Development Of Compression Methodology For 3d Video

Abstract

In the current thesis, a method to encode/decode multiview textures and associated depth maps together in a single bitstream is proposed. A methodology which uses Steganography approach in video compression is designed and implemented to compress the data in which instead of separate bit-streams each for depth and texture, only one bit-stream for texture (containing depth data) is developed. First, depth map is downsampled and then the encoded bits of depth maps have been merged into that of associated textures. Encoded bits of depth maps have been merged into that of associated textures in an innovative way with very minimal loss in PSNR of textures, 0.1 - 0.2 dB for high delay and 0.2 - 0.5 dB for low delay applications with overall bitrate gain of about 17% for high delay and 13% for low delay applications and for natural scenes. The encoding technique ensures backward compatibility (in texture only decode, e.g. MVC stereo high pro le) and correct extraction of downsampled encoded depth maps followed by decoding of downsampled depth maps. Benchmarking results in terms of compression e ciency and quality on HD sequences have been reported to explore further the viability of using this technique for Free-view point Video (FVV) or 3DV use cases. Experiments are also carried out to evaluate subjective quality of the synthesized views.