最大类间方差法，以其分割效果较好、适用范围较广、简单有效而引起人们普遍关注，且应用最为广泛。该方法基于类别可分离性，根据图像的一维灰度直方图，选取一个阈值使得类间方差最大。-Otsu method, with its split better, broader scope of application is simple and effective and caused widespread concern, and the most widely used. The method is based on class separability, according to the image of the one-dimensional histogram, select a threshold allows maximum between-class variance.