论文部分内容阅读
为了克服Slepian-Wolf编码不存在统一编码以及采用的无记忆信道模型不符合实际等问题,运用交互式编解码(IED)方式,建立了统一编码,构建了有记忆的有限状态信道(FSC)模型,同时设计了一种适合FSC的用于LDPC解码的置信传播算法。实验结果表明,实现后的IED相比非IED具有更低的误码率(BER),且与理论极限码率更加逼近,实际视频序列的测试结果显示,实现的IED分布式视频编码系统较LDPC算法有2dB左右的率失真性能增益。
In order to overcome the problem that Slepian-Wolf coding does not exist uniform coding and memoryless channel model does not fit into reality, a unified coding scheme is established by using IED. A finite state channel (FSC) model with memory is constructed At the same time, a FSC-based belief propagation algorithm for LDPC decoding is designed. Experimental results show that the IED has a lower bit error rate (BER) than the non-IED, and is closer to the theoretical limit bit rate. The actual video sequence test results show that the IED distributed video coding system achieves better performance than LDPC The algorithm has about 2dB rate distortion performance gain.