适合于高分辨力航测图像压缩的低复杂度算法  

Low Complexity Algorithm Suitable for Compressing High-resolution Aerial Image Data

作　　者：李其虎[1,2] 任国强[1] 吴钦章[1] 

机构地区：[1]中国科学院光电技术研究所,四川成都610209 [2]中国科学院研究生院,北京100149

出　　处：《电视技术》2011年第17期21-24,47,共5页Video Engineering

基　　金：国家"863"计划项目(2007AA802401)

摘　　要：针对高分辨力航测图像数据量庞大的问题,提出了一种基于小波分析的低复杂度图像压缩算法。该算法首先对图像进行5级二维5/3小波变换,以去除图像像素之间的相关冗余。依据正交小波变换的子带变换增益对变换后的图像系数进行最佳量化以去除视觉冗余,最后针对小波变换后图像系数的概率分布特点对小波系数最低频子带LL采用一种基于上下文的预测熵编码,对其他频带采用特定的系数扫描方式,并应用哥伦布指数编码联合自适应游程编码实现图像压缩。实验结果表明该算法在保证图像压缩质量前提下可以获得较高压缩比,且可以支持无损到有损的图像压缩,算法复杂度底,易于硬件实现。In order to solve problem that the high-resolution aerial image data is greatness, a low complexity algorithm of image compression is proposed based on wavelet analysis in this paper. Firstly, the co relational redundancy between the image pixels is removed by the two-dimensional 5/3 wavelet transform of five levels to image data. Then visual redundancy is removed by the optimal quantitative to transform image coefficients according as sub-band transform gain of wavelet transform. Finally, by the characteristic of the probability distribution that the wavelet transform coefficients, the prediction entropy coding based on context for lowest frequency sub-band. For other band with Columbus imtex encoded joint adaptive run-length coding and using specifically scan. Experimental results show that the algorithm has the premise of the image compression quality get higher compression ratio, and supports coding from loss to lossless, and low complex, and it is very suitable for hardware implementation.

关 键 词：图像压缩 DWT 最优量化 预测编码 

分 类 号：TN911.73[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]