一种基于人眼对比度敏感视觉特性的图像自适应量化方法

为了提高图像的压缩比和压缩质量，结合人眼对比度敏感视觉特性和图像变换域频谱特征，该文提出一种自适应量化表的构建方法。并将该表代替JPEG中的量化表，且按照JPEG的编码算法对3幅不同的彩色图像进行了压缩仿真实验验证，同时与JPEG压缩作对比分析。实验结果表明:与JPEG压缩方法相比，在相同的压缩比下，采用自适应量化压缩后，3幅解压彩色图像的SSIM和PSNR值分别平均提高了1.67%和4.96%。表明该文提出的结合人眼视觉特性的自适应量化是一种较好的、有实用价值的量化方法。

An Adaptive Quantization Method of Image Based on the Contrast Sensitivity Characteristics of Human Visual System

In order to improve the compression ratio and quality of the image, combined with the contrast sensitivity characteristics of human vision system and the spectrum characteristics of image in the transform domain, a method is proposed to form the adaptive quantization table in image compression. And according to the JPEG coding algorithm and replacing the quantization table in JPEG, simulations are carried out for three images by programming, whose results are compared with JPEG compression at the same time. The results show that: compared with JPEG compression, under the same compression ratio, average SSIM and PSNR of three decompressed images increase by 1.67% and 4.96% after being compressed using adaptive quantization, respectively. They show that the adaptive quantization based on HVS is a good and practical method.