摘要
边界特征是图像的重要信息,阈值是区分图像像素点的主要依据。研究了图像的二值化方法,阐述了一种基于数学形态学的自适应二值化方法。实验表明,该方法能够较好地保留原图的边界特征信息,二值化效果好。
Threshold is the main principle to differentiate pixels of an image. A new method of adaptive threshold selection is proposed which combines the edge feature that is the important information of the images based on the theory of mathematical morphology. Experiments show that this new algorithm keeps the original edge features good and gets good results.
出处
《科学技术与工程》
2007年第1期139-141,144,共4页
Science Technology and Engineering
关键词
二值化
阈值
数学形态学
边界特征
binarization threshold mathematical morphology edge feature
