基于改进遗传算法的自动阈值图像分割方法

Automatic Threshold Image Segmentation Approach Based on Improved Genetic Algorithm

图像分割法在图像分割的过程中只考虑象素的灰度值,没有考虑空间特性和存在计算复杂性过大的缺陷,影响图像效果。针对上述问题,提出一种基于改进遗传算法的自动阈值图像分割算法。方法对遗传算法中的编码方式、交叉算子、变异算子等参数进行了一些适当改进。把图像分割最佳阈值选取转换成优化问题。利用改进遗传算法的寻优高效性求解最佳阈值,实现图像分割。仿真结果证明,新算法极大地缩短了寻优时间,增强了图像分割过程中的抗噪性能,提高了图像分割的效率。从而有利于计算机视觉的后续处理,可以实现实时图像分割,具有实用价值。

An automatic threshold selection for image segmentation algorithm was presented based on improved genetic algorithm.It can overcome the shortcomings of the existing image segmentation methods which only considered pixel gray value without considering spatial limitations and computational complexity.In this new method,some improvements in genetic algorithm＇s coding,cross operator,mutation operator were made.The optimal threshold segmentation was converted into an optimization problem.The temperature parameters in annealing algorithm simulation were used to change the selection pressure,so as to optimize the selection algorithm.The improved genetic algorithm which was highly efficient in optimization also was used in order to find the optimal threshold value and accomplish the segmentation of images.Simulation experiments have proved that the new algorithm can greatly shorten the time for optimization,enhance the anti-noise capability in the segmentation process,and improve the efficiency of image segmentation.It can facilitate the post treatment of computer vision and can be applied to real-time image segmentation.So this method has practical value to some extent.