Abstract
In the modern video compression techniques, for achieving high compressive rate, the encoder would exploit all spatial (intra coding) and temporal (inter coding) correlation on every frame. The complexity to decode a frame is much less than to encode one. However, in some applications we would adopt some devices which lack memory and computing ability. Distributed Video Coding (DVC) is helpful in this situation. Based on the theoretic results of Distributed Source Coding (DSC), DVC focuses on a lower complexity encoder and a powerful decoder. This feature is very similar to the concept of Compressive Sensing (CS). Therefore, in this thesis, we discuss how to employ CS into DVC. We proposed a wavelet-based distributed video coding method with compressive sensing. To find a sparse representation for video frames, we use the difference on the wavelet domain. We also adopt skip block, quantization, and entropy coding to reduce the bit-rate. We use simple side information for helping decoder. The main contributions of our works are further reduction of complexity and coding time of the encoder. Compared with other DVC algorithms, the experimental results shows proposed encoder can save more energy. This makes the proposed encoder is able to be implemented in the devices with limited power or computational ability.