs14153809.
Full textAbstract:
To investigate forest resources, it is necessary to identify the tree species. However, it is a challenge to identify tree species using 3D point clouds of trees collected by light detection and ranging (LiDAR). PointNet++, a point cloud deep learning network, can effectively classify 3D objects. It is important to establish high-quality individual tree point cloud datasets when applying PointNet++ to identifying tree species. However, there are different data processing methods to produce sample datasets, and the processes are tedious. In this study, we suggest how to select the appropriate method by designing comparative experiments. We used the backpack laser scanning (BLS) system to collect point cloud data for a total of eight tree species in three regions. We explored the effect of tree height on the classification accuracy of tree species by using different point cloud normalization methods and analyzed the effect of leaf point clouds on classification accuracy by separating the leaves and wood of individual tree point clouds. Five downsampling methods were used: farthest point sampling (FPS), K-means, random, grid average sampling, and nonuniform grid sampling (NGS). Data with different sampling points were designed for the experiments. The results show that the tree height feature is unimportant when using point cloud deep learning methods for tree species classification. For data collected in a single season, the leaf point cloud has little effect on the classification accuracy. The two suitable point cloud downsampling methods we screened were FPS and NGS, and the deep learning network could provide the most accurate tree species classification when the number of individual tree point clouds was in the range of 2048–5120. Our study further illustrates that point-based end-to-end deep learning methods can be used to classify tree species and identify individual tree point clouds. Combined with the low-cost and high-efficiency BLS system, it can effectively improve the efficiency of forest resource surveys.
40
Politz, F., M. Sester, and C. Brenner. "GEOMETRY-BASED POINT CLOUD CLASSIFICATION USING HEIGHT DISTRIBUTIONS." ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences V-2-2020 (August 3, 2020): 259–66. http://dx.doi.org/10.5194/isprs-annals-v-2-2020-259-2020.
Full textAbstract:
Abstract. Semantic segmentation is one of the main steps in the processing chain for Airborne Laser Scanning (ALS) point clouds, but it is also one of the most labour intensive steps, as it requires many labelled examples to train a classifier. National mapping agencies (NMAs) have to acquire nationwide ALS data every couple of years for their duties. Having point clouds cover different terrains such as flat or mountainous regions, a classifier often requires a refinement using additional data from those specific terrains. In this study, we present an algorithm, which is able to classify point clouds of similar terrain types without requiring any additional training data and which is still able to achieve overall F1-Scores of over 90% in most setups. Our algorithm uses up to two height distributions within a single cell in a rasterized point cloud. For each distribution, the empirical mean and standard deviation are calculated, which are the input for a Convolutional Neural Network (CNN) classifier. Consequently, our approach only requires the geometry of point clouds, which enables also the usage of the same network structure for point clouds from other sensor systems such as Dense Image Matching. Since the mean ground level varies with the observed area, we also examined five different normalisation methods for our input in order to reduce the ground influence on the point clouds and thus increase its transferability towards other datasets. We test our trained networks on four different tests sets with the classes’ ground, building, water, non-ground and bridge.
41
Feofilov, A. G., C. J. Stubenrauch, and J. Delanoë. "Ice water content vertical profiles of high-level clouds: classification and impact on radiative fluxes." Atmospheric Chemistry and Physics Discussions 15, no. 12 (June 16, 2015): 16325–69. http://dx.doi.org/10.5194/acpd-15-16325-2015.
Full textAbstract:
Abstract. In this work, we discuss the shape of ice water content (IWC) vertical profiles in high ice clouds and its effect on radiative properties of these clouds, both in short- and in long-wave bands (SW and LW). Based on the analysis of colocated satellite data, we suggest a minimal set of primitive shapes (rectangular, isosceles trapezoid, lower and upper triangle), which sufficiently well represents the IWC profiles. About 75% of all high-level ice clouds (P < 440 hPa) have an ice water path smaller than 100 g m−2, with a 10% smaller contribution from single layer clouds. Most IWC profiles (80%) can be represented by a rectangular or isosceles trapezoid shape. However, with increasing IWP, the number of lower triangle profiles (IWC rises towards cloud base) increases, reaching up to 40% for IWP values greater than 300 g m−2. The number of upper triangle profiles (IWC rises towards cloud top) is in general small and decreasing with IWP, with the maximum occurrence of 15% in cases of IWP less than 10 g m−2. We propose a statistical classification of the IWC shapes using ice water path (IWP) as a single parameter. We have estimated the radiative effects of clouds with the same IWP and with different IWC profile shapes for five typical atmospheric scenarios and over a broad range of IWP, cloud height, cloud vertical extent, and effective ice crystal diameter (De). We explain changes in outgoing LW fluxes at the top of the atmosphere (TOA) by the cloud thermal radiance while differences in TOA SW fluxes relate to the De vertical profile within the cloud. Absolute differences in net TOA and surface fluxes associated with these parameterized IWC profiles instead of assuming constant IWC profiles are in general of the order of 1–2 W m−2: they are negligible for clouds with IWP < 30 g m−2, but may reach 2 W m−2 for clouds with IWP >300 W m−2.
42
Araújo Palharini, Rayana Santos, and Daniel Alejandro Vila. "Climatological Behavior of Precipitating Clouds in the Northeast Region of Brazil." Advances in Meteorology 2017 (2017): 1–12. http://dx.doi.org/10.1155/2017/5916150.
Full textAbstract:
This study aims to analyze the climatological classification of precipitating clouds in the Northeast of Brazil using the radar on board the Tropical Rainfall Measuring Mission (TRMM) satellite. Thus, for this research a time series of 15 years of satellite data (period 1998–2012) was analyzed in order to identify what types of clouds produce precipitation estimated by Precipitation Radar (PR) and how often these clouds occur. From the results of this work it was possible to estimate the average relative frequency of each type of cloud present in weather systems that influence the Northeast of Brazil. In general, the stratiform clouds and shallow convective clouds are the most frequent in this region, but the associated rainfall is not as abundant as precipitation caused by deep convective clouds. It is also seen that a strong signal of shallow convective clouds modulates rainfall over the coastal areas of Northeast of Brazil and adjacent ocean. In this scenario, the main objective of this study is to contribute to a better understanding of the patterns of cloud types associated with precipitation and building a climatological analysis from the classification of clouds.
43
Wu, Hangbin, Huimin Yang, Shengyu Huang, Doudou Zeng, Chun Liu, Hao Zhang, Chi Guo, and Long Chen. "Classification of Point Clouds for Indoor Components Using Few Labeled Samples." Remote Sensing 12, no. 14 (July 8, 2020): 2181. http://dx.doi.org/10.3390/rs12142181.
Full textAbstract:
The existing deep learning methods for point cloud classification are trained using abundant labeled samples and used to test only a few samples. However, classification tasks are diverse, and not all tasks have enough labeled samples for training. In this paper, a novel point cloud classification method for indoor components using few labeled samples is proposed to solve the problem of the requirement for abundant labeled samples for training with deep learning classification methods. This method is composed of four parts: mixing samples, feature extraction, dimensionality reduction, and semantic classification. First, the few labeled point clouds are mixed with unlabeled point clouds. Next, the mixed high-dimensional features are extracted using a deep learning framework. Subsequently, a nonlinear manifold learning method is used to embed the mixed features into a low-dimensional space. Finally, the few labeled point clouds in each cluster are identified, and semantic labels are provided for unlabeled point clouds in the same cluster by a neighborhood search strategy. The validity and versatility of the proposed method were validated by different experiments and compared with three state-of-the-art deep learning methods. Our method uses fewer than 30 labeled point clouds to achieve an accuracy that is 1.89–19.67% greater than existing methods. More importantly, the experimental results suggest that this method is not only suitable for single-attribute indoor scenarios but also for comprehensive complex indoor scenarios.
44
Zhang, Chunjiao, Shenghua Xu, Tao Jiang, Jiping Liu, Zhengjun Liu, An Luo, and Yu Ma. "Integrating Normal Vector Features into an Atrous Convolution Residual Network for LiDAR Point Cloud Classification." Remote Sensing 13, no. 17 (August 29, 2021): 3427. http://dx.doi.org/10.3390/rs13173427.
Full textAbstract:
LiDAR point clouds are rich in spatial information and can effectively express the size, shape, position, and direction of objects; thus, they have the advantage of high spatial utilization. The point cloud focuses on describing the shape of the external surface of the object itself and will not store useless redundant information to describe the occupation. Therefore, point clouds have become the research focus of 3D data models and are widely used in large-scale scene reconstruction, virtual reality, digital elevation model production, and other fields. Since point clouds have various characteristics, such as disorder, density inconsistency, unstructuredness, and incomplete information, point cloud classification is still complex and challenging. To realize the semantic classification of LiDAR point clouds in complex scenarios, this paper proposes the integration of normal vector features into an atrous convolution residual network. Based on the RandLA-Net network structure, the proposed network integrates the atrous convolution into the residual module to extract global and local features of the point clouds. The atrous convolution can learn more valuable point cloud feature information by expanding the receptive field. Then, the point cloud normal vector is embedded in the local feature aggregation module of the RandLA-Net network to extract local semantic aggregation features. The improved local feature aggregation module can merge the deep features of the point cloud and mine the fine-grained information of the point cloud to improve the model’s segmentation ability in complex scenes. Finally, to resolve the imbalance of the distribution of the various categories of point clouds, the original loss function is optimized by adopting a reweighted method to prevent overfitting so that the network can focus on small target categories in the training process to effectively improve the classification performance. Through the experimental analysis of a Vaihingen (Germany) urban 3D semantic dataset from the ISPRS website, it is verified that the proposed algorithm has a strong generalization ability. The overall accuracy (OA) of the proposed algorithm on the Vaihingen urban 3D semantic dataset reached 97.9%, and the average reached 96.1%. Experiments show that the proposed algorithm fully exploits the semantic features of point clouds and effectively improves the accuracy of point cloud classification.
45
Vassel, Maiken, Luisa Ickes, Marion Maturilli, and Corinna Hoose. "Classification of Arctic multilayer clouds using radiosonde and radar data in Svalbard." Atmospheric Chemistry and Physics 19, no. 7 (April 16, 2019): 5111–26. http://dx.doi.org/10.5194/acp-19-5111-2019.
Full textAbstract:
Abstract. Multilayer clouds (MLCs) occur more often in the Arctic than globally. In this study we present the results of a detection algorithm applied to radiosonde and radar data from an 1-year time period in Ny-Ålesund, Svalbard. Multilayer cloud occurrence is found on 29 % of the investigated days. These multilayer cloud cases are further analysed regarding the possibility of ice crystal seeding, meaning that an ice crystal can survive sublimation in a subsaturated layer between two cloud layers when falling through this layer. For this we analyse profiles of relative humidity with respect to ice to identify super- and subsaturated air layers. Then the sublimation of an ice crystal of an assumed initial size of r=400 µm on its way through the subsaturated layer is calculated. If the ice crystal still exists when reaching a lower supersaturated layer, ice crystal seeding can potentially take place. Seeding cases are found often, in 23 % of the investigated days (100 % includes all days, as well as non-cloudy days). The identification of seeding cases is limited by the radar signal inside the subsaturated layer. Clearly separated multilayer clouds, defined by a clear interstice in the radar image, do not interact through seeding (9 % of the investigated days). There are various deviations between the relative humidity profiles and the radar images, e.g. due to the lack of ice-nucleating particles (INPs) and cloud condensation nuclei (CCN). Additionally, horizontal wind drift of the radiosonde and time restriction when comparing radiosonde and radar data cause further deviations. In order to account for some of these deviations, an evaluation by manual visual inspection is done for the non-seeding cases.
46
Kermarrec, Gael, Zhonglong Yang, and Daniel Czerwonka-Schröder. "Classification of Terrestrial Laser Scanner Point Clouds: A Comparison of Methods for Landslide Monitoring from Mathematical Surface Approximation." Remote Sensing 14, no. 20 (October 12, 2022): 5099. http://dx.doi.org/10.3390/rs14205099.
Full textAbstract:
Terrestrial laser scanners (TLS) are contact-free measuring sensors that record dense point clouds of objects or scenes by acquiring coordinates and an intensity value for each point. The point clouds are scattered and noisy. Performing a mathematical surface approximation instead of working directly on the point cloud is an efficient way to reduce the data storage and structure the point clouds by transforming “data” to “information”. Applications include rigorous statistical testing for deformation analysis within the context of landslide monitoring. In order to reach an optimal approximation, classification and segmentation algorithms can identify and remove inhomogeneous structures, such as trees or bushes, to obtain a smooth and accurate mathematical surface of the ground. In this contribution, we compare methods to perform the classification of TLS point clouds with the aim of guiding the reader through the existing algorithms. Besides the traditional point cloud filtering methods, we will analyze machine learning classification algorithms based on the manual extraction of point cloud features, and a deep learning approach with automatic extraction of features called PointNet++. We have intentionally chosen strategies easy to implement and understand so that our results are reproducible for similar point clouds. We show that each method has advantages and drawbacks, depending on user criteria, such as the computational time, the classification accuracy needed, whether manual extraction is performed or not, and if prior information is required. We highlight that filtering methods are advantageous for the application at hand and perform a mathematical surface approximation as an illustration. Accordingly, we have chosen locally refined B-splines, which were shown to provide an optimal and computationally manageable approximation of TLS point clouds.
47
Nazir, Danish, Muhammad Zeshan Afzal, Alain Pagani, Marcus Liwicki, and Didier Stricker. "Contrastive Learning for 3D Point Clouds Classification and Shape Completion." Sensors 21, no. 21 (November 6, 2021): 7392. http://dx.doi.org/10.3390/s21217392.
Full textAbstract:
In this paper, we present the idea of Self Supervised learning on the shape completion and classification of point clouds. Most 3D shape completion pipelines utilize AutoEncoders to extract features from point clouds used in downstream tasks such as classification, segmentation, detection, and other related applications. Our idea is to add contrastive learning into AutoEncoders to encourage global feature learning of the point cloud classes. It is performed by optimizing triplet loss. Furthermore, local feature representations learning of point cloud is performed by adding the Chamfer distance function. To evaluate the performance of our approach, we utilize the PointNet classifier. We also extend the number of classes for evaluation from 4 to 10 to show the generalization ability of the learned features. Based on our results, embeddings generated from the contrastive AutoEncoder enhances shape completion and classification performance from 84.2% to 84.9% of point clouds achieving the state-of-the-art results with 10 classes.
48
Schenzinger, Verena, and Axel Kreuter. "Reducing cloud contamination in aerosol optical depth (AOD) measurements." Atmospheric Measurement Techniques 14, no. 4 (April 12, 2021): 2787–98. http://dx.doi.org/10.5194/amt-14-2787-2021.
Full textAbstract:
Abstract. We propose a new cloud screening method for sun photometry that is designed to effectively filter thin clouds. Our method is based on a k-nearest-neighbour algorithm instead of scanning time series of aerosol optical depth. Using 10 years of data from a precision filter radiometer in Innsbruck, we compare our new method and the currently employed screening technique. We exemplify the performance of the two routines in different cloud conditions. While both algorithms agree on the classification of a data point as clear or cloudy in a majority of the cases, the new routine is found to be more effective in flagging thin clouds. We conclude that this simple method can serve as a valid alternative for cloud detection, and we discuss the generalizability to other observation sites.
49
Zeng, Shan, Mark Vaughan, Zhaoyan Liu, Charles Trepte, Jayanta Kar, Ali Omar, David Winker, et al. "Application of high-dimensional fuzzy <i>k</i>-means cluster analysis to CALIOP/CALIPSO version 4.1 cloud–aerosol discrimination." Atmospheric Measurement Techniques 12, no. 4 (April 12, 2019): 2261–85. http://dx.doi.org/10.5194/amt-12-2261-2019.
Full textAbstract:
Abstract. This study applies fuzzy k-means (FKM) cluster analyses to a subset of the parameters reported in the CALIPSO lidar level 2 data products in order to classify the layers detected as either clouds or aerosols. The results obtained are used to assess the reliability of the cloud–aerosol discrimination (CAD) scores reported in the version 4.1 release of the CALIPSO data products. FKM is an unsupervised learning algorithm, whereas the CALIPSO operational CAD algorithm (COCA) takes a highly supervised approach. Despite these substantial computational and architectural differences, our statistical analyses show that the FKM classifications agree with the COCA classifications for more than 94 % of the cases in the troposphere. This high degree of similarity is achieved because the lidar-measured signatures of the majority of the clouds and the aerosols are naturally distinct, and hence objective methods can independently and effectively separate the two classes in most cases. Classification differences most often occur in complex scenes (e.g., evaporating water cloud filaments embedded in dense aerosol) or when observing diffuse features that occur only intermittently (e.g., volcanic ash in the tropical tropopause layer). The two methods examined in this study establish overall classification correctness boundaries due to their differing algorithm uncertainties. In addition to comparing the outputs from the two algorithms, analysis of sampling, data training, performance measurements, fuzzy linear discriminants, defuzzification, error propagation, and key parameters in feature type discrimination with the FKM method are further discussed in order to better understand the utility and limits of the application of clustering algorithms to space lidar measurements. In general, we find that both FKM and COCA classification uncertainties are only minimally affected by noise in the CALIPSO measurements, though both algorithms can be challenged by especially complex scenes containing mixtures of discrete layer types. Our analysis results show that attenuated backscatter and color ratio are the driving factors that separate water clouds from aerosols; backscatter intensity, depolarization, and mid-layer altitude are most useful in discriminating between aerosols and ice clouds; and the joint distribution of backscatter intensity and depolarization ratio is critically important for distinguishing ice clouds from water clouds.
50
Zhang, Zhiwen, Teng Li, Xuebin Tang, Xiangda Lei, and Yuanxi Peng. "Introducing Improved Transformer to Land Cover Classification Using Multispectral LiDAR Point Clouds." Remote Sensing 14, no. 15 (August 7, 2022): 3808. http://dx.doi.org/10.3390/rs14153808.
Full textAbstract:
The use of Transformer-based networks has been proposed for the processing of general point clouds. However, there has been little research related to multispectral LiDAR point clouds that contain both spatial coordinate information and multi-wavelength intensity information. In this paper, we propose networks for multispectral LiDAR point cloud point-by-point classification based on an improved Transformer. Specifically, considering the sparseness of different regions of multispectral LiDAR point clouds, we add a bias to the Transformer to improve its ability to capture local information and construct an easy-to-implement multispectral LiDAR point cloud Transformer (MPT) classification network. The MPT network achieves 78.49% mIoU, 94.55% OA, 84.46% F1, and 0.92 Kappa on the multispectral LiDAR point cloud testing dataset. To further extract the topological relationships between points, we present a standardization set abstraction (SSA) module, which includes the global point information while considering the relationships among the local points. Based on the SSA module, we propose an advanced version called MPT+ for the point-by-point classification of multispectral LiDAR point clouds. The MPT+ network achieves 82.94% mIoU, 95.62% OA, 88.42% F1, and 0.94 Kappa on the same testing dataset. Compared with seven point-based deep learning algorithms, our proposed MPT+ achieves state-of-the-art results for several evaluation metrics.