Cervical cell nucleus segmentation method and system based on HSV channel and multi-dimensional feature screening

By using a method based on HSV channels and multidimensional feature selection, the problems of insufficient utilization of color features and insufficient parameter adaptability in cervical cell nucleus segmentation are solved, achieving high-precision, low-false-positive cell nucleus segmentation that can adapt to full-view images under different conditions.

CN122454568APending Publication Date: 2026-07-24HEER MEDICAL TECH DEV CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202610589033.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-04-29
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Existing technologies for cervical cell nuclear segmentation suffer from problems such as insufficient utilization of color features, lack of multi-dimensional morphological feature screening, lack of adaptive parameter settings, and lack of cytoplasmic environment verification, leading to inaccurate segmentation, high false positive rate, and incomplete boundaries.

Method used

A method based on HSV channels and multidimensional feature selection was adopted. By converting the image to HSV space, extracting the V channel brightness component, performing threshold segmentation and morphological operations, and combining shape feature selection and cytoplasm verification, the segmentation results were ensured to conform to biological characteristics.

Benefits of technology

It improves segmentation accuracy, reduces false positive rate, ensures the integrity and biological rationality of cell nuclear boundaries, adapts to cell nuclei of different sizes and shapes, and reduces sensitivity to light and staining batches.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a cervical cell nucleus segmentation method based on an HSV channel and multi-dimensional feature screening, and the method comprises the steps: converting a cell image from RGB to HSV, and extracting a V channel brightness component; performing threshold segmentation and morphological operation on the V channel to obtain a cytoplasm mask; meanwhile, contrast enhancement, threshold segmentation and morphological operation are carried out on the V channel to obtain a binary mask; extracting a contour, and sequentially carrying out area preliminary screening and shape feature screening; a mononuclear mask is created for the contour meeting the requirement, the number of overlapped pixels of the surrounding area and the cytoplasm mask is calculated after expansion, and if the number exceeds a dynamic threshold value, it is judged that the cell nucleus is qualified; and finally drawing a contour to obtain a segmentation result. According to the method, the low-brightness characteristic of the cell nucleus is fully utilized, and the distinction degree of the cell nucleus and the background is enhanced; through multi-dimensional screening and cytoplasm environment verification, the cell nucleus contour conforming to the biological characteristics is obtained, deep learning is not needed, the interpretability is high, and cell nucleuses of different sizes and forms in the full-view image can be accurately segmented.
Need to check novelty before this filing date? Find Prior Art