An Improved Decoding Algorithm for Distributed Video Coding

Abstract

An improved decoding algorithm based on joint bitplane decoding for distributed video coding is proposed. The new algorithm uses the decoded bitplane results of quantization coefficients as prior knowledge, and decodes the other bit-planes to avoid correlation weakening between sources when decoding bit-planes independently. It makes full use of correlation between Wyner-Ziv frames and key frames; therefore the rate-distortion performance of distributed video coding can be enhanced without increasing decoding complexity. Simulations indicate that, compared with the independent bit-planes decoding system, the proposed algorithm can achieve 13% to 16% bit-rate saving under the same condition, and performance improves more apparently with the increase of quantization levels.

An LDPC-based improved decoding scheme for distributed video codec

“…Each point in the residual frame belongs to one confidence area. Simultaneously, with the estimated residual frame R, we can obtain the CNM parameterα , which will be sent along with the SI to the LDPCA decoder to calculate the LLR values for all bits in every bitplane with the methods in [10]. Then, according to the corresponding points in R of the pixels in SI belonging to different confidence areas, the LLR values will be revised with our proposed scheme.…”

“…In this section, we will describe in detail the specific realization steps of our improved decoding scheme based on that in [10]. Compared with the decoding scheme in [10], we propose to utilize the video inherent character to divide the points in the residual frame R into different confidence areas.…”

“…Compared with the decoding scheme in [10], we propose to utilize the video inherent character to divide the points in the residual frame R into different confidence areas. According to the one-to-one correspondence between the points in R and the pixels in SI, the pixels corresponding to the different confidence areas will have different LLR initialization methods, and the binary representations of one pixel located in different bitplanes also have different LLR initialization methods.…”

“…In these methods, decoding each bitplane independently will weaken the correlation between different bitplanes, and the system performance will be affected also. In [10], the authors propose an improved decoding scheme based on traditional LDPC WZ codec. It makes use of the decoded bitplanes as the prior information to help decoding the succeeding bitplanes to avoid correlation weakening and to enhance the RD performance of DVC without increasing decoding complexity.…”

“…In this paper, we propose an improved decoding scheme for DVC based on the decoding scheme in [10]. Considering the video inherent character, we set different confidence areas in the residual frame to partition the pixels in side information according to the movement of objects in the video sequence.…”