A Genetic Fuzzy Clustering Image Segmentation Method

Abstract

The invention discloses a genetic fuzzy clustering image segmentation method. Aiming at the problem that the fuzzy C-means algorithm is sensitive to noise and the cluster centers are too close to each other due to the influence of noise, a fuzzy clustering image segmentation method based on genetic algorithm is proposed. The segmentation method first uses the gray level and neighborhood information to perform anti-noise preprocessing on the original image; then uses the genetic fuzzy clustering algorithm to obtain the preliminary optimal cluster center; finally, according to the obtained preliminary optimal cluster center, uses The histogram of the anti-noise image corrects the cluster center, calculates the membership degree of each pixel in the image, and obtains the segmentation result. In the anti-noise preprocessing, the method adopts the improved gray similarity function to ensure the anti-noise effect under the noise with large intensity; Segment images with severe noise interference and small targets to be segmented, obtain correct cluster centers, and obtain accurate segmentation results.

Description

technical field [0001] The invention belongs to the field of image processing and application, in particular to a genetic fuzzy clustering image segmentation method. This method can effectively suppress the noise, and can improve the image segmentation accuracy when the noise interference is serious, and prevent the mis-segmentation caused by the cluster centers being too close. Background technique [0002] Fuzzy C-Means algorithm (Fuzzy C-Means, FCM) is a clustering algorithm with fuzzy decision-making ability, which is very effective for the segmentation of fuzzy boundary areas, and has been widely used in the field of image processing in recent years. The FCM algorithm tends to converge to local extremums. In response to this problem, Bezdek et al. proposed a method to optimize FCM using genetic algorithms to ensure the global optimal solution (Bezdek J.C., Optimization of fuzzy clustering criteria using genetic algorithms, IEEETransaction on Evolutionary Computation, 2...

AI Technical Summary

Problems solved by technology

[0004] Noise interference will not only cause noise in the segmentation results and affect the accuracy of segmentation, but also interfere with the algorithm's selection of cluster centers

In response to this problem, Krishnapuram et al. (Kirshnapuram R.A., Possibilistic approach to clustering.IEEE Transactions on Fuzzy Systems, 1, 1993:98-100) proposed a Possiblistic Clustering Method (PCM), which removes the single The constraint that the sum of the membership degrees of pixels for each cluster center must be 1 allows some pixels to have a small membership degree relative to each cluster center, so that it does not affect the calculation of the cluster center, but it is difficult to apply PCM alone image segmentation

Images