Research On Rate Control Algorithm In Feedback-Free Distributed Video Coding

As a new video compression coding technology, Distributed Video Coding(DVC) is a hot research topic in recent decades. Compared with traditional video coding, like MPEG-x, H.26 x, it shifts computational complexity from encoder to decoder, and applies channel coding techniques like Turbo codes. Therefore the system has the characteristic of simple encoding, complex decoding, high compression efficiency and good error-resilient performance. Thus, DVC is well-suited for the wireless video surveillance applications, where computing power, storage capacity and power consumption are limited.Although DVC compression system with feedback channel achieves the ideal performance, its constant feedback retransmission greatly increases the decoding delay. Besides, it cannot work when there is no feedback channel. Thus there is a very important meaning for us to study feedback-free DVC system. The DVC compression system without feedback channel decodes only once so that it can reduce the decoding delay and computational complexity, which is very suitable for one-way video applications and real time video scene. However, in feedback-free DVC system, the encoder cannot receive the information of the decoder, so it may not able to decode correctly or it costs too much to decode. How to deal with these situations which are not good to improve system performance and how to estimate the rate correctly are the research emphasis in this paper.This paper firstly introduces the basic principles and system framework of feedback-free Distributed Video Coding, and then focuses on the signal characteristics of bit planes and subbands. And then based on these characteristics, this paper proposes a hybrid coding method and a rate estimation method. The main work is as follows:1. Analyze those bit planes which cost too much to decode in the DVC compression system without feedback channel and use different ways to deal with different bit planes according to the error rate. And then propose a hybrid coding method of entropy code and Turbo code. In addition, when the quantization is low, the quantitative symbols of AC subband coefficients are mostly zero and sequential. This paper uses run length coding compression to improve the coding efficiency effectively. Simulation results demonstrate that the hybrid code method can improve the system rate distortion(RD) performance of different video sequences, especially for those intense sequences. For example, the proposed method can achieve significant RD improvement by up to 1.5 d B.2. Do a deep research on feedback-free DVC compression system of big GOP and use the above algorithm to improve its RD performance. And then propose a rate allocation algorithm based on the importance of bit plane so that we can allocate the rate more correct and improve the performance of the feedback-free DVC system.