DESARROLLO DE M ETODOS Y ALGORITMOS DE RECONSTRUCCI ON DE OBJETOS TRIDIMENSIONALES USANDO VIDEO A COLOR CON IMPLEMENTACI ON EN HARDWARE ESPEC IFICO

Abstract

Actually, several applications in signal and video processing have been proposed. Feature extraction are important in those applications. One of the emerging areas in video processing is the generation of content with tridimensional perception. In the rst part of this work, a brief review of di erent techniques used in the generation of tridimensional perception content is presented and di erent algorithms in the same topic are analized.We present di erent approaches used in the visualization of tridimensional videos and its features. In the second part, we present di erent algorithms to the disparity map computation, which will be used as important information in the 3D video sequences generation, additionally, some algorithms based on the quality results and computational cost are analized. In the third part of the work, we introduce the proposed algortihm which is based in the wavelet based on atomic functions with the multi level strategy and the anaglyph synthesis as a visualization method. In the fourth part, we can nd the images and video sequences in synthetic and real form and the objective assessments. We include the resulting images and videos in order to perform a subjective evaluation, taking into account the ghosting, loss of color and other artifacts. Finally, the implementation of di erent algorithms are presented, including the proposed algorithm in speci c hardware.