基于灰度共生矩阵的阈值分割算法

针对目前比较成熟的分割方法存在的局限性,提出一种基于灰度共生矩阵的阈值分割方法。该方法利用灰度共生矩阵中包含的图像灰度值的分布和边缘信息,定义一个新的统计量——边缘强度,将其作为确定分割阈值时的一个重要参数,并与传统分割方法进行实验比较。实验结果表明:该方法在处理模糊边界时能取得比传统分割方法更好的分割效果,很好地保留需要的目标边缘。

Algorithm of Threshold Segmentation Based on Gray Level Co-Occurrence Matrix

Aimed at the limitation of mature segmentation method, a threshold segmentation based on gray leve co-occurrence matrix （GLCM） was proposed. The method utilizes the distribution of gray level values and edge informatiol contained in GLCM, then defines a new statistic： edge magnitude as an important parameter for determining the threshold value. A comparison is executed between the new segmentation method and the traditional one. The experimental result show that this method is better than the traditional segmentation method in dealing with fuzzy boundary, and can preserving well target edge,