What Affects Decoding Complexity of Distributed Video Codec Based on Turbo Code

Abstract

The distributed video coding (DVC) is a new coming video compression technology that utilizes a different computing complexity mode from the traditional video codec. As a new paradigm there are some fundamental and hard questions in DVC that are not sufficiently addressed, for example the complexity balance between DVC encoder and decoder. This will be an important factor for some promising real time applications. In this context, this paper studies the factors which affect the decoding complexities of turbo code based DVC, especially focuses on quantization, side information and turbo code bit rates. Analyzing and simulation results show that precise side information will decrease the turbo decoding complexity, large reductions in computations can be traded against relatively small increases in bit rate, and uniform quantizer possesses the advantage than the non-uniform quantizer in computation load at a expense of a little decrease in rate distortion performance.