遗传算法优化归一化划分准则的图像分割

Image segmentation approach of Normalized Cut and genetic algorithm

传统Normalized Cut准则的图像分割需求解广义特征方程,二值化分割效果不佳。为改善图像分割效果,将NormalizedCut准则作为优化函数,使用遗传算法进行优化,通过最优化染色体确定分割结果。实验表明该方法能获得较高精度的分割结果。

Image segmentation using traditional Normalized Cut criterion needs to compute the generalized eigenvector, and the result of the binary segmentation is not satisfactory.In order to improve the image segmentation,this paper uses genetic algorithm to optimize the Normalized Cut criterion,and gets the segmentation results through the optimal chromosome.Experiments show that this method can obtain precision segmentation results.