摘要
1975年美国喷气推进实验室根据与美国国家儿童卫生和人类发展协会签订的合同,研制出世界上第一台染色体计算机识别系统,接着比利时、意大利、日本等国也研制了相应的系统,所有这些均由于染色体的复杂性而采用了人机交互方式分析,而完全的自动分析尚难实现.因此,用于改善染色体图像质量的预处理,具有重要的意义.本文利用Freeman码跟踪染色体图像边缘,在此基础上提取染色体.及研究了染色体图像的精化处理.1 染色体提取为了提取染色体,必须对其边界轮廓进行跟踪,一般用Freeman方向码(图1)跟踪.任一曲线链码符号为:
A method for segmenting the image of human chromosome is proposed. The image is first processed by two-valued processing and then the edge of the image is tracked and segmented with Freeman codes. The chromosome is then extracted. This procedure is called coarse processing. The edge of the image so extracted is however indistinct. For this reason, it is proposed that the image be further processed by three-valued segmentation, for which the statistical properties of the shade band, light band, intermediate band and indistinct edge band are used. But, in three-valued segmentation, the intermediate band and indistinct edge band are of the same level. By eliminating the indistinct edge band and reserving the other three, a distinct image will be obtained. This is a procedure called exact processing.
出处
《华中理工大学学报》
CSCD
北大核心
1992年第3期181-183,共3页
Journal of Huazhong University of Science and Technology
基金
"七五"攻关项目:"75-68-5-20
显微图像分析处理系统"子专题
关键词
人
染色体图像
分割法
计算机
human chromosome
freeman codes
coarse processing
exact processing