基于边缘信息的二维Otsu阈值分割算法的改进

Improvement of 2-D Otsu threshold segmentation algorithm based on edge information

在灰度图像的处理中,Otsu法是普遍的阈值分割方法,在对Ostu阈值分割方法不断改进的过程中,以二维的Otsu阈值分割方法应用最为广泛,在二维的Otsu方法的统计量中加入均值的方式虽然在一定的程度上消除了噪声,但对于背景与目标区域不明显时,二维Otsu阈值分割不能得到理想的效果。文中在二维Otsu方法改进的过程中,在不增加算法复杂度的情况下使用Sobel算子加入边缘信息,可以有效刻画边缘信息,同时可以很有效减少边缘处的噪声。

In the gray image processing,Otsu method is a common threshold segmentation method. In the process of Ostu threshold segmentation method improvement,the two-dimensional Otsu threshold segmentation method is the most widely used. The addition of the mean in the two-dimensional Otsu method statistic eliminates noise to a certain extent,but for the background and the target area is not obvious,two-dimensional Otsu threshold segmentation can not achieve the desired effect. In this paper,without increasing the complexity of the algorithm,edge information from Sobel it is effectively used to describe edge in the process of two-dimensional Otsu method. At the same time it is very effective at the edge of the noise reduction.