图象 处 理和模式识别中经常根据物体的形状来识别物

体或对物体进行分类。形状特征可以分为区域特征和边界特

征两大类，矩不变量是最基本的形状特征。Hu提出的矩不变

量是图象区域内部细节的描述，是一种区域特征。它是建立

在对一个区域内部灰度值的统计分析基础上的，需要目标区

域的所有象素参与运算-Image processing and pattern recognition are often based on the shape of objects to identify objects or to classify objects. Shape characteristics and regional characteristics can be divided into two major categories of border characteristics, moment invariants is the most basic features of the shape. Hu moment invariants proposed by the image region is described in detail, the characteristics of a region. It is based on a gray value of region on the basis of statistical analysis, the needs of the target region to participate in all-pixel operations