基于改进蛙跳算法的图像对比度增强方法

Image contrast enhancement method based on improved shuffled frog leaping algorithm

为改善图像对比度,提出了一种基于改进蛙跳算法的图像对比度增强新方法。该方法利用分段线性变换增强图像的原理,对蛙跳算法进行深入分析并进行改进后,将改进蛙跳算法与二维Otsu法相结合,利用蛙群的并行搜索机制自动选取图像的双阈值,然后再以图像的对比度作为改进蛙跳算法的适应度函数,自适应地搜索分段线性变换的斜率,并以之增强图像。实验结果表明,该方法有效改善原图像的对比度,且优于直方图均衡化法及基于基本蛙跳算法和人工鱼群算法的增强方法。

To improve poor image contrast, this paper proposes an image enhancement method based on improved Shuf-fled Frog Leaping algorithm（SFL）. According to the enhancement rule of traditional piecewise linear transformation, this method first analyzes and improves the basic SFL algorithm. Then a pair of thresholds of the original image is found auto-matically by combining the improved SFL algorithm with 2D Otsu method. Finally, regarding image contrast as the fit-ness function of the improved SFL algorithm, the transformative slopes of the piecewise linear transformation are searched automatically so as to get the enhancement image. Experimental results show that the proposed method not only can effectively better image contrast, but also is superior to some traditional methods, such as histogram equalization method and the enhancement methods based on basic SFL algorithm or artificial fish swarm algorithm.