Dissertations / Theses on the topic 'Detection'
To see the other types of publications on this topic, follow the link: Detection.
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Detection.'
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 dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
1
Kapoor, Prince. "Shoulder Keypoint-Detection from Object Detection." Thesis, Université d'Ottawa / University of Ottawa, 2018. http://hdl.handle.net/10393/38015.
Full textAbstract:
This thesis presents detailed observation of different Convolutional Neural Network
(CNN) architecture which had assisted Computer Vision researchers to achieve state-of-the-art performance on classification, detection, segmentation and much more to
name image analysis challenges. Due to the advent of deep learning, CNN had
been used in almost all the computer vision applications and that is why there is
utter need to understand the miniature details of these feature extractors and find
out their pros and cons of each feature extractor meticulously. In order to perform
our experimentation, we decided to explore an object detection task using a particular
model architecture which maintains a sweet spot between computational cost and
accuracy. The model architecture which we had used is LSTM-Decoder. The
model had been experimented with different CNN feature extractor and found their
pros and cons in variant scenarios. The results which we had obtained on different
datasets elucidates that CNN plays a major role in obtaining higher accuracy and
we had also achieved a comparable state-of-the-art accuracy on Pedestrian Detection
Dataset.
In extension to object detection, we also implemented two different model architectures which find shoulder keypoints. So, One of our idea can be explicated as
follows: using the detected annotation from object detection, a small cropped image
is generated which would be feed into a small cascade network which was trained
for detection of shoulder keypoints. The second strategy is to use the same object detection model and fine tune their weights to predict shoulder keypoints. Currently,
we had generated our results for shoulder keypoint detection. However, this idea
could be extended to full-body pose Estimation by modifying the cascaded network
for pose estimation purpose and this had become an important topic of discussion
for the future work of this thesis.
2
Laxhammar, Rikard. "Conformal anomaly detection : Detecting abnormal trajectories in surveillance applications." Doctoral thesis, Högskolan i Skövde, Institutionen för informationsteknologi, 2014. http://urn.kb.se/resolve?urn=urn:nbn:se:his:diva-8762.
Full textAbstract:
Human operators of modern surveillance systems are confronted with an increasing amount of trajectory data from moving objects, such as people, vehicles, vessels, and aircraft. A large majority of these trajectories reflect routine traffic and are uninteresting. Nevertheless, some objects are engaged in dangerous, illegal or otherwise interesting activities, which may manifest themselves as unusual and abnormal trajectories. These anomalous trajectories can be difficult to detect by human operators due to cognitive limitations. In this thesis, we study algorithms for the automated detection of anomalous trajectories in surveillance applications. The main results and contributions of the thesis are two-fold. Firstly, we propose and discuss a novel approach for anomaly detection, called conformal anomaly detection, which is based on conformal prediction (Vovk et al.). In particular, we propose two general algorithms for anomaly detection: the conformal anomaly detector (CAD) and the computationally more efficient inductive conformal anomaly detector (ICAD). A key property of conformal anomaly detection, in contrast to previous methods, is that it provides a well-founded approach for the tuning of the anomaly threshold that can be directly related to the expected or desired alarm rate. Secondly, we propose and analyse two parameter-light algorithms for unsupervised online learning and sequential detection of anomalous trajectories based on CAD and ICAD: the sequential Hausdorff nearest neighbours conformal anomaly detector (SHNN-CAD) and the sequential sub-trajectory local outlier inductive conformal anomaly detector (SSTLO-ICAD), which is more sensitive to local anomalous sub-trajectories. We implement the proposed algorithms and investigate their classification performance on a number of real and synthetic datasets from the video and maritime surveillance domains. The results show that SHNN-CAD achieves competitive classification performance with minimum parameter tuning on video trajectories. Moreover, we demonstrate that SSTLO-ICAD is able to accurately discriminate realistic anomalous vessel trajectories from normal background traffic.
3
Albrektsson, Fredrik. "Detecting Sockpuppets in Social Media with Plagiarism Detection Algorithms." Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-208553.
Full textAbstract:
As new forms of propaganda and information control spread across the internet, the need for novel ways of identifying them increases as well. One increasingly popular method of spreading false messages on microblogs like Twitter is to disseminate them from seemingly ordinary, but centrally controlled and coordinated user accounts – sockpuppets. In this paper we examine a number of potential methods for identifying these by way of applying plagiarism detection algorithms for text, and evaluate their performance against this type of threat. We identify one type of algorithm in particular – that using vector space modeling of text – as particularly useful in this regard.Allteftersom nya former av propaganda och informationskontroll sprider sig över internet krävs också nya sätt att identifiera dessa. En allt mer populär metod för att sprida falsk information på mikrobloggar som Twitter är att göra det från till synes ordinära, men centralt kontrollerade och koordinerade användarkonton – på engelska kända som “sockpuppets”. I denna undersökning testar vi ett antal potentiella metoder för att identifiera dessa genom att applicera plagiatkontrollalgoritmer ämnade för text, och utvärderar deras prestanda mot denna sortens hot. Vi identifierar framför allt en typ av algoritm – den som nyttjar vektorrymdsmodellering av text – som speciellt användbar i detta avseende.
4
Le, Anhtuan. "Intrusion Detection System for detecting internal threats in 6LoWPAN." Thesis, Middlesex University, 2017. http://eprints.mdx.ac.uk/21958/.
Full textAbstract:
6LoWPAN (IPv6 over Low-power Wireless Personal Area Network) is a standard developed by the Internet Engineering Task Force group to enable the Wireless Sensor Networks to connect to the IPv6 Internet. This standard is rapidly gaining popularity for its applicability, ranging extensively from health care to environmental monitoring. Security is one of the most crucial issues that need to be considered properly in 6LoWPAN. Common 6LoWPAN security threats can come from external or internal attackers. Cryptographic techniques are helpful in protecting the external attackers from illegally joining the network. However, because the network devices are commonly not tampered-proof, the attackers can break the cryptography codes of such devices and use them to operate like an internal source. These malicious sources can create internal attacks, which may downgrade significantly network performance. Protecting the network from these internal threats has therefore become one of the centre security problems on 6LoWPAN. This thesis investigates the security issues created by the internal threats in 6LoWPAN and proposes the use of Intrusion Detection System (IDS) to deal with such threats. Our main works are to categorise the 6LoWPAN threats into two major types, and to develop two different IDSs to detect each of this type effectively. The major contributions of this thesis are summarised as below. First, we categorise the 6LoWPAN internal threats into two main types, one that focuses on compromising directly the network performance (performance-type) and the other is to manipulate the optimal topology (topology-type), to later downgrade the network service quality indirectly. In each type, we select some typical threats to implement, and assess their particular impacts on network performance as well as identify performance metrics that are sensitive in the attacked situations, in order to form the basis detection knowledge. In addition, on studying the topology-type, we propose several novel attacks towards the Routing Protocol for Low Power and Lossy network (RPL - the underlying routing protocol in 6LoWPAN), including the Rank attack, Local Repair attack and DIS attack. Second, we develop a Bayesian-based IDS to detect the performance-type internal threats by monitoring typical attacking targets such as traffic, channel or neighbour nodes. Unlike other statistical approaches, which have a limited view by just using a single metric to monitor a specific attack, our Bayesian-based IDS can judge an abnormal behaviour with a wiser view by considering of different metrics using the insightful understanding of their relations. Such wiser view helps to increase the IDS’s accuracy significantly. Third, we develop a Specification-based IDS module to detect the topology-type internal threats based on profiling the RPL operation. In detail, we generalise the observed states and transitions of RPL control messages to construct a high-level abstract of node operations through analysing the trace files of the simulations. Our profiling technique can form all of the protocol’s legal states and transitions automatically with corresponding statistic data, which is faster and easier to verify compare with other manual specification techniques. This IDS module can detect the topology-type threats quickly with a low rate of false detection. We also propose a monitoring architecture that uses techniques from modern technologies such as LTE (Long-term Evolution), cloud computing, and multiple interface sensor devices, to expand significantly the capability of the IDS in 6LoWPAN. This architecture can enable the running of both two proposed IDSs without much overhead created, to help the system to deal with most of the typical 6LoWPAN internal threats. Overall, the simulation results in Contiki Cooja prove that our two IDS modules are effective in detecting the 6LoWPAN internal threats, with the detection accuracy is ranging between 86 to 100% depends on the types of attacks, while the False Positive is also satisfactory, with under 5% for most of the attacks. We also show that the additional energy consumptions and the overhead of the solutions are at an acceptable level to be used in the 6LoWPAN environment.
5
Olsson, Jonathan. "Detecting Faulty Piles of Wood using Anomaly Detection Techniques." Thesis, Luleå tekniska universitet, Institutionen för system- och rymdteknik, 2021. http://urn.kb.se/resolve?urn=urn:nbn:se:ltu:diva-83061.
Full textAbstract:
The forestry and the sawmill industry have a lot of incoming and outgoing piles of wood. It's important to maintain quality and efficiency. This motivates an examination of whether machine learning- or more specifically, anomaly detection techniques can be implemented and used to detect faulty shipments. This thesis presents and evaluates some computer vision techniques and some deep learning techniques. Deep learning can be divided into groups; supervised, semi-supervised and unsupervised. In this thesis, all three groups were examined and it covers supervised methods such as Convolutional Neural Networks, semi-supervised methods such as a modified Convolutional Autoencoder (CAE) and lastly, an unsupervised technique such as Generative Adversarial Network (GAN) was being tested and evaluated. A version of a GAN model proved to perform best for this thesis in terms of the accuracy of faulty detecting shipments with an accuracy rate of 68.2% and 79.8\% overall, which was satisfactory given the problems that were discovered during the progress of the thesis.
6
Prevot, Yohan. "Arterial perfusion detection method by synchronous detection." [Tampa, Fla] : University of South Florida, 2005. http://purl.fcla.edu/usf/dc/et/SFE0001385.
Full text
7
Chau, Sam. "Investigation of silicon PIN-detector for laser pulse detection." Thesis, Linköping University, Department of Science and Technology, 2004. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-325.
Full textAbstract:
This report has been written at SAAB Bofors Dynamics (SBD) AB in Gothenburg at the department of optronic systems.
In military observation operations, a target to hit is chosen by illumination of a laser designator. From the targetpoint laser radiation is reflected on a detector that helps identify the target. The detector is a semiconductor PIN-type that has been investigated in a laboratory environment together with a specially designed laser source. The detector is a photodiode and using purchased components, circuits for both the photodiode and the laserdiode has been designed and fabricated. The bandwidth of the op-amp should be about 30 MHz, in the experiments a bandwidth of 42 MHz was used. Initially the feedback network, which consists of a 5.6 pF capacitor in parallel with a 1-kohm resistor determined the bandwidth. To avoid the op-amp saturate under strong illuminated laser radiation the feedback network will use a 56-pF capacitor and a 100-ohm resistor respectively.
The laser should be pulsed with 10-20 ns width, 10 Hz repetition frequency, about 800 nm wavelength and a maximum output power of 80 mW. To avoid electrical reflection signals at measurement equipment connections, the laser circuit includes a resistor of about 50 ohm, that together with the resistance in the laserdiode forms the right termination that eliminate the reflection signals. The wire-wound type of resistor shall be avoided in this application and instead a surface mounted type was beneficial with much lower inductance. The detector showed a linear behaviour up to 40-mW optical power. Further investigation was hindered by the breakdown of the laserdiodes. The function generator limits the tests to achieve 80 mW in light power. In different experiments the responsivity of the photodiode is different from the nominal value, however it would have required more time to investigate the causes.
8
Chang, Pi-Jung. "Double Chooz neutrino detector: neutron detection systematic errors and detector seasonal stability." Diss., Kansas State University, 2013. http://hdl.handle.net/2097/16861.
Full textAbstract:
Doctor of PhilosophyDepartment of Physics
Glenn Horton-Smith
In March 2012, the Double Chooz reactor neutrino experiment published its most precise result so far: sin[superscript]2 2theta13 = 0.109 +/- 0.030(stat.) +/- 0.025(syst.). The statistical significance is 99.8% away from the no-oscillation hypothesis. The systematic uncertainties from background and detection efficiency are smaller than the first publication of the Double Chooz experiment. The neutron detection efficiency, one of the biggest contributions in detection systematic uncertainties, is a primary topic of this dissertation. The neutron detection efficiency is the product of three factors: the Gd-capture fraction, the efficiency of time difference between prompt and delayed signals, and the efficiency of energy containment. [superscript]252 Cf is used to determine the three factors in this study. The neutron detection efficiency from the [superscript]252 Cf result is confirmed by the electron antineutrino data and Monte Carlo simulations. The systematic uncertainty from the neutron detection efficiency is 0.91% used in the sin[superscript]2 2theta13 analysis. The seasonal variation in detector performance and the seasonal variations of the muon intensity are described in detail as well. The detector stability is confirmed by observation of two phenomena: 1) the [electron antineutrino] rate, which is seen to be uncorrelated with the liquid scintillator temperature, and 2) the daily muon rate, which has the expected correspondence with the effective atmospheric temperature. The correlation between the muon rate and effective atmospheric temperature is further analyzed in this thesis to determine the ratio of kaon to pion in the local atmosphere. An upper limit on instability of the neutron detection efficiency is established in the final chapter. The systematic error, 0.13%, from the relative instability is the deviation of the calibration runs. This thesis concludes with the potential systematic errors of neutron detection efficiency and estimation of how these potential systematic errors affect the result of sin[superscript]2 2theta13.
9
Wang, Jinghui. "Evaluation of GaN as a Radiation Detection Material." The Ohio State University, 2012. http://rave.ohiolink.edu/etdc/view?acc_num=osu1343316898.
Full text
10
Frascarelli, Antonio Ezio. "Object Detection." Thesis, Mälardalens högskola, Inbyggda system, 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:mdh:diva-28259.
Full textAbstract:
During the last two decades the interest about computer vision raised steadily with multiple applications in fields like medical care, automotive, entertainment, retail, industrial, and security. Objectdetection is part of the recognition problem, which is the most important scope of the computervision environment.The target of this thesis work is to analyse and propose a solution for object detection in a real timedynamic environment. RoboCup@Home will be the benchmarking event for this system, which willbe equipped on a robot competing in the 2018 event. The system has to be robust and fast enoughto allow the robot to react to each environment change in a reasonable amount of time.The input hardware used to achieve such system comprise of a Microsoft Kinect, which providesan high definition camera and fast and reliable 3D scanner. Through the study and analysis ofstate-of-the-art algorithms regarding machine vision and object recognition, the more suitable oneshave been tested to optimise the execution on the targeted hardware. Porting of the application toan embedded platform is discussed.
11
Macey, Paul Michael. "Apnoea detection." Thesis, University of Canterbury. Electrical and Electronic Engineering, 1998. http://hdl.handle.net/10092/6888.
Full textAbstract:
This thesis is concerned with the detection of apnoeas in infants from an abdominal breathing signal, where an apnoea is a pause in breathing during sleep. Apnoea detection is performed by analysing breathing signals recorded during sleep studies. An abdominal breathing signal recorded by the BabyLog polysomnographic system is used for this research. A reference set of apnoeas is formed by three human experts identifying apnoeas five seconds and longer within ten overnight recordings of breathing. There was a 10% disagreement on the identification of events. Based on this reference set, the performances of existing methods of apnoea detection were evaluated, and found to have low incidence of false negatives but high incidence of false positives. An existing algorithm was developed, and an application of this algorithm as part of a study of low risk infants is presented. Properties that represent most apnoeas as found in an abdominal breathing signal are described. Human experts are consulted to determine what properties of the signal they use to recognise apnoeas, and a collection of deterministic, or shape, properties is condensed to represent expert opinion. An apnoea is modeled as a flat region with four properties: flatness, duration, thinness and smoothness. Mathematical descriptions that discriminate between apnoea and non-apnoea events of each property are formulated. An expert system for the classification of events is then developed, based on property measures being classified by a neural network. The system has achieved 95% to 98% accuracy for a false detection rate of 15% to 40%. Applications include scoring apnoeas for sleep studies, an aid to clinicians in diagnosing breathing problems, and developing standard definitions of breathing signals corresponding to apnoeas.
12
Vella, Mark Joseph. "Distress detection." Thesis, University of Strathclyde, 2012. http://digitool.lib.strath.ac.uk:80/R/?func=dbin-jump-full&object_id=27534.
Full textAbstract:
Web attacks pose a prime concern for cybersecurity, and whilst attackers are leveraging modern technologies to launch unpredictable attacks with serious consequences, web attack detectors are still restricted to the classical misuse and anomaly detection methods. As a result, web attack detectors have limited resilience to novel attacks or produce impractical amounts of daily false alerts. Advances in intrusion detection techniques have so far only partly alleviated the problem as they are still tied to existing methods. This thesis proposes Distress Detection (DD), a detection method providing novel web attack resilience while suppressing false alerts. It is partly inspired by the workings of the human immune system, that is capable to respond against previously unseen infections. The premise is that within the scope of an attack objective (the attack's end result), attack HTTP requests are associated with features that are necessary to reach that objective, rendering them suspicious. Their eventual execution must generate system events that are associated with the successful attainment of their objective, called the attack symptoms. Suspicious requests and attack symptoms are modeled on the generic signs of ongoing infections that enable the immune system to respond to novel infections, however they are not exclusive to attacks. The suppression of false alerts is left to an alert correlation process based on the premise that attack requests can be distinguished from the rest through a link that connects their features with their consequent attack symptoms. The provision of novel attack resilience and false alert suppression is demonstrated through three prototype distress detectors, identifying DD as promising for effective web attack detection, despite some concerns about the level of diffculty of their implementation process.
13
Mustamo, P. (Pirkko). "Object detection in sports:TensorFlow Object Detection API case study." Bachelor's thesis, University of Oulu, 2018. http://urn.fi/URN:NBN:fi:oulu-201802081173.
Full textAbstract:
Object detection is widely used in the world of sports, its users including training staff, broadcasters and sports fans. Neural network based classifiers are used together with other object detection techniques. The aim of this study was to explore the modern open source based solutions for object detection in sports, in this case for detecting football players. TensorFlow Object Detection API, an open source framework for object detection related tasks, was used for training and testing an SSD (Single-Shot Multibox Detector) with Mobilenet- model. The model was tested as a) pre-trained and b) with fine-tuning with a dataset consisting of images extracted from video footage of two football matches. Following hypotheses were examined: 1) Pre-trained model will not work on the data without fine-tuning. 2) Fine-tuned model will work reasonably well on the given data. 3) Fine-tuned model will have problems with occlusion and players pictured against the rear wall. 4) Using more variable training data will improve results on new images. The results of this study indicate that: 1) The pre-trained model was useless for detecting players in the test images. 2) A fine-tuned model worked reasonably well. 3) Problem areas were players in clusters and/or pictured against the rear wall. 4) A model trained with data from one game was able to detect players in footage from another game. The overall model performance did not much improve by training the model with data from two games. Other model types (such as Faster R-CNN model) should be tested on the dataKohteentunnistusta käytetään yleisesti urheilumaailmassa, mm. valmennuksessa, televisiolähetyksissä sekä fanikäytössä. Neuroverkkoihin perustuvia menetelmiä käytetään yhdessä muiden tekniikoiden kanssa. Tämän tutkimuksen päämäärä oli tarkastella moderneja avoimen lähdekoodin ratkaisuja kohteentunnistukseen urheilussa, tässä tapauksessa jalkapalloilijoiden tunnistuksessa. TensorFlow Object Detection API perustuu avoimeen lähdekoodiin ja tarjoaa työkaluja kohteentunnistukseen. Sen avulla opetettiin ja testattiin SSD (Single-Shot Multibox Detector) with Mobilenet- mallia sekä a) valmiiksi treenattuna että b) hienosäädettynä aineistolla, joka koostui kahdesta jalkapallo-otteluvideosta poimituista kuvista. Työssä tarkasteltiin seuraavia hypoteeseja: 1) Valmiiksi opettettu malli ei toimi ilman hienosäätöä omalle aineistolle. 2) Hienosäädetty malli toimii kohtuullisen hyvin omalle aineistolle. 3) Hienosäädetyllä mallilla on ongelmia toisensa peittävien tai takaseinää vasten kuvattujen pelaajien tunnistamisessa. 4) Mallin opettaminen vaihtelevammalla aineistolla parantaa tuloksia uudenlaisia esineitä tunnistettaessa. Tutkimuksen tulosten perusteella: 1) Valmiiksi opetettu malli oli hyödytön tämän datan käsittelyssä. 2) Hienosäädetty malli toimi kohtalaisen hyvin. 3) Hienosäädetyllä mallilla oli ongelmia toisensa peittävien tai takaseinää vasten kuvattujen pelaajien tunnistamisessa. 4) Yhdestä pelistä saadulla aineistolla opetettu malli tunnisti pelaajat toisesta pelistä kohtalaisen hyvin. Mallin toiminta ei juurikaan parantunut kun se opetettiin molemmista peleistä koostetulla aineistolla. Muita mallityyppejä (kuten Faster R-CNN model) pitäisi testata tällä datalla
14
Zilch, Lloyd W. "Image charge detection and image charge detection mass spectrometry." [Bloomington, Ind.] : Indiana University, 2008. http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqdiss&rft_dat=xri:pqdiss:3344616.
Full textAbstract:
Thesis (Ph. D.)--Indiana University, Dept. of Chemistry, 2008.Title from home page (viewed Oct. 8, 2009). Source: Dissertation Abstracts International, Volume: 70-02, Section: B, page: 0994. Adviser: Martin F. Jarrold.
15
Yerra, Rajiv. "Detecting Similar HTML Documents Using A Sentence-Based Copy Detection Approach." Diss., CLICK HERE for online access, 2005. http://contentdm.lib.byu.edu/ETD/image/etd977.pdf.
Full text
16
Jenner, Mareike. ""Follow the evidence"? : methods of detection in American TV detective drama." Thesis, Aberystwyth University, 2013. http://hdl.handle.net/2160/973dbcaf-5796-42c5-a044-b51252c91b66.
Full textAbstract:
This thesis deals with methods of detection i.e. the mode of investigation employed to catch a criminal in American detective dramas on television. It divides methods of detection into the categories of ‘rational-scientific’ and ‘irrational-subjective’. ‘Rational-scientific’ methods of detection are linked to the literary tradition of Golden Age fiction and suggest an analytical distance to the crime. ‘Irrational-subjective’ methods are linked to a hard-boiled tradition and suggest (often emotional) ‘closeness’ to the victim, suspects or witnesses. Drawing on the work of Michel Foucault, John Fiske and Jason Mittell, this thesis views genre as discourse. As such, television genre is viewed as always changing and intersecting with a variety of other discourses, for example, representing social and political debates, shifts within the television industry and mirroring ideologies of ‘truth-finding’. It analyses methods of detection as a discourse internal to the genre, as a genre convention, as well as external to the genre i.e. as relating to discourses regarding social, political and industrial developments. It also explores how methods of detection, as an expression of ideologies of ‘truth-finding’, reveal how a specific series may be positioned in relationship to modern post-Enlightenment and postmodern discourses. A number of texts from different historical moments (Dragnet [NBC, 1951-1959], Quincy, M.E. [NBC, 1976-1983], CSI: Crime Scene Investigation [CBS, 2000- ], Hill Street Blues [NBC, 1981-1987], Twin Peaks [ABC, 1990-1991] and The Shield [fX, 2002-2008]) are analysed as examples of how individual genre texts represent these shifts in attitudes towards ‘truth-finding’. In a final step, this thesis analyses The Wire (HBO, 2002-2008) and Dexter (Showtime, 2006- ) as dramas that represent a more recent shift in the representation of ideologies of ‘truth-finding’ that may formulate ‘alternative’ methods of detection and a possible epistemological shift in postmodern culture.
17
Granath, Linus, and Andreas Strid. "Detecting the presence of people in a room using motion detection." Thesis, Malmö högskola, Fakulteten för teknik och samhälle (TS), 2016. http://urn.kb.se/resolve?urn=urn:nbn:se:mau:diva-20099.
Full textAbstract:
Companies face a problem where employees reserve rooms and do not show up, which leadsto money and resources loss for the companies. An application capable of detecting thepresence of people in a room could solve this problem.This thesis details the process of building an Android application capable of detectingthe presence of people in a static room using motion detection. The application wasdeveloped through a five-staged process and evaluated by performing experiments whichmeasured the accuracy of the application.The finished application is installed on a Sony Xperia M4 Aqua device which is mountedhigh up on a wall in a conference room where the application takes images of the room. Theapplication is connected to a Google Drive account where the application uploads acquiredimages with an appropriate label. The application achieved an accuracy of 94.18% in anexperiment where 550 images where taken automatically by the application in differentconference rooms with and without people inside them
18
Liu, Jessamyn. "Anomaly detection methods for detecting cyber attacks in industrial control systems." Thesis, Massachusetts Institute of Technology, 2020. https://hdl.handle.net/1721.1/129055.
Full textAbstract:
Thesis: S.M., Massachusetts Institute of Technology, Sloan School of Management, Operations Research Center, September, 2020Cataloged from PDF version of thesis.
Includes bibliographical references (pages 119-123).
Industrial control systems (ICS) are pervasive in modern society and increasingly under threat of cyber attack. Due to the critical nature of these systems, which govern everything from power and wastewater plants to refineries and manufacturing, a successful ICS cyber attack can result in serious physical consequences. This thesis evaluates multiple anomaly detection methods to quickly and accurately detect ICS cyber attacks. Two fundamental challenges in developing ICS cyber attack detection methods are the lack of historical attack data and the ability of attackers to make their malicious activity appear normal. The goal of this thesis is to develop methods which generalize well to anomalies that are not included in the training data and to increase the sensitivity of detection methods without increasing the false alarm rate. The thesis presents and analyzes a baseline detection method, the multivariate Shewhart control chart, and four extensions to the Shewhart chart which use machine learning or optimization methods to improve detection performance. Two of these methods, stationary subspace analysis and maximized ratio divergence analysis, are based on dimensionality reduction techniques, and an additional model-based method is implemented using residuals from LASSO regression models. The thesis also develops an ensemble method which uses an optimization formulation to combine the output of multiple models in a way that minimizes detection delay. When evaluated on 380 samples from the Kasperskey Tennessee Eastman process dataset, a simulated chemical process that includes disruptions from cyber attacks, the ensemble method reduced detection delay on attack data by 12% (55 minutes) on average when compared to the baseline method and was 9% (42 minutes) faster on average than the method which performed best on training data.
by Jessamyn Liu.
S.M.
S.M. Massachusetts Institute of Technology, Sloan School of Management, Operations Research Center
19
Pretz, John. "Detection of atmospheric muon neutrinos with the IceCube 9-String Detector." College Park, Md. : University of Maryland, 2006. http://hdl.handle.net/1903/4163.
Full textAbstract:
Thesis (Ph. D.) -- University of Maryland, College Park, 2006.Thesis research directed by: Physics. Title from t.p. of PDF. Includes bibliographical references. Published by UMI Dissertation Services, Ann Arbor, Mich. Also available in paper.
20
Erickson, Anna S. "Remote detection of fissile material : Cherenkov counters for gamma detection." Thesis, Massachusetts Institute of Technology, 2011. http://hdl.handle.net/1721.1/76496.
Full textAbstract:
Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Nuclear Science and Engineering, 2011.Cataloged from PDF version of thesis.
Includes bibliographical references (p. 161-167).
The need for large-size detectors for long-range active interrogation (Al) detection has generated interest in water-based detector technologies. AI is done using external radiation sources to induce fission and to detect, identify, and characterize special nuclear material (SNM) through the gamma rays and neutrons emitted. Long-range applications require detectors with a large solid angle and an ability to significantly suppress lowenergy background from linear electron accelerators. Water Cherenkov Detectors (WCD) were selected because of their transportability, scalability, and an inherent energy threshold. The main objective of this thesis was to design a large-size WCD capable of detecting gamma rays and to demonstrate particle energy discrimination ability. WCD was modeled in detail using Geant4 for optimization purposes. The experimental detector is composed of an aluminum body with a high efficiency (98.5%) diffuse reflector. Cherenkov photons are detected with six 8" hemispherical Hamamatsu photomultiplier tubes (PMT). PMTs are calibrated using two monoenergetic LEDs. The detector was shown to successfully detect gamma rays of energies above the Cherenkov threshold. The detector was able to discriminate between various sources, such as ⁶⁰Co and ²³²Th, even though WCD are known for their poor energy resolution. The detector design and analysis was completed, and it was demonstrated both computationally and experimentally that it is possible to use WCD to detect and characterize gamma rays. One of the accomplishments of this thesis was demonstration of event reconstruction capability of the detector system. A full-detector model was created using Geant4 simulation toolkit. The performance of the detector was predicted using the model and then experimentally verified. The qualitative agreement between the model and the experiment was observed. The event reconstruction was an important part of the detector performance analysis. Post-experimental data processing was done using ROOT.
by Anna S. Erickson.
Ph.D.
21
Kaatz, Miriam. "Impedimetric DNA detection : towards improved detection schemes for sensor integration." Thesis, University of Edinburgh, 2015. http://hdl.handle.net/1842/17890.
Full textAbstract:
Detection of DNA by electrochemical impedance spectroscopy (EIS) has been reported by many authors and assays have been developed using lab setups. However, as for most detection assay methods there are issues to address to enable the development for the sensor market: Long time-to-result & high complexity for labelled assays and a lack of sensitivity and reproducibility for label-free assays. This work considers two different approaches to address the issues of time-to-result and assay complexity. The first part presents work on achieving rapid sequence-specific electrochemical detection of DNA hybridisation to complementary DNA on an electrode surface. To accomplish assay sensitivity to low DNA target concentrations, a signal amplification strategy is often necessary. One approach is to couple an enzyme to the hybridised target molecules and to deposit insoluble dyes in the subsequent enzymatic reaction, which enhances sensitivity through an increase in the impedance signal in presence of a redox mediator. The time typically taken for this process (20 – 40 min) precludes the use outside lab setups. Therefore, a protocol for sensitive detection in the presence of redox mediator is demonstrated on a practical timescale required for use in sensor applications. Based on these results a model for the fundamental understanding of the amplification reaction is presented which explains the retention of sensitivity at these enhanced timescales. This also enabled further optimisation of the assay for application in single base pair mismatch detection in biologically relevant sequences. Moreover, direct detection of the precipitate formation is demonstrated which enables real-time measurement of the enzymatic reaction without redox agent addition and with enhanced mismatch discrimination. The second part investigates the possibility to detect DNA non-sequence-specifically by non-Faradaic means. This approach aims at reducing assay complexity by establishing whether it is possible to sense the presence of polymeric DNA in solution by measuring changes in the properties of the electrochemical double layer without DNA surface hybridisation. In a sensor setup this approach could be linked to a polymerase chain reaction (PCR) to discriminate polymer from nucleotide monomer and thereby enable PCR progress to be monitored. In this work the response in the electrochemical double layer at the interface of blocked metal electrodes and solutions containing DNA are studied by means of EIS. Blocking layers were applied to the electrode surface to prevent unspecific adsorption of molecules and ions to the metal surface whilst preserving the sensitivity to detection of changes in the double layer. The characteristics of surface blocking layers on disposable electrodes are studied as they are key to understand the double layer properties at a blocked surface. A number of self-assembled monolayers are compared with respect to their temperature stability and their blocking characteristics at different potentials and ion concentrations. This established the basis to study the effect of the presence of, initially, a model polyelectrolyte and, ultimately, DNA on the double layer. Polyelectrolyte detection is successfully shown for the model polyelectrolyte, polyacrylic acid. DNA detection was more challenging and possible causes for deviation from the polyacrylic acid response are discussed.
22
Gonzalez-Garcia, Abel. "Image context for object detection, object context for part detection." Thesis, University of Edinburgh, 2018. http://hdl.handle.net/1842/28842.
Full textAbstract:
Objects and parts are crucial elements for achieving automatic image understanding. The goal of the object detection task is to recognize and localize all the objects in an image. Similarly, semantic part detection attempts to recognize and localize the object parts. This thesis proposes four contributions. The first two make object detection more efficient by using active search strategies guided by image context. The last two involve parts. One of them explores the emergence of parts in neural networks trained for object detection, whereas the other improves on part detection by adding object context. First, we present an active search strategy for efficient object class detection. Modern object detectors evaluate a large set of windows using a window classifier. Instead, our search sequentially chooses what window to evaluate next based on all the information gathered before. This results in a significant reduction on the number of necessary window evaluations to detect the objects in the image. We guide our search strategy using image context and the score of the classifier. In our second contribution, we extend this active search to jointly detect pairs of object classes that appear close in the image, exploiting the valuable information that one class can provide about the location of the other. This leads to an even further reduction on the number of necessary evaluations for the smaller, more challenging classes. In the third contribution of this thesis, we study whether semantic parts emerge in Convolutional Neural Networks trained for different visual recognition tasks, especially object detection. We perform two quantitative analyses that provide a deeper understanding of their internal representation by investigating the responses of the network filters. Moreover, we explore several connections between discriminative power and semantics, which provides further insights on the role of semantic parts in the network. Finally, the last contribution is a part detection approach that exploits object context. We complement part appearance with the object appearance, its class, and the expected relative location of the parts inside it. We significantly outperform approaches that use part appearance alone in this challenging task.
23
Jakobsson, Uno, and Martin Andréasson. "Wildlife Detection Network." Thesis, Högskolan i Halmstad, Sektionen för ekonomi och teknik (SET), 2012. http://urn.kb.se/resolve?urn=urn:nbn:se:hh:diva-17765.
Full textAbstract:
Traffic accidents where wild animals are involved represents between 60 and 80 percent of all reported accidents, depending on location in Sweden. In a country like Sweden, with a lot of forest, there is always risk of a collision with a wild animal. Imagine if you, as a road user, had the possibility to receive warnings when the risk of an accident according to statistics is extra high. Wildlife Detection Network is a wildlife warning system with an information service, which makes the whole concept unique. When an animal is approaching the road, it is registered by sensors, and warning lights along the road are lit to inform drivers of the potential danger. In conclusion, this is a direct warning to all drivers on the road where the system is placed. When an animal is registered by the sensors, information containing time, date, weather circumstances and coordinates are sent to a database. The database stores information about the animal activity in the area, and will read out activity patterns for the animals. For example, the risk for a collision might be higher between 6.00 and 8.00 AM when the temperature is about ten degrees. When you approach the measured area in your car, you will receive a warning in you smartphone or GPS-unit. The warning tells you that the risk of encountering a wild animal along the road is high during the current circumstances. The associated service works as a complement for those that further wants to reduce the risk of a wildlife accident. We are well aware of that wildlife accidents are a very complex and in particular intractable problem. The two of us behind Wildlife Detection Network are proud of our concept and we are hopeful that our system will contribute to a decrease in wildlife accidents in the future.
24
Ståhl, Björn. "Online Anomaly Detection." Thesis, Blekinge Tekniska Högskola, Avdelningen för för interaktion och systemdesign, 2006. http://urn.kb.se/resolve?urn=urn:nbn:se:bth-2825.
Full textAbstract:
Where the role of software-intensive systems has shifted from the traditional one of fulfilling isolated computational tasks, larger collaborative societies with interaction as primary resource, is gradually taking its place. This can be observed in anything from logistics to rescue operations and resource management, numerous services with key-roles in the modern infrastructure. In the light of this new collaborative order, it is imperative that the tools (compilers, debuggers, profilers) and methods (requirements, design, implementation, testing) that supported traditional software engineering values also adjust and extend towards those nurtured by the online instrumentation of software intensive systems. That is, to adjust and to help to avoid situations where limitations in technology and methodology would prevent us from ascertaining the well-being and security of systems that assists our very lives. Coupled with most perspectives on software development and maintenance is one well established member of, and complement to, the development process. Debugging; or the art of discovering, localising, and correcting undesirable behaviours in software-intensive systems, the need for which tend to far outlive development in itself. Debugging is currently performed based on a premise of the developer operating from a god-like perspective. A perspective that implies access and knowledge regarding source code, along with minute control over execution properties. However, the quality as well as accessibility of such information steadily decline with time as requirements, implementation, hardware components and their associated developers, all alike fall behind their continuously evolving surroundings. In this thesis, it is argued that the current practice of software debugging is insufficient, and as precursory action, introduce a technical platform suitable for experimenting with future methods regarding online debugging, maintenance and analysis. An initial implementation of this platform will then be used for experimenting with a simple method that is targeting online observation of software behaviour.
25
Wen, Shihua. "Semiparametric cluster detection." College Park, Md. : University of Maryland, 2007. http://hdl.handle.net/1903/7204.
Full textAbstract:
Thesis (Ph. D.) -- University of Maryland, College Park, 2007.Thesis research directed by: Mathematical Statistics. Title from t.p. of PDF. Includes bibliographical references. Published by UMI Dissertation Services, Ann Arbor, Mich. Also available in paper.
26
Malek, Esmaeili Mani. "Multimedia copy detection." Thesis, University of British Columbia, 2013. http://hdl.handle.net/2429/44581.
Full textAbstract:
Asmultimedia-sharing websites are becoming increasingly popular, content providers
get more concerned about the illegal distribution of their copyrighted contents. The
recent content-based multimedia fingerprinting technology has evolved as an important
tool for automatically detecting illegal copies of audio, image, and video signals.
Multimedia fingerprints are signatures that are extracted from an audio, image,
or video signal as a compact identifier of the signal. Therefore fingerprints should
have enough discriminating ability to identify a multimedia object among others.
At the same time they should be robust to modifications a multimedia signal might
be subjected, such as compression, cropping, format change, scaling, and other
signal processing operations. Robustness requires the fingerprints of a signal to
only depend on the signals perceptual content and not on its format, size, quality,
etc. This thesis proposes copy detection systems for audio and video signals and
addresses the robustness as well as the discrimination ability of these systems.
We first address audio fingerprinting and propose an algorithm that can detect
small snippets of audio signals. Simulation results show that, the extracted fingerprints
are robust to audio modifications including pitch shift and tempo change.
For severe modifications that existing algorithms have poor detection rates (around
20%), our proposed algorithm yields detection rates above 80%.
We then address video fingerprinting and propose an algorithm that extracts
robust and discriminant binary fingerprints. Simulation results show that the proposed
algorithm is faster and more accurate than the state-of-the-art with a high
true positive rate of over 97% and a low false positive rate below 1%.
Another challenge in multimedia fingerprinting is fingerprint retrieval, i.e. searching a huge fingerprint database (millions of fingerprints), for an accurate match for
a query fingerprint in a fast fashion. We propose a fast and accurate Nearest Neighbour
(NN) search algorithm for binary fingerprints (Hamming space). Tested on a
very large database of 80 million images, we showed that the proposed algorithm
is about 3 times faster than the state-of-the-art while at the same time it is 10 times
more accurate.
27
Foisy, André. "Robust collision detection." Thesis, McGill University, 1994. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=28746.
Full textAbstract:
The problem of locating collisions between computer modeled moving rigid bodies is considered. Each rigid body is modeled by the union of polytopes. Safe and reliable collision detection algorithms are constructed.Since the computer representation of real numbers is finite, an interval projective point is used to encompass all localization errors of a modeled point. An interval point is the elementary geometrical form from which all others are constructed. The Euclidean convex set spanned by an interval point is also a polytope.
The construction of a polytope relies on a robust convex hull algorithm. The computed hull is guaranteed to contain all interval projective points.
An extrusion based collision detection algorithm builds an AND-OR decision tree. Each leaf is a univariate function that expresses the relation between a moving point and a moving plane. Interval zero finding methods are applied to find the overlap and non-overlap portions of the trajectories of moving polytopes.
The swept-volume based collision detection algorithm depends on the construction of a convex approximation that comprises the real swept volume. To obtain a convex approximation, the convex hull algorithm is applied to the bounding volumes of the vertices of a moving polytope. Each bounding volume is an interval projective point.
Finally, both collision detection algorithms are tested in the context of generate-and-test path planning.
28
Espinosa-Romero, Arturo. "Situated face detection." Thesis, University of Edinburgh, 2001. http://hdl.handle.net/1842/6667.
Full textAbstract:
In the last twenty years, important advances have been made in the field of automatic face processing, given the importance of human faces for personal identification, emotional expression and verbal and non verbal communication. The very first step in a face processing algorithm is the detection of faces; while this is a trivial problem in controlled environments, the detection of faces in real environments is still a challenging task. Until now, the most successful approaches for face detection represent the face as a grey-level pattern, and the problem itself is considered as the classification between "face" and "non-face" patterns. Satisfactory results have been achieved in this area. The main disadvantage is that an exhaustive search has to be done on each image in order to locate the faces. This search normally involves testing every single position on the image at different scales, and although this does not represent an important drawback in off-line face processing systems, in those cases where a real-time response is needed it is still a problem. In the different proposed methods for face detection, the "observer" is a disembodied entity, which holds no relationship with the observed scene. This thesis presents a framework for an efficient location of faces in real scenes, in which, by considering both the observer to be situated in the world, and the relationships that hold between the two, a set of constraints in the search space can be defined. The constraints rely on two main assumptions; first, the observer can purposively interact with the world (i.e. change its position relative to the observed scene) and second, the camera is fully calibrated. The first source constraint is the structural information about the observer environment, represented as a depth map of the scene in front of the camera. From this representation the search space can be constrained in terms of the range of scales where a face might be found as different positions in the image. The second source of constraint is the geometrical relationship between the camera and the scene, which allows us to project a model of the subject into the scene in order to eliminate those areas where faces are unlikely to be found. In order to test the proposed framework, a system based on the premises stated above was constructed. It is based on three different modules: a face/non-face classifier, a depth estimation module and a search module. The classifier is composed of a set of convolutional neural networks (CNN) that were trained to differentiate between face and non-face patterns, the depth estimation modules uses a multilevel algorithm to compute the scene depth map from a sequence of images captured the depth information and the subject model into the image where the search will be performed in order to constrain the search space. Finally, the proposed system was validated by running a set of experiments on the individual modules and then on the whole system.
29
Kalal, Zdenek. "Tracking learning detection." Thesis, University of Surrey, 2011. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.540948.
Full text
30
Ohlson, Frida, and Nadim Al-Mosawi. "Occupant Detection System." Thesis, Högskolan i Halmstad, Bio- och miljösystemforskning (BLESS), 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:hh:diva-28617.
Full textAbstract:
The purpose of the ODS (Occupant Detection System) project is to develop a car safety camera system that is able to discriminate human occupants in order to activate safety features depending on the human size. In order to discriminate the size of an occupant anthropometric measurements need to be performed. The aim of this study was to investigate the possibility of anthropometric measurements of human height and weight with a Kinect for Windows v2 sensor for discrimination of different occupants inside a vehicle. The goal was to find valid anthropometric methods for determination of human height and weight from landmarks on the upper body, then test if it is possible to perform these methods with the Kinect for Windows v2 sensor. The execution of this study was performed as a literature study with anthropometric tests on ATD-dummies (anthropomorphic test device) and on human test subjects. Measurements were performed first physically and then with the Kinect v2 sensor to obtain data in form of distance of 8 regions of the body. Three tests were performed, first on dummies, second was a pilot study and last the measurement study. The result revealed that it is hard to estimate human weight from body landmarks due to lack of information, therefor no tests were performed in this area. For height the result showed that the most valid methods were the measurements on arm span and ulna, both on physical measurements and with the camera. The conclusion is that it is possible estimating height from body landmarks but the positioning of the camera needs to be change in order for the measurements to be more accurate. This study has contributed to a greater understanding of measurement technology, automotive safety and anthropometric measurements.
31
Elsenbroich, Corinna Julia. "Instinct for detection." Thesis, King's College London (University of London), 2006. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.430430.
Full text
32
Lafis, S. "Rapid microbial detection." Thesis, Cranfield University, 1992. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.357505.
Full text
33
Doney, George Daniel. "Acoustic boiling detection." Thesis, Massachusetts Institute of Technology, 1994. http://hdl.handle.net/1721.1/28110.
Full text
34
Alzarooni, K. M. A. "Malware variant detection." Thesis, University College London (University of London), 2012. http://discovery.ucl.ac.uk/1347243/.
Full textAbstract:
Malware programs (e.g., viruses, worms, Trojans, etc.) are a worldwide epidemic. Studies and statistics show that the impact of malware is getting worse. Malware detectors are the primary tools in the defence against malware. Most commercial anti-malware scanners maintain a database of malware patterns and heuristic signatures for detecting malicious programs within a computer system. Malware writers use semantic-preserving code transformation (obfuscation) techniques to produce new stealth variants of their malware programs. Malware variants are hard to detect with today's detection technologies as these tools rely mostly on syntactic properties and ignore the semantics of malicious executable programs. A robust malware detection technique is required to handle this emerging security threat. In this thesis, we propose a new methodology that overcomes the drawback of existing malware detection methods by analysing the semantics of known malicious code. The methodology consists of three major analysis techniques: the development of a semantic signature, slicing analysis and test data generation analysis. The core element in this approach is to specify an approximation for malware code semantics and to produce signatures for identifying, possibly obfuscated but semantically equivalent, variants of a sample of malware. A semantic signature consists of a program test input and semantic traces of a known malware code. The key challenge in developing our semantics-based approach to malware variant detection is to achieve a balance between improving the detection rate (i.e. matching semantic traces) and performance, with or without the e ects of obfuscation on malware variants. We develop slicing analysis to improve the construction of semantic signatures. We back our trace-slicing method with a theoretical result that shows the notion of correctness of the slicer. A proof-of-concept implementation of our malware detector demonstrates that the semantics-based analysis approach could improve current detection tools and make the task more di cult for malware authors. Another important part of this thesis is exploring program semantics for the selection of a suitable part of the semantic signature, for which we provide two new theoretical results. In particular, this dissertation includes a test data generation method that works for binary executables and the notion of correctness of the method.
35
Kemp, E., M. Floyd, E. McCord-Duncan, Beth Ann Bailey, Ivy A. Click, and J. Gorniewicz. "IPV Detection Strategies." Digital Commons @ East Tennessee State University, 2007. https://dc.etsu.edu/etsu-works/6406.
Full text
36
Floyd, M., E. Kemp, E. McCord-Duncan, Beth Ann Bailey, Ivy A. Click, and J. Gorniewicz. "IPV Detection Strategies." Digital Commons @ East Tennessee State University, 2007. https://dc.etsu.edu/etsu-works/6416.
Full text
37
Ent, Petr. "Voice Activity Detection." Master's thesis, Vysoké učení technické v Brně. Fakulta informačních technologií, 2009. http://www.nusl.cz/ntk/nusl-235483.
Full textAbstract:
Práce pojednává o využití support vector machines v detekci řečové aktivity. V první části jsou zkoumány různé druhy příznaků, jejich extrakce a zpracování a je nalezena jejich optimální kombinace, která podává nejlepší výsledky. Druhá část představuje samotný systém pro detekci řečové aktivity a ladění jeho parametrů. Nakonec jsou výsledky porovnány s dvěma dalšími systémy, založenými na odlišných principech. Pro testování a ladění byla použita ERT broadcast news databáze. Porovnání mezi systémy bylo pak provedeno na databázi z NIST06 Rich Test Evaluations.
38
Matos, Diogo Silva. "Person detection system." Master's thesis, Universidade de Aveiro, 2017. http://hdl.handle.net/10773/23853.
Full textAbstract:
mestrado em Engenharia Eletrónica e TelecomunicaçõesO RADAR é para fins militares já relativamente antigo que sofreu um grande impulso durante a Segunda Guerra Mundial. Hoje em dia existe um forte desenvolvimento no RADAR em aplicações de navegação ou vigilância/segurança. Esta dissertação surge no seguimento destas novas aplicações, em que se pretende o desenvolvimento de um RADAR de baixo custo que permita ao utilizador detetar pessoas, bem como, os seus movimentos através de paredes ou objetos opacos. O desenvolvimento deste RADAR recaiu em tecnologias emergentes como antenas adaptativas e rádio definido por software que permitem uma grande versatilidade e adaptação em termos de aplicações. A utilização de um RADAR com múltiplas entradas e múltiplas saídas fornece uma maior diversidade de informação que garante mais probabilidades de deteção. A aplicação de técnicas digitais de beamforming, possibilita conhecer a posição e o movimento da pessoa. Com a implementação destas técnicas um protótipo capaz de detetar pessoas e os seus movimentos através de paredes e tijolos foi desenvolvido com sucesso solucionando o problema inicial. Na fase de projeto de RADAR houve necessidade de caracterizar a propagação de ondas de rádio em materiais de construção, como tijolos e madeira, medindo-se a sua atenuação. Deste modo foi possível fazer o balanço de potencia para varios canarios.
The RADAR is already relatively old for military purposes that underwent a major development during World War II. Nowadays there is a strong development in RADAR in navigation or surveillance/security applications. This dissertation follows on from these new applications, which aim to develop a low cost RADAR that allows the user to detect people as well as their movements through walls or opaque objects. The development of this RADAR has relied on emerging technologies such as adaptive antennas and SDR that allow for great versatility and adaptation in terms of applications. The use of a MIMO RADAR provides a greater diversity of information that guarantees more probabilities of detection and the application of digital techniques of beamforming, allows to know the position and the movement of the person. With the implementation of these techniques a prototype capable of detecting people and their movements through walls and bricks was successfully developed solving the initial problem. In the RADAR design phase it was necessary to characterize the propagation of radio waves in building materials, such as bricks and wood, by measuring their attenuation. In this way it was possible to perform the power balance for several scenarios.
39
O'Keefe, Eion Seiorse. "Polymer chemiluminescence detection." Thesis, University of Sussex, 1987. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.238667.
Full text
40
Park, Sean. "Neural malware detection." Thesis, Federation University Australia, 2019. http://researchonline.federation.edu.au/vital/access/HandleResolver/1959.17/173759.
Full textAbstract:
At the heart of today’s malware problem lies theoretically infinite diversity created by metamorphism. The majority of conventional machine learning techniques tackle the problem with the assumptions that a sufficiently large number of training samples exist and that the training set is independent and identically distributed. However, the lack of semantic features combined with the models under these wrong assumptions result largely in overfitting with many false positives against real world samples, resulting in systems being left vulnerable to various adversarial attacks. A key observation is that modern malware authors write a script that automatically generates an arbitrarily large number of diverse samples that share similar characteristics in program logic, which is a very cost-effective way to evade detection with minimum effort. Given that many malware campaigns follow this paradigm of economic malware manufacturing model, the samples within a campaign are likely to share coherent semantic characteristics. This opens up a possibility of one-to-many detection. Therefore, it is crucial to capture this non-linear metamorphic pattern unique to the campaign in order to detect these seemingly diverse but identically rooted variants. To address these issues, this dissertation proposes novel deep learning models, including generative static malware outbreak detection model, generative dynamic malware detection model using spatio-temporal isomorphic dynamic features, and instruction cognitive malware detection. A comparative study on metamorphic threats is also conducted as part of the thesis. Generative adversarial autoencoder (AAE) over convolutional network with global average pooling is introduced as a fundamental deep learning framework for malware detection, which captures highly complex non-linear metamorphism through translation invariancy and local variation insensitivity. Generative Adversarial Network (GAN) used as a part of the framework enables oneshot training where semantically isomorphic malware campaigns are identified by a single malware instance sampled from the very initial outbreak. This is a major innovation because, to the best of our knowledge, no approach has been found to this challenging training objective against the malware distribution that consists of a large number of very sparse groups artificially driven by arms race between attackers and defenders. In addition, we propose a novel method that extracts instruction cognitive representation from uninterpreted raw binary executables, which can be used for oneto- many malware detection via one-shot training against frequency spectrum of the Transformer’s encoded latent representation. The method works regardless of the presence of diverse malware variations while remaining resilient to adversarial attacks that mostly use random perturbation against raw binaries. Comprehensive performance analyses including mathematical formulations and experimental evaluations are provided, with the proposed deep learning framework for malware detection exhibiting a superior performance over conventional machine learning methods. The methods proposed in this thesis are applicable to a variety of threat environments here artificially formed sparse distributions arise at the cyber battle fronts.Doctor of Philosophy
41
Barascud, Nicolas. "Auditory pattern detection." Thesis, University College London (University of London), 2014. http://discovery.ucl.ac.uk/1458387/.
Full textAbstract:
The work presented in this doctoral thesis uses behavioural methods and neuroimaging to investigate how human listeners detect patterns and statistical regularities in complex sound sequences. Temporal pattern analysis is essential to sensory processing, especially listening, since most auditory signals only have meaning as sequences over time. Previous evidence suggests that the brain is sensitive to the statistics of sensory stimulation. However, the process through which this sensitivity arises is largely unknown. This dissertation is organised as follows: Chapter 1 reviews fundamental principles of auditory scene analysis and existing models of regularity processing to constrain the scientific questions being addressed. Chapter 2 introduces the two neuroimaging techniques used in this work, magnetoencephalography (MEG) and functional Magnetic Resonance Imaging (fMRI). Chapters 3-6 are experimental sections. In Chapter 3, a novel stimulus is presented that allows probing listeners’ sensitivity to the emergence and disappearance of complex acoustic patterns. Pattern detection performance is evaluated behaviourally, and systematically compared with the predictions of an ideal observer model. Chapters 4 and 5 describe the brain responses measured during processing of those complex regularities using MEG and fMRI, respectively. Chapter 6 presents an extension of the main behavioural task to the visual domain, which allows pattern detection to be compared in audition and vision. Chapter 7 concludes with a general discussion of the experimental results and provides directions for future research. Overall, the results are consistent with predictive coding accounts of perceptual inference and provide novel neurophysiological evidence for the brain's exquisite sensitivity to stimulus context and its capacity to encode high-order structure in sensory signals.
42
Rogers, Stuart Craig. "Defect Detection Microscopy." BYU ScholarsArchive, 2010. https://scholarsarchive.byu.edu/etd/2256.
Full textAbstract:
The automotive industry's search for stronger lighter materials has been hampered in its desire to make greater use of Magnesium alloys by their poor formability below 150°C. One current challenge is to identify the complex structure and deformation mechanisms at work and determine which of these are primary contributors to the nucleation of defects. Orientation Imaging Microscopy has been the most accessible tool for microstructural analysis over the past 15 years. However, using OIM to analyze defect nucleation sites requires prior knowledge of where the defects will occur because once the defects nucleate the majority of microstructural information is destroyed. This thesis seeks to contribute to the early detection of nucleation sites via three mechanisms: 1. Detection of cracks that have already nucleated, 2. Detection of surface topography changes that may indicate imminent nucleation and 3. Beam control strategies for efficiently finding areas of interest in a scan. Successive in-situ OIM scans of a consistent sample region while strain is increased, while using the three techniques developed in this thesis, will be employed in future work to provide a powerful defect analysis tool. By analyzing retrieved EBSD patterns we are able to locate defect / crack sites via shadowing on the EBSD patterns. Furthermore, topographical features (and potentially regions of surface roughening) can be detected via changes in intensity metrics and image quality. Topographical gradients are currently only detectable in line with the beam incidence. It is therefore suggested that the tensile specimens to be examined are orientated such that the resulting shear bands occur preferentially to this direction. The ability to refine the scan around these areas of interest has been demonstrated via an off-line adaptive scan routine that is implemented via the custom scan tool. A first attempt at a defect detection framework has been outlined and coded into MATLAB. These tools offer a first step to accessing the information about defect nucleation that researchers are currently seeking.
43
Butler, Joseph G. "Automated Fingertip Detection." BYU ScholarsArchive, 2012. https://scholarsarchive.byu.edu/etd/3164.
Full textAbstract:
One of the oldest biometrics that has been used to uniquely identify a person is their fingerprint. Recent developments in research on fingerprint collection have made it possible to collect fingerprint data from a stand-off digital image. Each of the techniques developed so far have relied on either a very controlled capture environment to ensure only a single fingertip is collected or manual cropping of the image down to the fingertip. The main body of the research focuses on extracting the fingerprint itself. If fingerprint collection via digital image is ever to be fielded in the real world on such devices as smart phones or tablets it will be necessary for the software to automatically detect a single or multiple fingertips in an image and isolate them for extracting the fingerprint. We introduce an automatic fingertip detection algorithm that couples image processing techniques with a machine learning capability to successfully identify varying numbers of fingertips in digital images. Our algorithm proves that while it is difficult to remove all constraints from the capture environment it is achievable with the method we have developed and we can achieve a recall of 69.77% at a precision of 78.95%. This gives us the important capability to detect varying numbers of fingertips in an image and provide a crucial piece in what could be a complete automated fingerprint recognition system.
44
Park, Jea Woo. "Lithography Hotspot Detection." PDXScholar, 2017. https://pdxscholar.library.pdx.edu/open_access_etds/3781.
Full textAbstract:
The lithography process for chip manufacturing has been playing a critical role in keeping Moor's law alive. Even though the wavelength used for the process is bigger than actual device feature size, which makes it difficult to transfer layout patterns from the mask to wafer, lithographers have developed a various technique such as Resolution Enhancement Techniques (RETs), Multi-patterning, and Optical Proximity Correction (OPC) to overcome the sub-wavelength lithography gap.
However, as feature size in chip design scales down further to a point where manufacturing constraints must be applied to early design phase before generating physical design layout. Design for Manufacturing (DFM) is not optional anymore these days. In terms of the lithography process, circuit designer should consider making their design as litho-friendly as possible.
Lithography hotspot is a place where it is susceptible to have fatal pinching (open circuit) or bridging (short circuit) error due to poor printability of certain patterns in a design layout. To avoid undesirable patterns in layout, it is mandatory to find hotspots in early design stage.
One way to find hotspots is to run lithography simulation on a layout. However, lithography simulation is too computationally expensive for full-chip design. Therefore, there have been suggestions such as pattern matching and machine learning (ML) technique for an alternative and practical hotspot detection method. Pattern matching is fast and accurate. Large hotspot pattern library is utilized to find hotspots. Its drawback is that it can not detect hotspots that are unseen before. On contrast, ML is effective to find previously unseen hotspots, but it may produce false positives.
This research presents a novel geometric pattern matching methodology using edge driven dissected rectangles and litho award machine learning for hotspot detection.
1. Edge Driven Dissected Rectangles (EDDR) based pattern matching
EDDR pattern matching employs member concept inside a pattern bounding box. Unlike the previous pattern matching, the idea proposed in this thesis uses simple Design Rule Check (DRC) operations to create member rectangles for pattern matching. Our approach shows significant speedup against a state-of-art commercial pattern matching tool as well as other methods. Due to its simple DRC edge operation rules, it is flexible for fuzzy pattern match and partial pattern match, which enable us to check previously unseen hotspots as well as the exact pattern match.
2. Litho-aware Machine Learning
A new methodology for machine learning (ML)-based hotspot detection harnesses lithography information to build SVM (Support Vector Machine) during its learning process. Unlike the previous research that uses only geometric information or requires a post-OPC (Optical Proximity Correction) mask, our method utilizes detailed optical information but bypasses post-OPC mask by sampling latent image intensity and use those points to train an SVM model. Our lithography-aware machine learning guides learning process using actual lithography information combined with lithography domain knowledge. While the previous works for SVM modeling to identify hotspots have used only geometric related information, which is not directly relevant to the lithographic process, our SVM model was trained with lithographic information which has a direct impact on causing pinching or bridging hotspots. Furthermore, rather than creating a monolithic SVM trying to cover all hotspot patterns, we utilized lithography domain knowledge and separated hotspot types such as HB(Horizontal Bridging), VB (Vertical Bridging), HP(Horizontal Pinching), and VP(Vertical Pinching) for our SVM model. Out results demonstrated high accuracy and low false alarm, and faster runtime compared with methods that require a post-OPC mask. We also showed the importance of lithography domain knowledge to train ML for hotspot detection.
45
Garcia, Hurtado J. (Juan). "Indoor outdoor detection." Master's thesis, University of Oulu, 2019. http://jultika.oulu.fi/Record/nbnfioulu-201906062479.
Full textAbstract:
Abstract. This thesis shows a viable machine learning model that detects Indoor or Outdoor on smartphones. The model was designed as a classification problem and it was trained with data collected from several smartphone sensors by participants of a field trial conducted. The data collected was labeled manually either indoor or outdoor by the participants themselves. The model was then iterated over to lower the energy consumption by utilizing feature selection techniques and subsampling techniques. The model which uses all of the data achieved a 99 % prediction accuracy, while the energy efficient model achieved 92.91 %. This work provides the tools for researchers to quantify environmental exposure using smartphones.
46
姚志霖. "Vehicle detection using endpoint detection method for FMCW radar detector." Thesis, 2010. http://ndltd.ncl.edu.tw/handle/77448172476508925617.
Full text
47
Karademir, SARUHAN. "DETECTING PDF JAVASCRIPT MALWARE USING CLONE DETECTION." Thesis, 2013. http://hdl.handle.net/1974/8387.
Full textAbstract:
One common vector of malware is JavaScript in Adobe Acrobat (PDF) files. In this thesis, we investigate using near-miss clone detectors to find this malware. We start by collecting a set of PDF files containing JavaScript malware and a set with clean JavaScript from the VirusTotal repository. We use the NiCad clone detector to find the classes of clones in a small subset of the malicious PDF files. We evaluate how clone classes can be used to find similar malicious files in the rest of the malicious collection while avoiding files in the benign collection. Our results show that a 10% subset training set produced 75% detection of previously known malware with 0% false positives. We also used the NiCad as a pattern matcher for reflexive calls common in JavaScript malware. Our results show a 57% detection of malicious collection with no false positives. When the two experiments’ results are combined, the total coverage of malware rises to 85% and maintains 100% precision. The results are heavily affected by the third-party PDF to JavaScript extractor used. When only successfully extracted PDFs are considered, recall increases to 99% and precision remains at 100%.Thesis (Master, Electrical & Computer Engineering) -- Queen's University, 2013-09-30 11:50:15.156
48
Lotspeich, Erica H. "Evaluation of the Odor Compounds Sensed by Explosive-Detecting Canines." 2011. http://hdl.handle.net/1805/2473.
Full textAbstract:
Indiana University-Purdue University Indianapolis (IUPUI)Trained canines are commonly used as biological detectors for explosives; however, there are some areas of uncertainty that have led to difficulties in canine training and testing. Even though a standardized container for determining the accuracy of explosives-detecting canines has already been developed, the factors that govern the amount of explosive vapor that is present in the system are often uncertain. This has led to difficulties in comparing the sensitivity of canines to one another as well as to analytical instrumentation, despite the fact that this container has a defined headspace and degree of confinement of the explosive. For example, it is a common misconception that the amount of explosive itself is the chief contributor to the amount of odor available to a canine. In fact, odor availability depends not only on the amount of explosive material, but also the explosive vapor pressure, the rate with which the explosive vapor is transported from its source and the degree to which the explosive is confined. In order to better understand odor availability, headspace GC/MS and mass loss experiments were conducted and the results were compared to the Ideal Gas Law and Fick’s Laws of Diffusion. Overall, these findings provide increased awareness about availability of explosive odors and the factors that affect their generation; thus, improving the training of canines. Another area of uncertainty deals with the complexity of the odor generated by the explosive, as the headspace may consist of multiple chemical compounds due to the extent of explosive degradation into more (or less) volatile substances, solvents, and plasticizers. Headspace (HS) and solid phase microextraction (SPME) coupled with gas chromatography/mass spectrometry (GC/MS) were used to determine what chemical compounds are contained within the headspace of an explosive as well as NESTT (Non-Hazardous Explosive for Security Training and Testing) products. This analysis concluded that degradation products, plasticizers, and taggants are more common than their parent explosive.
49
Chiu, Chin-Chi, and 邱敬棋. "Using Edge Detection Combined with Feature Detection for Moving Object Detection." Thesis, 2017. http://ndltd.ncl.edu.tw/handle/b32ygd.
Full textAbstract:
碩士國立臺灣師範大學
機電工程學系
105
This thesis is detecting object for moving images. Nowadays, there are many methods for moving object detection on surveillance, and the method used is to find features and then to use the motion of those features between images to calculate features points moving. But the feature points sometimes are more difficult to define because the objects moving are easy to make images blur. Especially, when the objects may not be known in advance. In this thesis, using SURF algorithm defines the features of motional images because it detecting speed is faster than SIFT. But whether it is SIFT or SURF when the detected object moves, the matching result is not as good as expected because the objects may have incorrect feature points on moving. In the thesis, we provide edge and feature detection to combine for increasing the feature matching. In addition, this study we use a lot of different detection to detect and calculate the correct feature points to analyze. In experiment, we can further understand our methods getting the better ability to identify compared to the traditional methods.
50
Teh-Chung, Chen. "Detecting Visually Similar Web Pages: Application to Phishing Detection." Phd thesis, 2011. http://hdl.handle.net/10048/1682.
Full textAbstract:
We propose a novel approach for detecting visual similarity between two web pages. The proposed approach applies Gestalt theory and considers a webpage as a single indivisible entity. The concept of supersignals, as a realization of Gestalt principles, supports our contention that web pages must be treated as indivisible entities. We objectify, and directly compare, these indivisible supersignals using algorithmic complexity theory. We apply our new approach to the domain of anti-Phishing technologies, which at once gives us both a reasonable ground truth for the concept of “visually similar,” and a high-value application of our proposed approach.
Phishing attacks involve sophisticated, fraudulent websites that are realistic enough to fool a significant number of victims into providing their account credentials. There is a constant tug-of-war between anti-Phishing researchers who create new schemes to detect Phishing scams, and Phishers who create countermeasures. Our approach to Phishing detection is based on one major signature of Phishing webpage which can not be easily changed by those con artists –Visual Similarity. The only way to fool this significant characteristic appears to be to make a visually dissimilar Phishing webpage, which also reduces the successful rate of the Phishing scams or their criminal profits dramatically. For this reason, our application appears to be quite robust against a variety of common countermeasures Phishers have employed. To verify the practicality of our proposed method, we perform a large-scale, real-world case study, based on “live” Phish captured from the Internet.
Compression algorithms (as a practical operational realization of algorithmic complexity theory) are a critical component of our approach. Out of the vast number of compression techniques in the literature, we must determine which compression technique is best suited for our visual similarity problem. We therefore perform a comparison of nine compressors (including both 1-dimensional string compressors and 2-dimensional image compressors). We finally determine that the LZMA algorithm performs best for our problem.
With this determination made, we test the LZMA-based similarity technique in a realistic anti-Phishing scenario. We construct a whitelist of protected sites, and compare the performance of our similarity technique when presented with a) some of the most popular legitimate sites, and b) live Phishing sites targeting the protected sites. We found that the accuracy of our technique is extremely high in this test; the true positive and false positive rates reached 100% and 0.8%, respectively.
We finally undertake a more detailed investigation of the LZMA compression technique. Other authors have argued that compression techniques map objects to an implicit feature space consisting of the dictionary elements generated by the compressor. In testing this possibility on live Phishing data, we found that derived variables computed directly from the dictionary elements were indeed excellent predictors. In fact, by taking advantage of the specific characteristic of dictionary compression algorithm, we slightly improve on our accuracy when using a modified/refined LZMA algorithm for our already perfect NCD classification application.Software Engineering and Intelligent Systems