Image segmentation method based on characteristic importance sorting spectral clustering

Abstract

The invention discloses an image segmentation method based on characteristic importance sorting spectral clustering, aiming to solve the problems of high complexity and poor stability of the existing spectrum clustering. The realization process comprises the following steps: (1) extracting grey level value characteristic, grey level co-occurrence characteristic or wavelet characteristic from an image to be segmented; (2) performing normalization to the characteristic data; (3) calculating the importance of all the samples according to the normalized characteristic data; (4) sequencing the importance of all the samples, selecting 100 samples with high importance as a sample subset for sampling; (5) according to a method, using the selected sample subset to solve the characteristic vector space of all the samples which are processed through spectral mapping, using the characteristic vector corresponding to k characteristic values of the given clustering number for dimensionality reduction; and (6) performing k-means clustering to the obtained data, and outputting the final image segmentation result. Compared with the existing spectral clustering method, the method has stable result and low complexity, the image segmentation result is obviously increased and the method can be used for target detection and target recognition.

Description

technical field [0001] The invention belongs to the technical field of image processing, relates to image segmentation, and can be used for target detection and target recognition on texture images and SAR images. Background technique [0002] Clustering is the process of distinguishing and classifying things according to certain requirements and rules. In this process, there is no prior knowledge about categories, and there is no teacher's guidance. The similarity between things is used as the classification of categories criteria, and thus fall into the category of unsupervised classification. Cluster analysis refers to the use of mathematical methods to study and process the classification of given objects. It is a kind of multivariate statistical analysis and an important branch of unsupervised pattern recognition. It divides a sample set without a class mark into several subsets according to certain criteria, so that similar samples are classified into one class as muc...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the shortcomings of the above-mentioned existing problems, and propose an image segmentation method based on feature importance sorting spectral clustering, by evaluating the importance of samples and sampling samples with high importance to maximize the retention of image information , ignoring low-importance samples to reduce computational complexity, thereby increasing the speed of image segmentation without increasing computational complexity, and obtaining stable and better image segmentation results

Images