A Fast SAR Image Segmentation Method Based on Fuzzy Clustering of Key Pixels

Abstract

The invention provides a fast SAR (Synthetic Aperture Radar) image segmentation method based on key pixel fuzzy clustering. The method is used for solving the problems of long operating time and low segmentation accuracy of the existing SAR image segmentation method. The method comprises the steps of: 1, inputting an SAR image I to be segmented and a segmentation class number c; 2, performing Gaussian filtering on the image to obtain a filtered image X; 3, dividing the image X into a key pixel set S and a non-key pixel set L according to a local maximum pixel rule; 4, performing fuzzy clustering on key pixels by using spatial information; 5, determining class labels of non-key pixels by using the key pixel clustering result; 6, obtaining an intermediate segmentation result C in combinationwith the class labels of the key pixels and the non-key pixels; and 7, smoothing the result C by using local neighborhood information to obtain a final segmentation result. The method improves the accuracy of the SAR image segmentation result, reduces the segmentation time, and provides a foundation for later SAR image understanding and interpretation.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a SAR image segmentation method, which can be used for image semantic recognition and image search. Background technique [0002] Synthetic Aperture Radar (SAR) images have been widely used in military and civilian fields in recent years because of their imaging process which is not affected by weather and time. SAR image segmentation is a prerequisite and basic technology for SAR image understanding and interpretation, so there is a great demand for accurate and fast SAR image segmentation. SAR image segmentation is the process of dividing a SAR image into a certain number of non-overlapping, uniform regions. The pixels in the same region have similar characteristics, and the pixels in different regions have different characteristics. However, the widespread speckle noise in SAR images makes the precise segmentation of SAR images a challenge. At the same time, as the s...

AI Technical Summary

Problems solved by technology

The traditional fuzzy clustering method only processes a single pixel. For SAR images rich in speckle noise, the segmentation process is seriously affected by noise, which makes the accuracy of the final segmentation result very low.

[0004] With the increasing requirements for SAR image segmentation in modern society, various traditional segmentation algorithms have the disadvantages of low accuracy, poor robustness to noise, and slow running time, which makes the segmentation results obtained by traditional methods unsatisfactory. Requirements, researchers have made some improvements to these defects, such as improvements to traditional fuzzy clustering: Gong Maoguo et al. published a paper "FuzzyC- means clustering with local information and kernel metric for imagesegmentation", this paper adds neighborhood items to the original fuzzy C-means clustering objective function, and introduces local spatial distance and gray level difference information, making the segmentation process robust to noise At the same time, the kernel method is added to the similarity measurement process, so that the segmentation process can further suppress the influence of speckle noise

However, for images affected by severe speckle noise, this method will still be affected by noise, resulting in lower segmentation accuracy

At the same time, the existing clustering methods for image segmentation process each pixel in the image, which makes the iterative process of the clustering method very slow, and this phenomenon is more serious in the process of large-scale image segmentation

Therefore, traditional methods are difficult to achieve fast and accurate segmentation of SAR images

Images