Study On Unequal Error Protection Schemes Based On Pictures And Videos

Nowadays, both multimedia communication and wireless communication are developing largely. They are merging to one system rapidly. Because both of them have their own systems, the conflict between them will make influence to the whole system while merging. Multimedia information, such as picture and video data, is compressed highly, so the relativity among the data is small. However, wireless channels are very complex. They have high bit-error-rate and limited bandwidth. Therefore, there are many problems during the transmission of multimedia information based on wireless channels.To overcome the merge problem, the combination of the source coding and channel coding based on wireless environment should be considered. Therefore, the quality of reconstructed video could be improved under the limited system resources. Unequal Error Protection (UEP) is one of the promising techniques to address this issue.In this work, the related techniques of UEP are researched. Meanwhile, several new UEP schemes based on wireless channels are proposed. The main jobs and innovations are as follows:1. An UEP scheme based on SPIHT-Coded image transmission over MIMO system is proposed.SPIHT-Coded image information is divided into different substreams according to their different importance. Each substream is transmitted differentially over MIMO system based on MIMO Adaptive Channel Assignment Policy with changing-length pilot sequences (P_ACAP). Simulation results show that the scheme with P_ACAP can achieve UEP for image transmission automatically. Meanwhile, by selecting pilot sequences with changing length, the better quality of reconstructed image could be achieved even when the wireless channel condition is not good.2. An UEP scheme based on H.264/AVC Data Partition (DP) image transmission over MIMO-OFDM system is proposed. The combination of MIMO and OFDM improves significantly capacity of the system and overcome the frequency selective effects of broadband wireless channel. UEP is provided by channel package policy for Data Partitioning (DP) defined in H.264 standard. Simulation results show that the scheme can obviously improve the error resilience compared to EEP, especially when the channel is in bad condition. So the scheme is suitable to fluctuant channel with large range.3. An UEP scheme based on Scalable Video Coding (SVC) extension of the H.264/AVC Standard and LDPC channel coding is proposed.In the proposed scheme, SVC coded video information is divided into different substreams in terms of time and qulity scalability. These substreams are channel-coded using Low Density Parity Check (LDPC) Codes as forward error correction (FEC) coding to provide UEP. In which, an Optimal Bit Allocation Algorithm (OBAA) is proposed to get optimal LDPC channel code rates allocation. In order to implement OBAA, we develop a LDPC probability model of decoding failure based on AWGN channels. Due to only LDPC codes are used to provide the unequal error protection, the efficiency of channel transmission is improved.4. An enhanced UEP scheme with the changing quantization parameters based on the third UEP scheme is proposed.Motivated by optimally allocating the bits to achieve resilience against channel induced corruptions, a Joint Optimal Bit Allocation Algorithm (JOBAA) is proposed. Simulation results show that the proposed UEP scheme with changing QP can achieve better quality of reconstructed video than UEP scheme with fixed QP under the same constraints of transmission bit rate budget.