摘要
采用差图象动目标检测、窗口自适应分割等技术实现了精子目标的快速检测定位和分割，克服了因目标亮暗不均造成的分割误差和目标丢失。利用形态特征鉴别静止精子和大小差不多的杂质，为高效实用的ＣＡＳＡ（ＣｏｍｐｕｔｅｒＡｉｄｅｄＳｐｅｒｍＡｎａｌｙｓｉｓ）系统奠定了基础。
Fast Detection and Segmentation of Moving Sperm targets was performed based on MAD analysis and adaptive thresholding. Static sperms was discriminated from false target via their differences in shape features so as to build a good foundation for an effective Computer Aided Sperm Analysis system (CASA).
出处
《中国图象图形学报（A辑）》
CSCD
1999年第5期405-408,共4页
Journal of Image and Graphics
关键词
精子运动分析
动目标检测
图象分割
Sperm motion analysis, Moving targets detection ,Image segmentation
