摘要
本文给出了物体的三维矩描述,提出了连续切片三维空间中的二值离散群体的相似矩匹配检测方法和微机实现途径,理论分析和实验结果均表明,相似矩匹配检测法不仅极大地减少了三维二值离散群体检测对计算机内存的需求,且具有实现容易、计算量小、准确率高等优点。
This paper gives moment's description of 3-D object and presents similar moment matching detection method for binary discrete objects from 3D-space of series slices. Theoretical analysis and experimental results show that it saves computer memory extremely and can easily be realized, and it decreases calculation and improves detecting ratio. Number of objects, start section and end section centers of each object are obtained after processed whole series slices.
关键词
模式识别
生物医学
离散群体
检测
Pattern recognition, Moment's descriptor, Similar moment matching detection
