基于g-l-PSO算法的灰度图像增强方法

A Gray Image Enhancing Algorithm Based on g-l-PSO

规则化Beta函数给出了灰度图像对比度变换函数的统一表达形式,但是Beta函数的参数需要根据具体图像而确定,难于找出合理的参数值.首先分析了传统PSO算法的不足,使用种群的局部最优解为遗传依据,全局最优解为变异依据,提出综合的g-l-PSO算法.将g-l-PSO算法用于确定Beta函数的参数,通过适应度评估,可以获取最适合当前图像的参数值.通过三幅Pout图像的增强效果分析,可知该灰度图像增强方法切实可行.

Regular Beta function gives a uniform expression of gray image contrast transform functions.But the two parameters of Beta function relate to specific images and they are difficult to find reasonable parameter values.This paper analyzed the traditional PSO algorithm and proposed a g-l-PSO algorithm.Using the local optimal solution for the population genetic and the variation was based on the global optimal solution.The Beta function parameters were determined g-l-PSO through the fitness assessment.This work can obtain the most suitable parameter values for current image.Three typical images enhancing experiments show that this approach is feasible.