Grey scale characteristic graph-based automatic separation method for conglutinated chromosomes

Abstract

The invention provides an automatic separation method for conglutinated chromosomes. The method firstly carries out immersion and corrosion processing operation based on a grey scale characteristic graph of chromosomes, and then carries out conditioned expansion recovery processing operation. The invention designs a grey scale characteristic graph-based morphological processing method to separate the chromosomes which are seriously conglutinated but are neither sheltered nor overlapped. The method solves the technical problems and fills up the technical vacancy of separating the chromosomes which are seriously conglutinated but are neither sheltered nor overlapped in the prior art, meets the requirements of accurately separating the conglutinated chromosomes and detecting the number of the chromosomes in medical fields of chromosome analysis, cell research, pathological analysis, and the like, and further improves the automaticity in a chromosome analysis system. The method provides individual chromosomes with good continuous boundary effect and independence for the automatic separation processing in the chromosome analysis system, and particularly keeps concave and convex details of original chromosome image boundary completely and clearly so as to improve the separation automaticity of the conglutinated chromosomes and the accuracy of the chromosome number counting.

Description

【Technical field】 [0001] The invention belongs to the technical field of image processing, and relates to an automatic segmentation technology of cohesive chromosomes based on grayscale feature images, in particular to a segmentation method for severe cohesive chromosomes in a chromosome analysis system. 【Background technique】 [0002] In the automatic chromosome analysis system, the chromosomes collected at the end of prophase and metaphase always have adhesion phenomenon. In this regard, early chromosome analysis systems often used manual segmentation to separate chromosomes, providing individual independent chromosomes for the next step of chromosome classification and pairing. With the development of image processing technology, automatic segmentation methods are gradually introduced into it. At present, a variety of segmentation techniques and methods for cohesive and overlapping chromosomes have been developed and developed at home and abroad. There is a segmentation ...

AI Technical Summary

Problems solved by technology

If the traditional watershed algorithm is used to obtain the gradient information from the gray image as the input image to achieve segmentation, if the gradient calculation of the original cohesive chromosome gray image does not change much or the gradient calculation is not performed, it is easy to produce The over-segmentation of cohesive chromosomes is wrongly segmented into multiple chromosome blocks, but if filtering is performed, it is easy to combine multiple cohesive chromosomes into a whole

[0003] Therefore, it is necessary to provide a unique segmentation method to solve the problem of chromosome segmentation with serious adhesion and no occlusion and overlap.

Images