一种改进的基于类间方差的阈值分割法

The algorithm of threshold image segmentation based on the variance between two classes

在图像的目标和背景像素灰度均服从正态分布时 ,提出了一种改进的基于最大类间方差法的图像分割算法 .对新算法进行了测试并与最大类间方差法和最大熵法进行了比较 ,结果表明 ,新的改进算法在图像分割过程中具有速度快。

In consideration of the object and the background in image under normal distributions, the algorithm of threshold image segmentation based on the maximum variance method was presented. It was testified and compared with the Otsu method and the maximum entropy method. The results showed that this algorithm had many advantages such as high segmentation precision and fast computation speed.