摘要
阈值法是图像分割的一种重要方法,在图像处理与识别中广为应用。本文提出了一种基于最大相关准则的自动选择阈值方法。该方法通过使图像中目标和背景分布的相关量最大来选择阈值。仿真分析显示该算法计算量与熵方法比较大大减少,分割效果则比Yen等人提出的算法更佳。
Thresholding is an important form of image segmentation and is used m the image processing for many applications. In this paper, an automatic technique for thresholdiug of digital images based on the maximum correlation criterion is presented. This method maxiiuizes the correlatious associated with the distributions of the background and object, classes to obtaiu the optimal threshold. It is found that the proposed approach is more effective than that proposed by Yen, and its computational complexity is inuch lower cornpared with the metliods of ent,ropy.
出处
《电子与信息学报》
EI
CSCD
北大核心
2002年第1期136-139,共4页
Journal of Electronics & Information Technology
