Orientador: Antonio Augusto Fasolo Quevedo / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-14T05:51:27Z (GMT). No. of bitstreams: 1 Nunes_LigiaChristinaBorsatoGuimaraes_D.pdf: 5683432 bytes, checksum: b411bd5190e5d51df43c3c3b6a743ed8 (MD5) Previous issue date: 2008 / Resumo: Instrumentos de avaliação sistematizados em paralisia cerebral (PC) são fundamentais em avaliação, tratamento e pesquisa. O questionário de qualidade de vida pediátrico PedsQL tem por objetivo verificar o impacto de enfermidades na qualidade de vida. O PedsQL para PC aborda questões específicas relacionadas à condição. O Gross Motor Function Measure (GMFM) é um instrumento que visa quantificar a função motora grossa e é amplamente utilizado no mundo, inclusive no Brasil. A alteração dos padrões de marcha é comum em PCs. O padrão-ouro para avaliar marcha é a análise tridimensional; entretanto, é improvável torná-la rotina na prática clínica. Tendo em vista estas dificuldades, foi desenvolvida a escala visual de marcha de Edimburgo. Embora a importância desses instrumentos seja indiscutível e alguns estarem sendo utilizados no Brasil, não há tradução oficial, validação, nem adaptação cultural deles. Também faltam estudos de fidedignidade e da influência do treinamento na utilização de escalas de marcha visuais. Este trabalho tem como objetivos traduzir, validar, adaptar culturalmente os três instrumentos, verificar a correlação da escala de Edimburgo com a análise tridimensional e verificar a influência do treinamento na fidedignidade intra e inter-observadores e na concordância com o padrão-ouro. 64 pais e 36 crianças com paralisia cerebral participaram do estudo de validação do PedsQL PC brasileiro; 10 crianças foram filmadas e avaliadas através da versão brasileira do GMFM e duas crianças inglesas submeteram-se à avaliação de marcha em laboratório e através de vídeo convencional. Oito fisioterapeutas avaliaram os vídeos e conferiram escores às avaliações motoras de GMFM e 22 fisioterapeutas avaliaram os vídeos de marcha, segundo a Escala de Edimburgo. Verificou-se que as versões Brasileiras do PedsQL (primeira versão e modificada) apresentaram boa consistência interna, sensitividade e as duas versões, embora demonstrem resultados estatisticamente diferentes, apresentam uma correlação quase perfeita (0,998). É necessária uma investigação com maior número de crianças de variadas classificações funcionais. Quanto ao GMFM, a fidedignidade inter-avaliadores foi substancial (medianas de kappa= 0,69 e 0,71) e a fidedignidade intra-avaliadores foi quase perfeita (mediana de kappa =0,976), o que torna a versão da escala relativamente confiável, mesmo sem o auxílio do manual. Entretanto, os resultados de fidedignidade observados são menores que em trabalhos que utilizaram o manual, ainda não traduzido para o Brasil. Foi desenvolvido um pacote de treinamento para a Escala Visual de Marcha de Edimburgo em CD, com versões em inglês e português, de fácil manuseio e boa interatividade, embora em linguagem simples. A versão brasileira da escala de Edimburgo obteve índices substanciais de fidedignidade intra-avaliador e modestos entre avaliadores. A concordância com a avaliação instrumentada geral foi menor que 50%, para os grupos treinado e não treinado. O treinamento através do CD melhorou a fidedignidade intra e inter-avaliador e a concordância com a avaliação instrumentada. Estudos com maior número de pacientes são necessários para verificação das fidedignidades da Escala de Edimburgo, bem como a avaliação da versão em inglês do pacote de treinamento e a verificação de outras formas de treinamento. A disponibilização dos três instrumentos de avaliação em PC contribui para estudos futuros no Brasil. / Abstract: Systematized evaluation tools for children with Cerebral Palsy is very important in evaluation, treatment and research. The pediatrics Quality of Life Questionnaire PedsQL has the purpose of verify the diseases impact in quality of life. Its Cerebral Palsy version encompasses more specific questions regarding motor damage due to CP. The Gross Motor Function Measure GMFM) is a tool that aims to quantify Gross Motor Function. It is widely used around the world, including Brasil. Changes on the gait patterns are common in CP. The golden standard to evaluate gait is three-dimensional gait analysis, although it is not possible to perform this evaluation as a routine in clinics. The Edimburgh Visual Gait Scale was developed based on video. Although the importance of these tools is incontestable and some are being used in Brazil, there is no official translation, nor validation nor cultural adaptation of them. There is still a lack of intra and inter rater reliability studies of the three tools and the training effects on observational gait scales. The purposes of this work are to translate, validate and culturally adapt the above mentioned tools, to verify the correlation between the Edinburgh gait scale and 3D gait analysis as well as to study the training influence on inter and intrarater reliability and on agreement with golden standard. 64 parents and 36 children with CP took part of the PedsQL study; 10 children were filmed and evaluated trough the Brazilian version of GMFM and two English children undergone to laboratorial and video gait evaluation. Eight Physical Therapists evaluated the GMFM videos and 22 Physical Therapists evaluated the gait videos, according to the Edinburgh gait scale. The Brazilian versions of PedsQL (first and modified) show good internal consistency, sensitivity and, although the two brazilian versions returned different scores, they show an almost perfect correlation. Further investigation with more children of all functional classifications is needed. Regarding GMFM, the global inter rater reliability was substantial and the intrarater reliability was almost perfect, which makes the Brazilian scale feasible even without the manual. However, the reliability found on this work was smaller than works that used the GMFM manual, still not translated to Brazil. It was developed a training package to the Edinburgh Visual Gait Scale in a simple, easy handling and interactive CD, with versions in English and Portuguese. The Brazilian version of the Edinburgh Visual Gait Scale had substantial intra rater reliability and modest inter rater reliability. The mean agreement with 3D gait analysis was less than 50% for trained and untrained groups. Training with CD has improved intra and inter rater reliability as well as agreement with 3D analysis. Reliability studies with more volunteers as well as studies with the English version training package and other training approaches are needed. The three evaluation tools available now in Portuguese will contribute with more studies in Cerebral Palsy in Brazil. / Doutorado / Engenharia Biomedica / Doutor em Engenharia Elétrica

Submitted by Marlene Santos (marlene.bc.ufg@gmail.com) on 2016-08-10T19:31:03Z No. of bitstreams: 2 Dissertação - Weder Cabral Mendes - 2016.pdf: 1214460 bytes, checksum: 14588573f8f81fe4836a9945adacf37d (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2016-08-15T13:25:32Z (GMT) No. of bitstreams: 2 Dissertação - Weder Cabral Mendes - 2016.pdf: 1214460 bytes, checksum: 14588573f8f81fe4836a9945adacf37d (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) / Made available in DSpace on 2016-08-15T13:25:32Z (GMT). No. of bitstreams: 2 Dissertação - Weder Cabral Mendes - 2016.pdf: 1214460 bytes, checksum: 14588573f8f81fe4836a9945adacf37d (MD5) license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) Previous issue date: 2016-03-31 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / Gait recognition via computer vision attracted increasing interest for its noninvasive characteristic and mainly for your advantage of recognizing people at distance. Recognition is performing extracting features included in gait, this features are extracted from images sequence of people walking. The main challenges of gait recognition is to extract characteristics with unique information for each person, in additional, the use of accessories and clothes difficult the feature extraction process. This paper proposes a gait recognition method using information of people’s contours transformed in domain frequence by Discrete Fourier Transform. A lot of data are generated from the contours, thereby, three different techniques for dimensionality reduction CDA (Class Discrimination Ability), PCA (Principal Component Analysis) and PLS (Partial Least Squares) are employed to reduce the dimensionality of data and generate characteristics that are relevant to the recongnition system. Two classifiers, KNN (K-Nearest Neighbor) and LDA (Linear Discriminant Analysis) classify the characteristics that are returned by the dimensionality reduction methods. The accuracy are achieved by the combination of the dimensionality reduction methods and classifiers, the highest accuracy was 92:67%, which was achieved with the combination between the LDA and PCA (LDAPCA). Therefore, the results show that the information contained in the contours of silhouette are discriminant to recognize people by their gait. / O reconhecimento de pessoas através da marcha humana via visão computacional tem ganhado destaque por ser uma técnica biométrica não invasiva e principalmente por sua vantagem de reconhecer pessoas à distância. O reconhecimento é realizando extraindo características contidas na marcha de cada pessoa, essas características são extraídas de sequências de imagens da pessoa caminhando. Os principais desafios dessa técnica biométrica está em extrair as características com informações que consigam diferenciar uma pessoa da outra, além disso, o uso de acessórios e vestimentas dificultam o processo de extração de características. Este trabalho propõe um método de reconhecimento baseado na marcha humana utilizando informações dos contornos das pessoas transformados para o domínio da frequência por meio da Transformada Discreta de Fourier. Como são geradas muitos dados a partir dos contornos, três técnicas diferentes de redução de dimensionalidade CDA (Class Discrimination Ability), PCA (Principal Component Analysis) e PLS (Partial Least Squares) são empregadas para reduzir a quantidade de dados e gerar características que sejam relevantes para o sistema de reconhecimento. Dois classificadores, KNN (K-Nearest Neighbor) e LDA (Linear Discriminant Analysis) classificam as características retornadas pelos métodos de redução de dimensionalidade. As taxas de acurácia são obtidas pelos resultados gerados entre a combinação dos métodos de redução de dimensionalidade e os classificadores, a maior taxa de acurácia foi de 92;67%, a qual foi alcançada com a combinação entre o LDA e PCA (LDAPCA). Dessa forma, conclui-se que as informações contidas no contorno da silhueta no domínio da frequência são discriminantes para reconhecer pessoas através da marcha.

No description available.

The work done in this master's thesis, presents a new system for the recognition of human actions from a video sequence. The system uses, as input, a video sequence taken by a static camera. A binary segmentation method of the the video sequence is first achieved, by a learning algorithm, in order to detect and extract the different people from the background. To recognize an action, the system then exploits a set of prototypes generated from an MDS-based dimensionality reduction technique, from two different points of view in the video sequence. This dimensionality reduction technique, according to two different viewpoints, allows us to model each human action of the training base with a set of prototypes (supposed to be similar for each class) represented in a low dimensional non-linear space. The prototypes, extracted according to the two viewpoints, are fed to a $K$-NN classifier which allows us to identify the human action that takes place in the video sequence. The experiments of our model conducted on the Weizmann dataset of human actions provide interesting results compared to the other state-of-the art (and often more complicated) methods. These experiments show first the sensitivity of our model for each viewpoint and its effectiveness to recognize the different actions, with a variable but satisfactory recognition rate and also the results obtained by the fusion of these two points of view, which allows us to achieve a high performance recognition rate. / Le travail mené dans le cadre de ce projet de maîtrise vise à présenter un nouveau système de reconnaissance d’actions humaines à partir d'une séquence d'images vidéo. Le système utilise en entrée une séquence vidéo prise par une caméra statique. Une méthode de segmentation binaire est d'abord effectuée, grâce à un algorithme d’apprentissage, afin de détecter les différentes personnes de l'arrière-plan. Afin de reconnaitre une action, le système exploite ensuite un ensemble de prototypes générés, par une technique de réduction de dimensionnalité MDS, à partir de deux points de vue différents dans la séquence d'images. Cette étape de réduction de dimensionnalité, selon deux points de vue différents, permet de modéliser chaque action de la base d'apprentissage par un ensemble de prototypes (censé être relativement similaire pour chaque classe) représentés dans un espace de faible dimension non linéaire. Les prototypes extraits selon les deux points de vue sont amenés à un classifieur K-ppv qui permet de reconnaitre l'action qui se déroule dans la séquence vidéo. Les expérimentations de ce système sur la base d’actions humaines de Wiezmann procurent des résultats assez intéressants comparés à d’autres méthodes plus complexes. Ces expériences montrent d'une part, la sensibilité du système pour chaque point de vue et son efficacité à reconnaitre les différentes actions, avec un taux de reconnaissance variable mais satisfaisant, ainsi que les résultats obtenus par la fusion de ces deux points de vue, qui permet l'obtention de taux de reconnaissance très performant.

The work done in this master's thesis, presents a new system for the recognition of human actions from a video sequence. The system uses, as input, a video sequence taken by a static camera. A binary segmentation method of the the video sequence is first achieved, by a learning algorithm, in order to detect and extract the different people from the background. To recognize an action, the system then exploits a set of prototypes generated from an MDS-based dimensionality reduction technique, from two different points of view in the video sequence. This dimensionality reduction technique, according to two different viewpoints, allows us to model each human action of the training base with a set of prototypes (supposed to be similar for each class) represented in a low dimensional non-linear space. The prototypes, extracted according to the two viewpoints, are fed to a $K$-NN classifier which allows us to identify the human action that takes place in the video sequence. The experiments of our model conducted on the Weizmann dataset of human actions provide interesting results compared to the other state-of-the art (and often more complicated) methods. These experiments show first the sensitivity of our model for each viewpoint and its effectiveness to recognize the different actions, with a variable but satisfactory recognition rate and also the results obtained by the fusion of these two points of view, which allows us to achieve a high performance recognition rate. / Le travail mené dans le cadre de ce projet de maîtrise vise à présenter un nouveau système de reconnaissance d’actions humaines à partir d'une séquence d'images vidéo. Le système utilise en entrée une séquence vidéo prise par une caméra statique. Une méthode de segmentation binaire est d'abord effectuée, grâce à un algorithme d’apprentissage, afin de détecter les différentes personnes de l'arrière-plan. Afin de reconnaitre une action, le système exploite ensuite un ensemble de prototypes générés, par une technique de réduction de dimensionnalité MDS, à partir de deux points de vue différents dans la séquence d'images. Cette étape de réduction de dimensionnalité, selon deux points de vue différents, permet de modéliser chaque action de la base d'apprentissage par un ensemble de prototypes (censé être relativement similaire pour chaque classe) représentés dans un espace de faible dimension non linéaire. Les prototypes extraits selon les deux points de vue sont amenés à un classifieur K-ppv qui permet de reconnaitre l'action qui se déroule dans la séquence vidéo. Les expérimentations de ce système sur la base d’actions humaines de Wiezmann procurent des résultats assez intéressants comparés à d’autres méthodes plus complexes. Ces expériences montrent d'une part, la sensibilité du système pour chaque point de vue et son efficacité à reconnaitre les différentes actions, avec un taux de reconnaissance variable mais satisfaisant, ainsi que les résultats obtenus par la fusion de ces deux points de vue, qui permet l'obtention de taux de reconnaissance très performant.