VIEW SYNTHESIS OF THREE DIMENSIONAL VIDEOS USING VIDEO PLUS DEPTH Loghman, Maziar This thesis investigates the use of segmentation in view synthesis for three- dimensional video. View synthesis is the process of generating novel views of a scene, using set of views as the reference. Recently, several techniques have been suggested that use the depth maps for rendering virtual views. However, inaccuracy in depth maps causes annoying visual artifacts in depth-based view synthesis. This thesis presents an e cient depth image-based rendering technique based on segmentation using multi-level thresholding. In the proposed algorithm, rst all the images are segmented according to the depth and the pixels belonging to di erent objects are warped and blended independently. Based on multi-level thresholding, an algorithm for nding the ghost contour pixels is provided which simpli es the computations. A novel method for inpainting for disocclusions is introduced which uses the segmented images to nd the associated background boundary pixels. The experimental results show that the proposed algorithm improves the PSNR of the synthesized views up to 0:68 dB for the multi-view video test sequences and eliminates the annoying artifacts. M.S. in Electrical Engineering, May 2012 Kim, Joohee 2012-07-02 2012-05 Thesis application/pdf islandora:9063 http://hdl.handle.net/10560/2805 ECE / Electrical and Computer Engineering Illinois Institute of Technology en In Copyright http://rightsstatements.org/page/InC/1.0/ Restricted Access