基于DCT域的高压缩图像去块效应算法

块离散余弦变换(B lock D iscrete Cosine Transform)的主要缺点是在低比特率时其恢复图像的块边界上会出现明显可见的方块效应,降低了图像的视觉质量.为了尽量消除块效应并保护图像的边缘信息提出了一种基于DCT域的块效应消除算法.该算法充分利用了人类视觉系统(Hum an V isual System)特性,建立了块效应模型并给出一个简便的检测边缘标准,对于平滑区,对影响块效应的参数进行修正,然后用线性函数块代替阶跃函数块去除块效应,最后再对更新块和纹理区在DCT域中进行后滤波.仿真结果验证了本文算法的有效性.

Reduction of blocking artifacts for highly compressed images based on discrete cosine transform domain

A major drawback of block discrete cosine transform(BDCT) is that the decoded images,especially at very low bit rates,exhibit highly noticeable blocking artifacts near the block boundaries.A blocking reduction algorithm in DCT domain is proposed to reduce blocking artifacts as many as possible,while preserving edge information.The characteristic of Human visual system(HVS) is extensively utilized in this paper.The model of blocking artifacts is built and a convenient edge detection measurement is introduced. For smooth regions,the step block is replaced with linear block and the parameter that affects blocking artifacts is modified.The DCT-domain post-filtering method is applied to the updated blocks and texture regions.The experiment results show the effectiveness of the proposed algorithm.