Lossless embedded compression using multi-mode DPCM & averaging prediction for HEVC-like video codec

Video encoders and decoders for HEVC-like compression standards require huge external memory bandwidth, which composes a significant portion of the codec power consumption. To reduce the memory bandwidth, this paper presents a new lossless embedded compression algorithm and a high-throughput hardware architecture. Firstly, hybrid spatial-domain prediction is proposed to combine the merits of DPCM scanning and averaging. The prediction is then enhanced with multiple modes to accommodate various image characteristics. Finally, efficient residual regrouping is used to improve the compression performance based on semi-fixed length (SFL) coding. Experimental results show the proposed scheme can reduce data traffic by an average of 57.6%. The average compression ratio is 2.49, improved by at least 13.2% relative to the state-of-the-art algorithms. Although the complexity is increased, by applying several optimizations the hardware cost of designed architecture increases slightly. This work can be implemented with 54.2k gates cost for compressor and 46k gates for decompressor, which can support 4k×2k@120fps decoder.