【摘要】 颜色和形状信息是识别物体的重要信息 .本文提出用数字滤波器将图象空间分割为 H M和M\H 空间的一种方法 .该方法进一步将分析彩色图象为众所周知的测量分割部分几何特征的算法 ,包括通过节点的阶数确定分割部分的分布及动态特征 ,包括速度和移动方向的分布等更多还原

【Abstract】 Color and shape information are very important for us to recognize objects. In this paper, we represent a digital chromaticity filter as a device, which performs segmentation of the image space into two: \%HM\% and \%M\H\%. Segmentation by the above methods further reduces the analysis of a color image to well known algorithms for measuring geometric characteristics of segments, including their distribution by dimensions, areas, perimeters, and so on; the topological characteristics, including distribution of segments by the order of nodal points by connectivity, ranking, etc.; and the dynamic characteristics, including distribution by speed and direction of motion, etc.更多还原