A Multi-Surface Estimation Interval Type II Fuzzy Clustering Method for Magnetic Resonance Brain Image Segmentation

Abstract

The present invention is a multi-surface estimation interval type II fuzzy clustering MRI brain image segmentation method. The method first multiplies a plurality of base surfaces by weights to fit the offset field surface, and introduces the objective function as a multiplier item, through iteration The solution makes the fitted offset field surface gradually approach the real offset field; then, the method is upgraded to the interval type II fuzzy field, so that the method can better deal with the fuzzy boundary problem in the iterative process, and improve the ability of the method to deal with fuzzy problems ; Then consider the pixel information of the neighborhood space and add it to the objective function to increase the neighborhood space information. The method of the invention improves the ability to deal with blur problems, reasonably utilizes the neighborhood information, obtains obvious improvements in the offset field correction and segmentation results obtained on the magnetic resonance image, and has broad market prospects and application value.

Description

【Technical field】 [0001] The invention relates to a multi-surface estimation interval type-two fuzzy clustering magnetic resonance brain image segmentation method. The fuzzy clustering and image segmentation techniques are widely used in the field of image applications and belong to the field of digital image processing. 【Background technique】 [0002] Image segmentation is an image processing technology that divides an image into several characteristic and non-overlapping regions based on image feature information. The image regions obtained by image segmentation have common visual characteristics inside, and the image regions constitute the simplified form of the image, which is convenient for image understanding and image analysis. Image segmentation is an important preprocessing technique in the field of computer vision and image recognition. Therefore, it is of great significance to study fast, robust and accurate image segmentation methods. Traditional image segmenta...

AI Technical Summary

Problems solved by technology

However, these methods have insufficient constraints on the estimation of the migration field, resulting in the estimated migration field may not meet the smoothness constraints

Li Chunming and others proposed to use multiple base surfaces to fit the offset field surface and obtain the clustering result by minimizing the noise (see literature: Li Chunming, Gore, Davaziks. For the estimation and organization of the magnetic resonance image offset field Multiplicative intrinsic component optimization for segmentation. Magnetic Resonance Imaging. Volume 32, 913-923, 2014. (Li C, Gore J C, Davatzikos C, “Multiplicative intrinsic component optimization (MICO) for MRI bias fieldestimation and tissue segmentation,” Magn Reson Imaging .,vol.32,pp.913-923,2014.)), but in the correction and segmentation of magnetic resonance image offset field, there are still serious ambiguities, and the effect of the method is still limited

Images