基于自适应搜索的绝对矩块截断编码的图像压缩  

Absolute Moment Block Truncation Coding Image Compression Based on Adaptive Search

作　　者：张茗茗[1] 周诠[1] 呼延烺[1] ZHANG Mingming;ZHOU Quan;HUYAN Lang(China National Key Laboratory of Science and Technology on Space Microwave,Xi′an Institute of Space Radio Technology,Xi′an 710100,China)

机构地区：[1]西安空间无线电技术研究所空间微波技术国家级重点实验室

出　　处：《电讯技术》2019年第12期1404-1410,共7页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(61372175);空间微波技术国家重点实验室基金资助项目(2018SSFNKLSMT-13)

摘　　要：针对现有的固定大小的绝对矩块截断后图像质量不高的问题,提出了基于自适应搜索的压缩算法。首先根据中值确定范围搜索出图像中心阈值,使得图像中心阈值最能代表图像基本内容;然后根据图像中心阈值分段搜索出上下中心,这样在图像块中所有像素与上下中心距离偏差之和最小。该算法确定的上下中心与理想的上下中心一致,与绝对矩块截断编码(Absolute Moment Block Truncation Coding,AMBTC)算法的峰值信噪比最佳值相同,同时耗时较少,可以做到实时处理。仿真结果表明,提出的自适应搜索算法能够较好地提升图像的质量,同时复杂度较低。A compression scheme based on adaptive search for absolute block truncation in fixed size blocks is proposed to solve the problem of blurring display.Firstly,image center threshold is searched according to median range,which can best represent basic image content,and then upper and lower centers are searched according to threshold in segment,so that in each block the whole distortion between pixels and centers is the least.Upper and lower centers determined in this scheme are consistent with ideal centers,and optimal peak signal-to-noise ratio(PSNR)in absolute moment block truncation coding(AMBTC)compression is obtained.Moreover,it takes less time and achieves real-time processing.Simulation results show that the proposed adaptive searching scheme can improve image quality and reduce complexity.

关 键 词：图像压缩 绝对矩阵截断 自适应搜索 

分 类 号：TN919.81[电子电信—通信与信息系统]