Several Of Distributed Video Coding Technology Research

The traditional video coding scheme usually uses prediction (inter-frame and intra-frame), transformation, coding at the encoder to realize the video compression. However, these operations would get lots of calculations at the encoder and leads to a high demand for the encoder. So the processing method is not applicable for the encoder that just can provide limited resources, such as wireless sensor network with light mobile terminals. According to the deficiency of the traditional video coding scheme, the Distributed Video Coding (DVC) scheme is a good solution by reducing the coding operation complexity, which transfers the complexity of the calculations to the decoder.Some works based on distributed video system have been done as follows:1. Introduces the distributed video coding scheme and several key technologies. The adaptive coding scheme based on Turbo code distributed video system is realized, this scheme realizes the dynamic adjustment of quantitative parameters through calculating the motion complexity of non-key frame at the encoder. Results show that the reconstruction image of adaptive scheme has a better visual effect, the performance is improved by0.25-0.5dB than the non-adaptive scheme when the code rate is200kbps.2. Investigates several kinds of side information generation schemes. Analyzes and compares the average interpolation and motion compensation scheme to obtain side information. A key research on the bit plane updating algorithm is made. Results show that the bit plane updating algorithm can not only eliminate the side information fuzzy caused by the average interpolation but also have a good elimination on the block effect caused by motion compensation, and the performance is improved by0.1dB than the original side information when the code rate is300kbps.3. Investigates code rate algorithm of the non-key frame without feedback channel. This algorithm uses the mapping relationship between bit plane error probability and check number to allocate code rate for each bit plane in advance and complete decoding. Results show that this method can distribute accurately required code rate for each bit plane and the rate distortion performance just have a little change.