Color image segmentation method and system

Abstract

A method and system for color image segmentation are provided. The method comprises: calculating distances between pixel points of the color image based on pixel values of three channels of RGB of each pixel point of the color image and abscissa and ordinate values; Constructing a fused k-nearest neighbor graph fusing k-nearest neighbor graph and inter k neighborhood graph based on the calculateddistances between pixels of the color image; Performing clustering analysis on each pixel of the color image based on the constructed fused k-nearest neighbor graph; The color image is segmented according to the result of the clustering analysis. The color image segmentation method and system according to the invention can increase the computational complexity and obtain a better clustering effectof a non-linear data set, thereby achieving a better color image segmentation effect.

Description

technical field [0001] The present invention generally relates to the field of image processing, and more specifically relates to a color image cutting method and system based on a novel nonlinear clustering algorithm. Background technique [0002] The existing image segmentation methods no longer simply use edge difference segmentation, but use more unsupervised or semi-supervised algorithms to segment images. The new unsupervised or semi-supervised segmentation algorithm mainly improves in two aspects: 1. The feature processing of the original image; 2. The use of machine learning algorithms to improve the segmentation effect. The new unsupervised or semi-supervised segmentation algorithm has many advantages compared with the traditional segmentation algorithm, but there are still some problems: for example, 1. The step of feature processing is increased, which increases the computational complexity and computational time, requiring more 2. Most unsupervised and semi-supe...

AI Technical Summary

Problems solved by technology

[0003] Exemplary embodiments of the present invention aim to overcome the defects of high complexity of feature processing and poor effect on nonlinear data sets existing in the above-mentioned existing image segmentation algorithms

Images