边缘检测是图像预处理的重要内容之一,在对Pal和King经典模糊边缘检测算法改进的基础上,提出了一种基于遗传算法和Otsu进行图像阈值选取,以不同阂值为基准确定出线性隶属函数,对多峰图像确定多闽值隶属函数的方法,进行模糊增强,从而提取边缘,实验结果表明该算法不仅提高了边缘提取质量,而且缩短了阈值选取时间,是一种有效性、正确性的图像处理方法。
Image segmentation is one of basic research issues of image processing and computer vision.This paper makes some improvement to the Pal and King fuzzy edge-detection algorithm.A scheme for thresholding using genetic algorithms and Ostu is presented,according to optimal threshold a new line membership function is given,extending to more threshold.Then fuzzy enhancement is processing,at last rather fine image edge is gained.The experiment results show that this algorithm is not only higher segmentation quality but also higher computational speed,so it proves the algorithm is right and efficient.