基于图像处理的人类染色体自动分类系统设计  

An Image-processing-based Design for the Automatic Classifying System of Chromosome

作　　者：胡晓飞[1] 

机构地区：[1]南京邮电学院,江苏南京210003

出　　处：《连云港化工高等专科学校学报》2002年第4期24-26,共3页Journal of Lianyungang College of Chemical Technology

摘　　要：介绍了1种染色体自动分类软件的设计方法。系统主要实现染色体的自动识别、排序和配对,同时检测出变异染色体,进而诊断疾病。系统的设计运用了邻域平均法、中值滤波法、局部统计滤波、状态统计滤波以及四邻接法等图像滤波方法,以及梯度法、门限法和边缘跟踪法等图像锐化方法,通过比较各种方法的程序运行结果找到最优算法。A designing method for chromosome automatic classifying software is presented. The system accomplishes the automatic recognizing, sorting, and pairing, of chromosome, the pick-out of aberrant chromosomes and the diagnosis of diseases thereby. A lot of digital image processing methods are applied in the system. Five methods of image smoothing are adopted, such as neighborhood averaging, median filtering, local statistical filtering, state statistical filtering and four-adjacency method. Besides, three methods of edge tracing are adopted, such as gradient method, threshold method and edge-tracing method. The best algorithm are found on the grounds of the results of the program.

关 键 词：图像处理 人类染色体 自动分类系统 设计 软件 图像处理 

分 类 号：Q987.1[生物学—遗传学] TP391[生物学—人类学]