摘要
研究了黑麦(Secale cereale)染色体C-带图象电脑自动分析技术,对特征带纹的提取方法和染色体电脑图象带型模式建立的方法进行研究,最后就染色体分析的若干参数,进行了人工分析和电脑分析的比较,以确定电脑分析方法的可行性和可靠性。通过两种方法比较结果,经统计分析,其差异未达到显著标准,说明电脑分析的结果是可靠的。
In the present study, the photograph of C-banded mitotic chromosomes of Secale cereale cv. Jingzhou Heimai was used as the original medium (Plate, A). It was inputted to a computer system to establish a standardized computer karyotype model of rye which could make the research of cytology, chromosome genetics and genetic manipulation more effective and accurate. For aligning each chromosome from the arbitrary shape, following steps were taken. 1, segmenting chromosomes from the background by an image segmentation algorithm; 2, extracting individual chromosomes) 3, finding out the central skeleton of each chromosome by a thinning algorithm; 4 , fitting a curve represe nted by a function with central skeleton of chromosome by a curve fitting algorithm; 5, aligning the curve on horizontal axis; 6, the grey levels in aligned chromosome are estimated from original image by bilinear interpolation. In extracting banding features, an image entropy operator and a maximum entropy of histogram of chromosome image were used to segment the bands into three levels: dark bands, white bands and grey bands, which are the main symbols for recognizing chromosomes in a computer system. On this basis, the computer image model of C-banded rye chromosomes was established (Plate, B), which kept the original characteristics of the chromosomes and was characterized by its typicality, universality and practicality due to the modifications by automated processing. Several chromosomal parameters such as chrosomome lengths and band intervals measured by manual and automated methods were compared in pairs (Tables 1,2). The chromosomal parameters measured on the same chromosomes by these two methods did not show significant difference except on one chromosome, indicating the reliability and feasibility of the automated analysis.
出处
《江苏农业学报》
CSCD
1990年第3期17-24,共8页
Jiangsu Journal of Agricultural Sciences