Image segmentation method based on intuitionistic fuzzy c-means clustering

Abstract

The invention discloses an image segmentation method based on intuitionistic fuzzy C-means clustering. It mainly solves the problems of being sensitive to noise, easy to fall into local optimum, and the number of clusters need to be set in image segmentation. Curve fitting, screening all peak points of the fitting curve as a set of initial value ranges of cluster centers, and counting the maximum number of cluster centers; on the basis of , using the pixel position information and gray level information to construct the linearity in the intuitionistic fuzzy objective function The weighting function coefficients are used to obtain the membership degree matrix U; the optimal membership degree matrix is ​​obtained by evaluating U according to the graded distance index evaluation index, and the error detection strategy is used to screen the misclassified pixels for correct classification. The invention enhances the robustness to noise, can adaptively determine the number of image clusters, and can be used for preprocessing of image recognition and computer vision.

Description

technical field [0001] The invention belongs to the field of digital image processing, in particular to an image segmentation method, which can be used for image recognition and computer vision preprocessing. Background technique [0002] Since the 1970s and 1980s, many scholars have continued to pay attention to image segmentation. Image segmentation technology has become a basic technology in many fields. As long as it is related to extracting content in images, image segmentation technology cannot be lacking. The quality and effect of image segmentation work It will directly or indirectly affect subsequent image projects. There are many types of existing image segmentation methods, which can be summarized as threshold-based image segmentation methods, edge-based image segmentation methods, region-based image segmentation methods, and cluster-based image segmentation methods. The image segmentation method based on the image segmentation method and the image segmentation m...

AI Technical Summary

Problems solved by technology

Ahmed et al. added the spatial neighborhood information item to the objective function of fuzzy C-means, and proposed the FCM_S algorithm. Although the algorithm improves the robustness to noise, the computational complexity is high. In order to reduce the FCM_S algorithm Computational complexity, Chen Songcan and Zhang Daoqiang introduce neighborhood information into the objective function of the algorithm through mean filtering and median filtering, and propose FCM_S1 and FCM_S2 algorithms, see: Chen Songcan, Zhang Daoqiang. A stable combination based on kernel functions Spatial information fuzzy C-means image segmentation algorithm. American Institute of Electrical and Electronics Engineers System Control Processing Transactions. Volume 34, 1907–1916, 2004. (S.Chen and D.Zhang, "Robust Image Segmentation Using FCM withSpatial Constraints Based on New Kernel -induced Distance Measure, "IEEETrans.Syst, Man, Cybern, vol.34, pp1907-1916, 2004.); Among these two algorithms, the FCM_S1 algorithm has a better effect on Gaussian noise, but is less robust to salt and pepper noise. Although the FCM_S2 algorithm has a better effect on salt and pepper noise, it is less robust to Gaussian noise. Therefore, these two algorithms cannot be robust and universal to various types of noise.

Cai Weiling and others combined the spatial information and grayscale information of the image to construct a linear weighted sum image, and proposed a fast generation FCM algorithm, see: Cai Weiling, Chen Songcan, Zhang Daoqiang. A fast and robust blurring method for image segmentation that introduces local information C-means Clustering Algorithm. Pattern Recognition. Volume 40, 825-838, 2007. (W.Cai, S.Chen, and D.Zhang, "Fast and RobustFuzzy C-means Clustering Algorithms Incorporating Local Information for ImageSegmentation," Pattern Recognit., vol.40, no.3, pp.825-838, Mar.2007.), this method is robust to Gaussian noise and also robust to salt and pepper noise; but the above algorithms are Without considering more fuzziness of data, Charge et al. further found that using intuitionistic fuzzy set theory can consider more fuzziness of data and classify data more accurately, and proposed a fuzzy clustering method based on intuitionistic fuzzy data. See: A novel intuitionistic fuzzy C means clustering algorithm and its application tomedical images. Appl.Soft Comput. 11(2):1711-1717, 2011.); Since the fuzzy clustering method based on intuitionistic fuzzy data is also sensitive to noise, Verma et al. further introduced local spatial information into the intuitionistic fuzzy C-means algorithm, see: Wei Verma, Agrawal, Sharan. An improved intuitionistic fuzzy C-means algorithm combined with local spatial information for brain image segmentation. Applied Soft Computing. 543-557, 2016. (H.Verma, R.K.Agrawal, A.Sharan , "An Improved Intuitionistic Fuzzy C-means Clustering Algorithm Incorporating Local Information for Brain Image Segmentation," Appl.Soft Comput., 543–557, 2016)

[0004] Although the above improved method optimizes the anti-noise performance of the fuzzy clustering algorithm to a certain extent, it still has insufficient robustness to noise, is sensitive to the initial value of the cluster center, and cannot adaptively analyze the number of image clusters. Not enough

Images