An optimized transform and quantization scheme for HEVC intra lossless coding

The lossless coding framework of High Efficiency Video Coding (HEVC) comprises prediction and entropy coding, while omitting the transform process. However, residuals derived from prediction continue to demonstrate significant spatial redundancy, which has the potential to compromise the coding efficiency. This article presents an optimized scheme for HEVC intra lossless coding that integrates transform and quantization. In the proposed scheme, a residual block, generated by HEVC, is partitioned into two components: the coefficient block and the error block. The coefficient block, responsible for lossless coding, is obtained through the adoption of the transform and quantization processes employed in HEVC lossy coding. Subsequently, the error block is derived by predicting the residual block using the generated coefficient block. Both the coefficient block and the error block can be effectively encoded utilizing the existing entropy coding scheme in HEVC lossy coding. Rate‐distortion optimization is employed to determine whether the residual block will be partitioned into two components. The proposed scheme is implemented into the HM 12.1 software. Through the utilization of the discrete cosine transform kernel and the discrete sine transform kernel, with a quantization parameter set to 20, on HEVC standard test sequences and the all intra main configuration file, empirical findings substantiate that the proposed methodology attains an average bit‐rate reduction of 3.31%, with a maximum improvement of 12.29% in comparison to HEVC lossless coding. In this article, we propose an Intra lossless coding scheme based on transformation and quantization for High Efficiency Video Coding (HEVC). The initial residual block is divided into two parts: the coefficient block and the error block. The proposed mode is combined with the transform bypass mode to achieve better coding efficiency at transform units level. Compared with the HEVC lossless coding, the proposed scheme can achieve up to 12.29% bit‐rate saving.