基于视觉感知的高效视频编码标准帧内量化矩阵优化方法

An Improved Intra Quantization Matrix for High Efficiency Video Coding Based on Visual Perception

该文使用一种基于离散余弦变换(DCT)的调制传递函数(MTF)改进高效视频编码标准(HEVC)帧内量化矩阵,并在计算过程中采用一种新的空间频率计算方法。HEVC整数DCT变换是在DCT变换的基础上经过伸缩扩展以及微调后得到的,与DCT变换有所区别,因此针对这种差别进一步优化校正量化矩阵。仿真过程中设计一种基于结构相似性(SSIM)的Bjontegaard比特率差值(BDBR)方法来测试量化矩阵的编码性能,结果表明在相同的视频质量下,该文设计的量化矩阵能进一步降低码率。该方法在编码过程中只需要简单地改变量化矩阵,没有影响编码模型的算法结构,也不会增加编码的复杂度。

A Discrete Cosine Transform （DCT） based Modulation Transfer Function （MTF） is used to improve the intra quantization matrix for the High Efficiency Video Coding （HEVC） standard. A new method is used to calculate the spatial frequency in the calculation process. The integer DCT for the HEVC is achieved by scaling and hand-tuning the DCT matrix. Due to difference between these two transforms, the quantization matrices are optimized. The experimental results show that the proposed HEVC intra quantization matrix based on the visual perception can reduce more bit rate at the similar video quality by using a Structural SIMilarity （SSIM） based Bjontegaard Delta Bit Rate （BDBR） performance evaluation. Since only the quantization matrices are changed in the encoding process, the proposed algorithm does not affect the structure of the encoding algorithm and does not add the coding complexity.