Добірка наукової літератури з теми "Draco compression"

Point cloud videos play a crucial role in immersive applications enabled by holographic-type communication, which has been identified as an important service for 6G and beyond wireless systems and the metaverse. The significant volume of point cloud video demands efficient compression and transmission techniques to support the Quality of Experience (QoE) requirements of real-time immersive applications. A few Point Cloud Compression (PCC) techniques, such as MPEG PCC and Draco, have emerged in recent years, and studies have shown that each technique has its strengths and weaknesses under different system settings. This paper proposes a multi-codec rate adaptive point cloud streaming method to satisfy the QoE requirements of interactive and live applications considering available system resources. The proposed method leverages three common PCC techniques: MPEG V-PCC, MPEG G-PCC, and Draco. The performance of each PCC technique is evaluated under various test conditions, and then estimation models are constructed to predict the bit rate, the decoding time, and the quality of the reconstructed point cloud. Based on the user's quality requirements and available computational and communication resources, the proposed streaming method selects a codec along with appropriate compression parameters that can provide the minimum latency for streaming. Evaluation results demonstrate that the proposed method can provide better QoE than benchmark methods under various bandwidth and computation scenarios.

In this article, we present an efficient coding scheme for LiDAR point cloud maps. As a point cloud map consists of numerous single scans spliced together, by recording the time stamp and quaternion matrix of each scan during map building, we cast the point cloud map compression into the point cloud sequence compression problem. The coding architecture includes two techniques: intra-coding and inter-coding. For intra-frames, a segmentation-based intra-prediction technique is developed. For inter-frames, an interpolation-based inter-frame coding network is explored to remove temporal redundancy by generating virtual point clouds based on the decoded frames. We only need to code the difference between the original LiDAR data and the intra/inter-predicted point cloud data. The point cloud map can be reconstructed according to the decoded point cloud sequence and quaternion matrices. Experiments on the KITTI dataset show that the proposed coding scheme can largely eliminate the temporal and spatial redundancies. The point cloud map can be encoded to 1/24 of its original size with 2 mm-level precision. Our algorithm also obtains better coding performance compared with the octree and Google Draco algorithms.

Three-dimensional visualization of medical image data can enable doctors to observe images from more angles and higher dimensions. It is of great significance for doctors to assist in diagnosis and preoperative planning. Most 3D visualization systems are based on desktop applications, which are too dependent on hardware and operating system. This makes it difficult to use across platforms and maintain. Web-based systems tend to have limited capabilities. To this end, we developed a web application, which not only provides DICOM (Digital Imaging and Communications in Medicine) image browsing and annotation functions, but also provides three-dimensional post-processing functions of multiplanar reconstruction, volume rendering, lung parenchyma segmentation and brain MRI (Magnetic Resonance Imaging) analysis. In order to improve the rendering speed, we use the Marching Cube algorithm for 3D reconstruction in the background in an asynchronous way, and save the reconstructed model as glTF (GL Transmission Format). At the same time, Draco compression algorithm is used to optimize the glTF model to achieve more efficient rendering. After performance evaluation, the system reconstructed a CT (Computed Tomography) series of 242 slices and the optimized model was only 6.37mb with a rendering time of less than 2.5s. Three-dimensional visualization of the lung parenchyma clearly shows the volume, location, and shape of pulmonary nodules. The segmentation and reconstruction of different brain tissues can reveal the spatial three-dimensional structure and adjacent relationship of glioma in the brain, which has great application value in auxiliary diagnosis and preoperative planning.

Published mass models fitted to galaxy rotation curves are used to study the systematic properties of dark matter (DM) halos in late-type and dwarf spheroidal (dSph) galaxies. Halo parameters are derived by fitting non-singular isothermals to (V2 – V2vis)1/2, where V(r) is the observed rotation curve and Vvis is the rotation curve of the visible matter. the latter is calculated from the surface brightness assuming that the mass-to-light ratio M/L is constant with radius. “Maximum disk” values of M/L are adjusted to fit as much of the inner rotation curve as possible without making the halo have a hollow core. Rotation curve decomposition becomes impossible fainter than absolute magnitude Mb ≃ −14, where V becomes comparable to the velocity dispersion of the gas. To increase the luminosity range further, we include dSph galaxies, which are physically related to spiral and irregular galaxies. Combining the data, we find that DM halos satisfy well defined scaling laws analogous to the “fundamental plane” relations for elliptical galaxies. Halos in less luminous galaxies have smaller core radii rc, higher central densities ρ0, and smaller central velocity dispersions σ. Scaling laws provide new and detailed constraints on the nature of DM and on galaxy formation and evolution. Some simple implications include:1 – A single, continuous physical sequence of increasing mass extends from dSph galaxies with Mb ≃ −7.6 to Sc I galaxies with Mb ≃ −22.4.2 – the high DM densities in dSph galaxies are normal for such tiny galaxies. Since virialised density depends on collapse redshift zcoll, ρ0 ∝ (1 + zcoll)3, the smallest dwarfs formed at least Δzcoll ≃ 7 earlier than the biggest spirals.3 – the high DM densities of dSphs implies that they are real galaxies formed from primordial density fluctuations. They are not tidal fragments. Tidal dwarfs cannot retain even the low DM densities of their giant-galaxy progenitors. in contrast, dSphs have higher DM densities than do giant-galaxy progenitors.4 – the fact that, as luminosity decreases, dwarf galaxies become much more numerous and also more nearly dominated by DM raises the possibility that there exists a large population of objects that are completely dark. Such objects are a canonical prediction of cold DM theory. If they exist, “empty halos” are likely to be small and dense -that is, darker versions of Draco and UMi.5 – the slopes of the DM parameter correlations provide a measure on galactic mass scales of the slope n of the power spectrum |δk|2 ∝ kn of primordial density fluctuations. Our preliminary results, not yet corrected for baryonic compression of DM, give n ≃ –1.9 ± 0.2. This is consistent with cold DM theory.

Parent concrete coming from a wide range of sources can result in considerable differences in the properties of recycled coarse aggregate (RCA). In this study, the RCAs were obtained by crushing the parent concrete with water-to-cement ratios (W/Cparent) of 0.4, 0.5 and 0.6, respectively, and were strengthened by carbonation and nano-silica slurry wrapping methods. It was found that when W/Cparen was 0.3, 0.4 and 0.5, respectively, compared with the mortar in the untreated RCA, the capillary porosity of the mortar in the carbonated RCA decreased by 19%, 16% and 30%, respectively; the compressive strength of concrete containing the carbonated RCA increased by 13%, 11% and 13%, respectively; the chloride diffusion coefficient of RAC (DRAC) containing the nano-SiO2 slurry-treated RCA decreased by 17%, 16% and 11%; and that of RAC containing the carbonated RCA decreased by 21%, 25% and 26%, respectively. Regardless of being strengthened or not, both DRAC and porosity of old mortar in RCAs increased with increasing W/Cparent. For different types of RCAs, DRAC increased obviously with increasing water absorption of RCA. Finally, a theoretical model of DRAC considering the water absorption of RCA was established and verified by experiments, which can be used to predict the DRAC under the influence of different factors, especially the water absorption of RCA.

Abstract Vast masses of basic rocks are present as tectonic blocks and slices along the Eastern deep fault of the Vardar zone of Serbia. They are predominantly comprised of gabbro, with smaller part made up of diabase, and occurrences of granite, aplite and pegmatite dykes. Basic rock masses are trending along the line Kragujevac (Ždraljica)-Velika Pčelica-Bogalinac ~8 km west of Rekovac. A significantly smaller diabase massif is present along the same tectonic line, further toward SE, at Prevešt village by Kalenicka River, approximately 13 km south of Rekovac. Drača open pit mine is situated in this diabase massif. Geologic explorative works have confirmed the reserves of 1 846 695 t of stone mass for building purposes. For over a decade, Draca mine has been producing various types of building stone, mainly graded stone aggregate with favourable physico-mechanical properties for road-construction works. Chemical analyses and petrographic study have shown typical composition and fabric for this type of rock. Main constituents are plagioclase and pyroxene, with opaque minerals as accessory and varying secondary minerals – chlorite, calcite, in some places epidote and limonite. Pyrite enrichment is visible in some areas of the massif. Chlorite, calcite, epidote and pyrite are the products of propylitic alteration. Although products of alteration are present throughout the rock mass with variable intensity, as is typical for the basic rocks of the former ocean floor sequences, petrologic properties are favourable for building stone purposes. Physico-mechanical properties of diabase have favourable values and varying scattering degrees. Dry state uniaxial compressive strength average values from seven analyses vary in the range 130-169 MPa. Resistance to abrasion average values vary in the range 9.04-17.07 cm3/50cm2. Apparent density varies within the span 2759-2926 g/cm3 and real density 2804-2951 g/cm3. Water absorption values 0.08-1.04 %. Resistance to weathering through testing of stability using Sodium-sulphate values vary from 0.00 to 0.15 % and through frost resistance from 0.00 to 0.04 %. Porosity values are almost constant at 0.8 %. In more altered parts of the rock mass, porosity reaches 1.6 %. Graded crushed aggregate has favourable values of Los Angeles coefficient 14.2 and 14.3 % for gradation B. Taking into consideration all performed tests and analyses, it is concluded that diabase from Drača mine can be used as a building stone for production of aggregate for use in concrete and for road-construction (asphalt paving mixtures for moderate, light and very light traffic load as a top wearing layer; for lower and upper bearing layers; for classic and modern road foundations); for production of crushed and hewn stone for building; crushed stone for railroad ballast. Also, it can be and is used as a raw material for production of stone wool for thermal insulation purposes.

AbstractThis paper develops an arithmetic coding algorithm based on delta recurrent neural network for edge computing devices called DRAC. Our algorithm is implemented on a Xilinx Zynq 7000 Soc board. We evaluate DRAC with four datasets and compare it with the state-of-the-art compressor DeepZip. The experimental results show that DRAC outperforms DeepZip and achieves 5X speedup ratio and 20X power consumption saving.

Introduction: One of the important advantages of the distal transradial access (dTRA) is the significant reduction in the incidence of radial artery occlusion (RAO). There are few reports on the influencing factors for distal radial artery occlusion (dRAO) after cardiovascular interventions via the dTRA. Methods: This retrospective analysis included the clinical data of patients who underwent a cardiovascular intervention via the dTRA. The dRAO was evaluated by ultrasound within 24 hours after the procedure. Multivariate logistic analysis was used to explore the influencing factors for dRAO. Results: The incidence of dRAO was 3.5% (28/805) at 24 hours follow-up after the procedure. In the comparison between the 2 groups, the preoperative distal radial artery (DRA) internal diameter in the dRAO group was significantly smaller than that in the non-dRAO group (p=0.001). The prevalence of DRA inner diameter/sheath outer diameter <1 was significantly higher in the dRAO group than in the non-dRAO group (p=0.013). The number of puncture attempts was significantly greater in the dRAO group than in the non-dRAO group (p=0.007). Multivariate logistic analysis showed that DRA inner diameter/sheath outer diameter <1 was an independent risk factor for dRAO (OR=4.827, 95% CI=1.087–21.441, p=0.039). Conclusions: The incidence of dRAO 24 hours after cardiovascular intervention via the dTRA was 3.5%, and a DRA inner diameter/sheath outer diameter <1 was an independent risk factor for dRAO. Preoperative ultrasound assessment of vessel inner diameter and selection of a sheath with a smaller outer diameter may reduce the risk of dRAO. Clinical Impact The incidence of distal radial artery occlusion after cardiovascular intervention was 3.5%. The distal radial artery inner diameter/sheath outer diameter <1 was an independent risk factor for distal radial artery occlusion. Preoperative ultrasound assessment of vessel inner diameter and selection of a sheath with a smaller outer diameter may reduce the risk of distal radial artery occlusion. The number of puncture attempts and compression time were not related to distal radial artery occlusion.

Au cours des dernières décennies, les objets 3D sont devenus un élément essentiel de la vie quotidienne, tant dans le contexte privé que professionnel. Ces objets 3D sont souvent stockés sur le cloud et transférés sur des réseaux plusieurs fois au cours de leur existence, où ils sont susceptibles de faire l'objet d'attaques malveillantes. Par conséquent, des méthodes de sécurisation des objets 3D, comme le chiffrement ou l'insertion des données cachées, sont essentielles. Le chiffrement est utilisé pour protéger la confidentialité visuelle du contenu d’un objet 3D. Il est également possible d'utiliser des schémas de chiffrement sélectif, dans lesquels seulement une partie de l’objet 3D est chiffrée. L'insertion des données cachées est généralement utilisée pour protéger les droits d'auteur ou l'authenticité des objets 3D. Toutefois, lorsqu'un objet 3D est chiffré, un tiers, tel qu'un serveur, peut avoir besoin d'intégrer des données dans l'objet 3D confidentiel. Dans ce cas, les données sont cachées dans le domaine chiffré. Les objets 3D sont souvent constitués de millions de sommets, de sorte que le stockage et le partage en ligne sont coûteux. Par conséquent, la compression des objets 3D est essentielle. Dans ce travail, nous présentons trois contributions dans différents domaines de recherche.Premièrement, nous présentons notre travail sur une nouvelle méthode permettant d'obtenir un objet 3D marqué à partir d'une insertion de données cachées de haute capacité dans le domaine chiffré. Basée sur des propriétés homomorphiques du cryptosystème de Paillier, notre méthode permet d'insérer plusieurs messages secrets dans le domaine chiffré avec une haute capacité. Ces messages peuvent être extraits dans le domaine en clair après le déchiffrement de l'objet 3D. À notre connaissance, nous sommes les premiers à proposer une méthode d'insertion de données cachées dans le domaine chiffré où les données cachées de haute capacité sont conservées dans le domaine en clair après le déchiffrement de l'objet 3D. Le chiffrement et l'insertion de données cachées dans le domaine chiffré sont conformes au format et sans expansion de taille, malgré l'utilisation du cryptosystème de Paillier.Nous présentons ensuite notre travail sur une mesure d'évaluation du niveau de sécurité visuelle des objets 3D chiffrés sélectivement. Basé sur une nouvelle base de données composée d'objets 3D chiffrés sélectivement et évalués, nous proposons un modèle pour déterminer les paramètres de sécurité en fonction du niveau de sécurité souhaité. Enfin, nous détaillons notre score 3DVS qui sert à mesurer le niveau de sécurité visuelle des objets 3D chiffrés sélectivement.Nous présentons également, à notre connaissance, la première méthode permettant de déchiffrer hiérarchiquement un objet 3D chiffré en fonction d'un trousseau de clés généré. Ce trousseau se compose d'un ensemble de clés qui permettent un déchiffrement plus ou moins fort de l'objet 3D chiffré. Chaque objet 3D déchiffré hiérarchiquement a un niveau de sécurité visuelle différent, où l'objet 3D est plus ou moins accessible visuellement. Notre méthode est essentielle lorsqu'il s'agit d'empêcher des fuites des secrets commerciales au sein d'une entreprise ou par des attaquants extérieurs. Elle est également écologique et plus sécurisée que les méthodes traditionnelles de chiffrement sélectif.Enfin, nous présentons notre travail sur des méthodes conjointes de sécurité et de compression basées sur la méthode de compression d'objets 3D de Google, Draco, dans laquelle nous intégrons une étape de sécurité dans Draco, qui est en train de devenir la nouvelle norme de l'industrie. Ces étapes de sécurité sont le chiffrement, le chiffrement sélectif et le tatouage
Over the last few decades, 3D objects have become an essential part of everyday life, in both private and professional contexts. These 3D objects are often stored on the cloud and transferred over networks many times during their existence, where they are susceptible to malicious attacks. Therefore, 3D object security, such as encryption or data hiding, is essential. Encryption is used to protect the visual confidentiality of the 3D object's content. Selective encryption schemes can also be used, where part of a component, such as a part of each vertex, is encrypted. Data hiding is generally used to protect the copyright or the authenticity of the 3D object. However, when a 3D object is encrypted, a third party such as a server may need to embed data in the confidential 3D object. In this case, data hiding in the encrypted domain is performed. In many applications, 3D objects often consist of millions of vertices, and so storing and sharing them online is expensive, time consuming and not environmentally friendly. Consequently, 3D object compression is essential. In this work, we present three contributions in different research areas. First, we present our work on a new method to obtain a watermarked 3D object from high-capacity data hiding in the encrypted domain. Based on the homomorphic properties of the Paillier cryptosystem, our proposed method allows us to embed several secret messages in the encrypted domain with a high-capacity. These messages can be extracted in the plaintext domain after the 3D object decryption. To the best of our knowledge, we are the first to propose a data hiding method in the encrypted domain where the high-capacity watermark is conserved in the plaintext domain after the 3D object is decrypted. The encryption and the data hiding in the encrypted domain are format compliant and without size expansion, despite the use of the Paillier cryptosystem. Then, we present our work on an evaluation metric for the visual security level of selectively encrypted 3D objects. We present a new dataset composed of evaluated selectively encrypted 3D objects. We propose a model to determine the security parameters according to a desired security level. Finally, we detail our proposed 3DVS score which serves to measure the visual security level of selectively encrypted 3D objects. We also present a method which allows us to hierarchically decrypt an encrypted 3D object according to a generated ring of keys. This ring consists of a set of keys that allow a stronger or weaker decryption of the encrypted 3D object. Each hierarchically decrypted 3D object has a different visual security level, where the 3D object is more or less visually accessible. Our method is essential when it comes to preventing trade secrets from being leaked from within a company or by exterior attackers. It is also ecologically friendly and more secure than traditional selective encryption methods. Finally, we present our work on joint security and compression methods based on Google's 3D object compression method Draco, where we integrate a security step in Draco, which is becoming the new industry standard. These security steps are encryption, selective encryption and watermarking

The target of this thesis is sumarize the theory of metods used for composition image sequence into single high dynamic range image. Selected methods and algorithms are described in theoretical part and implemented in practical part of this thesis. It was used C++ language with using OpenCV library.