基于遗传算法的分形图像压缩技术研究

摘要

针对分形图像压缩过程中匹配编码效率和保证解码图像质量的冲突问题,构造一种基于遗传算法的适应度函数,对杂交算子和变异算子进行设计和优化.实验结果表明,新算法与Fisher自适应四叉树算法、均值聚类算法和自适应遗传等四种算法,以编码耗时(s)、压缩比和PSNR差值为指标进行比较,在保证解码图像质量的前提下,新算法匹配编码效率获得了令人满意的效果.该项研究成果为探索图像新型压缩算法提供了一种途径,具有重要参考借鉴价值.%Aiming at the conflict between efficiency of code matching in the process of fractal image compression and ensuring the quality of decoded images, we construct new fitness function which is based on genetic algorithm to design and optimise the crossover operator and mutation operator. Experimental results show that, the new algorithm can obtain satisfactory results while ensuring the quality of decoded image comparing with the algorithms of Fisher adaptive quadtree, K-means clustering algorithm and the adaptive genetic algorithm in terms of encoding time consuming (s), compression ratio and PSNR difference. This research outcome provides a new way for exploring the novel image compression algorithm and has an important reference value.