When forensic examiners try to identify the perpetrator of a felony, they use individual facial marks when comparing the suspect with the perpetrator. Facial marks are often used for identification and they are nowadays found manually. To speed up this process, it is desired to detect interesting facial marks automatically. This master thesis describes a method to automatically detect and separate permanent and non-permanent marks. It uses a fast radial symmetry algorithm as a core element in the mark detector. After candidate skin mark extraction, the false detections are removed depending on their size, shape and number of hair pixels. The classification of the skin marks is done with a support vector machine and the different features are examined. The results show that the facial mark detector has a good recall while the precision is poor. The elimination methods of false detection were analysed as well as the different features for the classifier. One can conclude that the color of facial marks is more relevant than the structure when classifying them into permanent and non-permanent marks. / När forensiker försöker identifiera förövaren till ett brott använder de individuella ansiktsmärken när de jämför den misstänkta med förövaren. Dessa ansiktsmärken identifieras och lokaliseras oftast manuellt idag. För att effektivisera denna process, är det önskvärt att detektera ansiktsmärken automatiskt. I rapporten beskrivs en framtagen metod som möjliggör automatiskt detektion och separation av permanenta och icke-permanenta ansiktsmärken. Metoden som är framtagen använder en snabb radial symmetri algoritm som en huvuddel i detektorn. När kandidater av ansiktsmärken har tagits, elimineras alla falska detektioner utifrån deras storlek, form och hårinnehåll. Utifrån studiens resultat visar sig detektorn ha en god känslighet men dålig precision. Eliminationsmetoderna av falska detektioner analyserades och olika attribut användes till klassificeraren. I rapporten kan det fastställas att färgskiftningar på ansiktsmärkena har en större inverkan än formen när det gäller att sortera dem i permanenta och icke-permanenta märken.

The usage of 3D modelling is increasing fast, both for civilian and military areas, such as navigation, targeting and urban planning. When creating a 3D model from satellite images, clouds canbe problematic. Thus, automatic detection ofclouds inthe imagesis ofgreat use. This master thesis was carried out at Vricon, who produces 3D models of the earth from satellite images.This thesis aimed to investigate if Support Vector Machines could classify pixels into cloud or non-cloud, with a combination of texture and color as features. To solve the stated goal, the task was divided into several subproblems, where the first part was to extract features from the images. Then the images were preprocessed before fed to the classifier. After that, the classifier was trained, and finally evaluated.The two methods that gave the best results in this thesis had approximately 95 % correctly classified pixels. This result is better than the existing cloud segmentation method at Vricon, for the tested terrain and cloud types.

Submitted by Sabrina de Cassia Martins null (sabrismartins@gmail.com) on 2017-02-08T15:06:43Z No. of bitstreams: 1 Tese-Sabrina Martins_VERSÃO FINAL.pdf: 4892311 bytes, checksum: ded93001eeec82424d2a73ec7626a3d8 (MD5) / Approved for entry into archive by LUIZA DE MENEZES ROMANETTO (luizamenezes@reitoria.unesp.br) on 2017-02-13T18:13:27Z (GMT) No. of bitstreams: 1 martins_sc_dr_sjrp.pdf: 4892311 bytes, checksum: ded93001eeec82424d2a73ec7626a3d8 (MD5) / Made available in DSpace on 2017-02-13T18:13:27Z (GMT). No. of bitstreams: 1 martins_sc_dr_sjrp.pdf: 4892311 bytes, checksum: ded93001eeec82424d2a73ec7626a3d8 (MD5) Previous issue date: 2016-12-12 / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / A presente tese tem como objeto de estudo a terminologia da Fauna e da Flora cuja formação envolve o uso do vocabulário das cores. Nosso estudo abrange dois temas gerais: por um lado, a contribuição dos nomes de cores para a variação denominativa em terminologia a partir de uma perspectiva interlinguística; por outro, a descrição de tais variantes em uma obra lexicográfica especializada e multilíngue. A fim de delimitar a nossa pesquisa, concentramo-nos na análise de dois subdomínios da Biologia: a Botânica, especificamente as Angiospermas (monocotiledôneas e eudicotiledôneas), e a Zoologia, exclusivamente os Vertebrados (peixes, anfíbios, répteis, aves e mamíferos). Dessa forma, visamos, em primeiro lugar, ampliar o Dicionário Onomasiológico de Expressões Cromáticas da Fauna e Flora por meio da inserção em sua microestrutura dos correspondentes das palavras-entrada em línguas inglesa e italiana. Seguindo a metodologia proposta em nosso Mestrado (MARTINS, 2013), utilizamos o Corpus Web para a busca e validação dos correspondentes. Em segundo, uma vez verificada a existência de correspondentes para cada entrada, realizamos sua análise estrutural, observando o papel desses itens para a divulgação dos conceitos da Fauna e da Flora em outras culturas. Em terceiro, examinamos as possíveis causas que atuam na formação de tais unidades e que propiciam ou não a escolha do item cor nas línguas em estudo, isto é, português, inglês e italiano. Por último, ainda em relação à obra lexicográfica, propomos formular a metodologia mais adequada para a implementação de uma plataforma on-line que contribuirá para minimizar as divergências terminológicas nessa área, para a normatização de conceitos e para o estabelecimento da comunicação entre culturas. Enfatizamos que o nosso interesse é motivado pelo papel ativo dos nomes de cores na expansão do léxico e a investigação dessa fatia lexical é uma forma de evidenciar o uso dos nomes de cores para a difusão da terminologia da Fauna e da Flora. Finalmente, visamos contribuir qualitativamente, tanto no aspecto teórico (por meio das reflexões apresentadas) quanto no metodológico, para o âmbito das pesquisas que se limitam ao estudo do léxico, bem como para o conjunto das obras lexicográficas produzidas no Brasil. / The present thesis has as its object of study the terminology of Fauna and Flora composed by color names. Hence, it deals with two major themes: on one hand, the contribution of color names to the denominative variation in Terminology from a interlinguistic perspective; on the other, the description of such variants in a multilingual and specialized lexicographical work. In order to delimit our research, we concentrate our analysis on two subdomains of Biology: Botany, specifically the Angiosperms (monocotyledons and dicotyledons), and Zoology, specifically the Vertebrates (fish, amphibians, reptiles, birds and mammals). In this manner, it aims, firstly, to extend the Onomasiological Dictionary of Chromatic Phrases of Fauna and Flora by inserting in the microstructure the equivalents of the entries in English and Italian languages. Following the methodology proposed in a prior research (MARTINS, 2013), we use the Corpus Web to collect and validate the equivalents. Secondly, once checked the existence of equivalents for each entry, it intends to realize their structural analysis, observing the role of these denominative variants to the propagation of Fauna and Flora concepts in other cultures. Thirdly, examining the probable causes that actuate in the formation of such units and that propitiate or not the choice of the color name in the languages in study, i.e., Portuguese, English and Italian. Finally, still regarding the lexicographical work, we propose to formulate an appropriate methodology for the implementation of an on-line platform that will contribute to minimize terminological variances in this area, to concepts normalization, as well as to the communication among cultures. We would like to emphasize that our interest is motivated by the active role of color names in the expansion of the lexicon and the study of this vocabulary demonstrates the importance of color names to the diffusion of Fauna and Flora terminology. To sum up, we expect to contribute qualitatively, as in the theoretical aspect (by means of the reflections presented in the following pages) as in the methodological aspect, to the range of the lexical studies, as well as to the set of lexicographical works produced in Brazil. / FAPESP: 2013/09232-1

Visual tracking is a classical computer vision problem with many important applications in areas such as robotics, surveillance and driver assistance. The task is to follow a target in an image sequence. The target can be any object of interest, for example a human, a car or a football. Humans perform accurate visual tracking with little effort, while it remains a difficult computer vision problem. It imposes major challenges, such as appearance changes, occlusions and background clutter. Visual tracking is thus an open research topic, but significant progress has been made in the last few years. The first part of this thesis explores generic tracking, where nothing is known about the target except for its initial location in the sequence. A specific family of generic trackers that exploit the FFT for faster tracking-by-detection is studied. Among these, the CSK tracker have recently shown obtain competitive performance at extraordinary low computational costs. Three contributions are made to this type of trackers. Firstly, a new method for learning the target appearance is proposed and shown to outperform the original method. Secondly, different color descriptors are investigated for the tracking purpose. Evaluations show that the best descriptor greatly improves the tracking performance. Thirdly, an adaptive dimensionality reduction technique is proposed, which adaptively chooses the most important feature combinations to use. This technique significantly reduces the computational cost of the tracking task. Extensive evaluations show that the proposed tracker outperform state-of-the-art methods in literature, while operating at several times higher frame rate. In the second part of this thesis, the proposed generic tracking method is applied to human tracking in surveillance applications. A causal framework is constructed, that automatically detects and tracks humans in the scene. The system fuses information from generic tracking and state-of-the-art object detection in a Bayesian filtering framework. In addition, the system incorporates the identification and tracking of specific human parts to achieve better robustness and performance. Tracking results are demonstrated on a real-world benchmark sequence.