In this dissertation, novel variational optical-flow and active-contour methods are investigated to address challenging problems in computer vision and shape-based imaging. Starting from traditional applications of these methods in computer vision, such as object segmentation, tracking, and detection, this research subsequently applies similar active contour techniques to the realm of shape-based imaging, which is an image reconstruction technique estimating object shapes directly from physical wave measurements. In particular, the first and second part of this thesis deal with the following two physically inspired computer vision applications. Optical Flow for Vision-Based Flame Detection: Fire motion is estimated using optimal mass transport optical flow, whose motion model is inspired by the physical law of mass conservation, a governing equation for fire dynamics. The estimated motion fields are used to first detect candidate regions characterized by high motion activity, which are then tracked over time using active contours. To classify candidate regions, a neural net is trained on a set of novel motion features, which are extracted from optical flow fields of candidate regions. Coupled Photo-Geometric Object Features: Active contour models for segmentation in thermal videos are presented, which generalize the well-known Mumford-Shah functional. The diffusive nature of heat processes in thermal imagery motivates the use of Mumford-Shah-type smooth approximations for the image radiance. Mumford-Shah's isotropic smoothness constraint is generalized to anisotropic diffusion in this dissertation, where the image gradient is decomposed into components parallel and perpendicular to level set curves describing the object's boundary contour. In a limiting case, this anisotropic Mumford-Shah segmentation energy yields a one-dimensional ``photo-geometric'' representation of an object which is invariant to translation, rotation and scale. These properties allow the photo-geometric object representation to be efficiently used as a radiance feature; a recognition-segmentation active contour energy, whose shape and radiance follow a training model obtained by principal component analysis of a training set's shape and radiance features, is finally applied to tracking problems in thermal imagery. The third part of this thesis investigates a physics-driven active contour approach for shape-based imaging. Adjoint Active Contours for Shape-Based Imaging: The goal of this research is to estimate both location and shape of buried objects from surface measurements of waves scattered from the object. These objects' shapes are described by active contours: A misfit energy quantifying the discrepancy between measured and simulated wave amplitudes is minimized with respect to object shape using the adjoint state method. The minimizing active contour evolution requires numerical forward scattering solutions, which are obtained by way of the method of fundamental solutions, a meshfree collocation method. In combination with active contours being implemented as level sets, one obtains a completely meshfree algorithm; a considerable advantage over previous work in this field. With future applications in medical and geophysical imaging in mind, the method is formulated for acoustic and elastodynamic wave processes in the frequency domain.

No description available.

Zugl.: Karlsruhe, Univ., Diss., 2008 / Hergestellt on demand. - Auch im Internet unter der Adresse http://uvka.ubka.uni-karlsruhe.de/shop/isbn/978-3-86644-260-3 verfügbar

Thesis (Ph. D.)--Rutgers University, 2007. / "Graduate Program in Computer Science." Includes bibliographical references (p. 63-66).

Zugl.: Bremen, Univ., Diss., 2009

Zugl.: Aachen, Techn. Hochsch., Diss., 2006.

Essen, Universiẗat, Diss., 2005--Duisburg.

Submitted by AMAURI GABRIEL DE JESUS JUNIOR null (engcomp.amauri@gmail.com) on 2016-04-26T22:38:30Z No. of bitstreams: 1 dissertacao.pdf: 2768291 bytes, checksum: 95f3a5b55640c18b626bd655f9d741ed (MD5) / Rejected by Felipe Augusto Arakaki (arakaki@reitoria.unesp.br), reason: Solicitamos que realize uma nova submissão seguindo as orientações abaixo: O arquivo submetido está sem a ficha catalográfica. A versão submetida por você é considerada a versão final da dissertação/tese, portanto não poderá ocorrer qualquer alteração em seu conteúdo após a aprovação. Corrija esta informação e realize uma nova submissão contendo o arquivo correto. Agradecemos a compreensão. on 2016-04-28T20:45:50Z (GMT) / Submitted by AMAURI GABRIEL DE JESUS JUNIOR null (engcomp.amauri@gmail.com) on 2016-04-29T13:43:21Z No. of bitstreams: 1 dissertacao.pdf: 2771977 bytes, checksum: 4fbc89c116e007f08725eca70f97172a (MD5) / Rejected by Felipe Augusto Arakaki (arakaki@reitoria.unesp.br), reason: Solicitamos que realize uma nova submissão seguindo as orientações abaixo: A versão final da dissertação deve ser submetida no formato PDF (Portable Document Format) e o arquivo não deve estar protegido. Por favor, corrija o arquivo PDF e realize uma nova submissão com o arquivo desprotegido. Agradecemos a compreensão. on 2016-04-29T17:34:45Z (GMT) / Submitted by AMAURI GABRIEL DE JESUS JUNIOR null (engcomp.amauri@gmail.com) on 2016-05-05T22:21:23Z No. of bitstreams: 1 dissertacao.pdf: 2771977 bytes, checksum: 4fbc89c116e007f08725eca70f97172a (MD5) / Approved for entry into archive by Juliano Benedito Ferreira (julianoferreira@reitoria.unesp.br) on 2016-05-09T20:13:36Z (GMT) No. of bitstreams: 1 jesusjunior_ag_me_prud.pdf: 2771977 bytes, checksum: 4fbc89c116e007f08725eca70f97172a (MD5) / Made available in DSpace on 2016-05-09T20:13:36Z (GMT). No. of bitstreams: 1 jesusjunior_ag_me_prud.pdf: 2771977 bytes, checksum: 4fbc89c116e007f08725eca70f97172a (MD5) Previous issue date: 2016-03-30 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Esse trabalho apresenta um estudo sobre técnicas de reinicialização para métodos level-set com aplicação em escoamentos de fluidos. O transporte da interface é feito com um método WENO de quinta ordem para as derivadas espaciais, e um Runge-Kutta de terceira ordem para a derivada temporal. Utilizamos malha uniforme, no contexto de diferenças finitas, e nos limitamos a estudar casos bidimensionais. Após o estudo da advecção da interface, apresentamos duas técnicas de reinicialização para restaurar a propriedade de função distância da função level-set sem, que haja perda de massa. A primeira das técnicas consiste na resolução de uma equação diferencial parcial, e a outra é puramente geométrica. Utilizando as Equações de Navier-Stokes, estudamos duas técnicas para o cálculo da força devido à tensão superficial: CSF e SSF. Apresentamos testes numéricos com velocidade prescrita para verificar as duas técnicas de reinicialização, e resolvemos problemas de escoamentos bifásicos com tensão superficial para verificar como as duas técnicas de reinicialização se comportam nesse tipo de problema. / This work presents a study about redistancing schemes for level-set methods with applications in fluid flow. The transport of the interface is performed with a fifth order WENO scheme combined with a third order accurate Runge-Kutta method. After the advection of the interface, we analise two redistancing methods to restore the distance function property of the level-set function. The first method consists in the solution of a partial differential equation, while the other is fully geometric. Using the Navier-Stokes equations, we study two different techniques for the calculation of the surface tension force: CSF and SSF. In order to numerically verify the two redistancing schemes, we present tests with prescribed velocity field. Finally, we apply the redistancing techniques for solving biphase fluid flows in the presence of the surface tension.

abstract: This dissertation describes a process for interface capturing via an arbitrary-order, nearly quadrature free, discontinuous Galerkin (DG) scheme for the conservative level set method (Olsson et al., 2005, 2008). The DG numerical method is utilized to solve both advection and reinitialization, and executed on a refined level set grid (Herrmann, 2008) for effective use of processing power. Computation is executed in parallel utilizing both CPU and GPU architectures to make the method feasible at high order. Finally, a sparse data structure is implemented to take full advantage of parallelism on the GPU, where performance relies on well-managed memory operations. With solution variables projected into a kth order polynomial basis, a k+1 order convergence rate is found for both advection and reinitialization tests using the method of manufactured solutions. Other standard test cases, such as Zalesak's disk and deformation of columns and spheres in periodic vortices are also performed, showing several orders of magnitude improvement over traditional WENO level set methods. These tests also show the impact of reinitialization, which often increases shape and volume errors as a result of level set scalar trapping by normal vectors calculated from the local level set field. Accelerating advection via GPU hardware is found to provide a 30x speedup factor comparing a 2.0GHz Intel Xeon E5-2620 CPU in serial vs. a Nvidia Tesla K20 GPU, with speedup factors increasing with polynomial degree until shared memory is filled. A similar algorithm is implemented for reinitialization, which relies on heavier use of shared and global memory and as a result fills them more quickly and produces smaller speedups of 18x. / Dissertation/Thesis / Doctoral Dissertation Aerospace Engineering 2015

Ingeniero Civil Matemático / El objetivo central de esta memoria es estudiar un problema inverso geométrico en mecánica de fluidos y realizar un procedimiento de reconstrucción numérica que permita recuperar distintos cuerpos rígidos inmersos en un fluido viscoso, siendo de especial interés el caso de cuerpos no convexos. Para llevar a cabo esta reconstrucción numérica se utiliza el llamado método level set. El método level set fue introducido por S. Osher y J. A. Sethian como un método simple y versátil para calcular y analizar el movimiento de una interface Γ bajo un campo de velocidades V, en dos y tres dimensiones, donde Γ es la frontera de una región Ω. Por otra parte en los problemas inversos geométricos, es decir, problemas donde la incógnita es una forma geométrica, el enfoque estándar para la solución de estos consiste en parametrizar la forma geométrica y aplicar métodos de regularización directamente a la parametrización. Este enfoque sufre de la limitación que para obtener aproximaciones convergentes se tiene que tener un conocimiento a priori de la estructura y topología de la forma geométrica buscada. Por esta razón, recientemente se han considerado enfoques alternativos para la solución de problemas de reconstrucción de formas geométricas, entre ellos el método level set, el cual fue utilizado inicialmente en el procesamiento de imágenes digitales. La presente memoria esta estructurada de la siguiente manera. En el Capítulo 1 se realiza una introducción al trabajo realizado. En el Capítulo 2 se hace una introducción a los problemas inversos, se define el problema inverso geométrico de detección de obstáculos dentro de un fluido y se muestran los resultados de identificabilidad y estabilidad para este problema. En el Capítulo 3 se estudia el método de los elementos finitos y la resolución del problema de Stokes usando dicho método, en donde se muestran el algoritmo de Uzawa y el algoritmo numérico para Stokes usado en esta memoria. En el Capítulo 4 se presenta el método de diferenciación con respecto al dominio, el cual resulta fundamental para posteriormente realizar el cálculo de la primera derivada local del funcional de costo asociado al problema inverso geométrico en estudio. En el Capítulo 5 se presenta el método level set, estudiando los movimientos por curvatura media y en dirección normal, la ecuación de reinicialización y la extensión del campo de velocidades. Además, se muestra su aplicación a la optimización de formas y se utiliza la diferenciación con respecto al dominio para deducir la expansión de primer orden del funcional de costo asociado al problema. En el Capítulo 6 se muestran los principales resultados numéricos obtenidos al usar el método level set, recuperando diferentes obstáculos (incluyendo algunos de geometría no convexa), para lo cual se ha utilizado el programa FreeFem. Finalmente, se presentan las principales conclusiones obtenidas de este trabajo de título.