摘要
为了提升绝对矩块截断编码(absolute moment block truncation coding,AMBTC)重构图像的质量,结合遗传算法提出一种基于块截断编码(block truncation coding,BTC)的压缩方法,通过搜索不重叠子图像的高低量化值,得到优化的编码组.实验结果表明,与AMBTC相比,新方法得到的重构图像的峰值信噪比提升了0.5~0.8dB.
In this paper,a compression scheme is proposed,which cooperates block truncation coding(BTC)with genetic algorithm(GA)to improve the quality of reconstructed images of absolute moment block truncation coding(AMBTC),in which optimal groups of codes are obtained via searching high/low quantization levels within non-overlapping sub-blocks.Experimental results show that compared with AMBTC the quality of reconstructed images acquired by proposed scheme is increased by 0.5~0.8 dB in peak signal to noise ratio(PSNR).
作者
回征
周诠
张晨光
HUI Zheng;ZHOU Quan;ZHANG Chenguang(Xi’an Institute of Space Radio Technology,Xi’an 710100,China)
出处
《扬州大学学报（自然科学版）》
CAS
北大核心
2018年第3期40-43,共4页
Journal of Yangzhou University：Natural Science Edition
基金
国家自然科学基金资助项目(61372175)
关键词
图像压缩
绝对矩块截断编码
遗传算法
峰值信噪比
digital image compression
absolute moment block truncation coding
genetic algorithm
peak signal to noise ratio
