基于函数级进化型硬件的无损图像压缩

Lossless image compression based on function-level evolvable hardware

为实现无损图像压缩,设计了基于函数级进化型硬件(FEHW:Function-levelEvolutionalHardware)无损图像压缩系统。该系统由进化单元、函数单元和分类单元组成,运用遗传算法(GA:GeneticAlgorithm)自适应地改变其配置结构,在压缩过程中以函数进化的方式,进行像素预测和误差预测,以此增强无损图像压缩的性能。运用LOCO(LowComplexityLosslessCompression),JPEG(JointPhotographicEx-pertsGroup),CALIC(Context-basedAdaptiveLossslessImageCoding)和GA无损压缩算法进行了仿真,给出了不同算法对9种标准图像无损压缩的比较结果,该方法平均压缩率为4.885bit/pixel。

The lossless image compress system based on FEHW(Function-level Evolutional Hardware) was designed to realize lossless image compress. This system is composed of evolution unit,function unit and classification unit, its configuration is self-adapt changed by GA(Genetic Algorithm). It can do pixel prediction and error prediction by function evolutional mode during lossless image compressing to improve the performance of lossless image compress. The LOCO(Low Complexity Lossless Compression),JPEG(Joint Photographic Experts Group),CALIC(Context-based Adaptive Losssless Image Coding) and GA algorithms have been simulated,the results of lossless compress are given. Comparisions of 9 types standard image by different algorithms are also given. The average compress ratios was 4.885 bit/pixel.