A lowcomplexity encoding method and encoder architecture are proposed for the low-density parity-check (LDPC) codes in space applications recommended by Consultant Committee for Space Data Systems. This method uses t...
详细信息
A lowcomplexity encoding method and encoder architecture are proposed for the low-density parity-check (LDPC) codes in space applications recommended by Consultant Committee for Space Data Systems. This method uses the generator matrix partition and decomposition to obtain a much smaller dense core matrix. Thus, only one quarter of dense block matrices need to be implemented using the traditional circulant encoding structure. Therefore, the encoder can be implemented with lower computation and storage complexity.
Existing encoder rate control (ERC) solutions have two technical limitations that prevent them from being widely used in real-world applications. One is that encoder side information (ESI) is required to be generated ...
详细信息
Existing encoder rate control (ERC) solutions have two technical limitations that prevent them from being widely used in real-world applications. One is that encoder side information (ESI) is required to be generated which increases the complexity at the encoder. The other is that rate estimation is performed at bit plane level which incurs computation overheads and latency when many bit planes exist. To achieve a low-complexity encoder, we propose a new ERC solution that combines an efficient encoder block mode decision (EBMD) for the distributed residual video coding (DRVC). The main contributions of this paper are as follows: 1) ESI is not required as our ERC is based on the analysis of the statistical characteristics of the decoder side information (DSI);2) a simple EBMD is introduced which only employs the values of residual pixels at the encoder to classify blocks into Intra mode, Skip mode, and WZ mode;3) an ERC solution using pseudo-random sequence scrambling is proposed to estimate rates for all WZ blocks at frame level instead of at bit plane level, i.e., only one rate is estimated;and 4) a quantization-index estimation algorithm (QIEA) is proposed to solve the problem of rate underestimation. The simulation results show that the proposed solution is not only low complex but also efficient in both the block mode decision and the rate estimation. Also, as compared to DISCOVER system and the state-of-the-art ERC solution, our solution demonstrates a competitive rate-distortion(RD)performance. Due to maintain the low-complexity nature of the encoder and have good RD performance, we believe that our ERC solution is promising in practice.
The High Efficiency Video Coding (HEVC) adopts increasing number of intra prediction modes and Coding Unit (CU) partitions, in order to diminish its cost at required bitrates. However, its attendant computational comp...
详细信息
ISBN:
(纸本)9781538644584
The High Efficiency Video Coding (HEVC) adopts increasing number of intra prediction modes and Coding Unit (CU) partitions, in order to diminish its cost at required bitrates. However, its attendant computational complexity is unfavorable in lots of real-world applications. In this paper, we propose a probability-based strategy that benefits a low-complexity HEVC intra-encoder and also provides guidance to design low-complexity framework of next generation encoder. The proposed strategy comprises three steps: a candidate mode list initialization based on the estimated winning probabilities of all intra modes, an early termination of intra prediction based on the estimated probability of obtaining the best intra mode, and a pre-decision of Coding Unit (CU) split based on the estimated distributions of the Rate-Distortion (RD) costs. Comprehensive experiments have validated the effectiveness of the proposed algorithm, with a promising simulation performance under Common Test Conditions (CTC).
Smart spaces is a scenario that is ideal for the application of wireless communication technology to benefit non-expert users. A prime example of such an application would be low-cost and real-time video surveillance ...
详细信息
Smart spaces is a scenario that is ideal for the application of wireless communication technology to benefit non-expert users. A prime example of such an application would be low-cost and real-time video surveillance of hallways or video conferencing. Traditional solutions for these scenarios use state-of-the-art video codecs like H.264/AVC in order to achieve the low bitrate which is essential in wireless communication. On the flip side, such encoders are computationally complex and need expensive hardware to run on thus making them inefficient choices for being deployed in clusters. Recent research has resulted in distributed video coding (DVC) being proposed as a solution for applications that have limited battery resources and low hardware complexity, thus necessitating a low-complexity encoder. It is now a popular topic in the research community and the past years have seen several implementations. However, current DVC solutions use iteratively decodable channel codes like low-density parity check - accumulate (LDPCA) codes or Turbo codes that have large latencies. In order to make real-time communication possible, the proposed architecture makes efficient use of skip blocks to reduce the bitrate, eliminates the iterative decoding nature of the Wyner-Ziv (WZ) channel and uses a simple data-hiding-based compression algorithm. This drastically cuts down on the time complexity of the decoding procedure while still maintaining a rate-distortion performance better than that of H.264/AVC intra-coding and other current DVC solutions.
暂无评论