A multi-modal medical image fusion method based on guided filtering and graph theory significance

Abstract

The invention requests to protect a multi-modal medical image fusion method based on guided filtering and graph theory significance, which is mainly used for fusing a multi-modal medical image and comprises the following steps: S1, obtaining a smooth sub-band image and a detail sub-band image of a to-be-fused image; S2, constructing a fusion rule recombination smoothing sub-band coefficient and adetail sub-band coefficient by using a graph theory saliency method; and S3, reconstructing the fused smooth sub-band coefficient and detail sub-band coefficient to obtain a fused image. The decomposition method based on the guide filtering algorithm is low in time complexity, guarantees image fusion, can effectively reduce noise interference in the decomposition process, and improves the robustness of the method. And meanwhile, the brightness, color and direction characteristics of the smooth sub-band image are extracted by utilizing a saliency method of the graph theory, so that the brightness information of the gray level image and the color information of the pseudo-color image can be reserved, the problems of brightness information loss and color distortion of the fused image in otherfusion methods are effectively solved, and the quality of the fused image is ensured.

Description

technical field [0001] The invention belongs to the technical field of image processing, and is used to fuse medical images of multiple modalities, mainly the fusion of gray images (MRI) and pseudo-color images (PET), that is, MRI-PET fusion and grayscale images (MRI) and gray The fusion of high-resolution images (CT) is the fusion of MRI-CT, especially involving the use of guided filtering algorithms and graph-theoretic saliency algorithms. Background technique [0002] At present, image fusion is a research hotspot in the field of image processing research. The main research methods can be summarized into two categories: image fusion methods based on space domain and image fusion methods based on frequency domain. [0003] The image fusion method based on space domain is to directly construct the fusion rule for the original image to obtain the fusion image. Commonly used image fusion methods based on space domain include simple weighted average, fusion method based on PCA...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims to solve the problems of loss of brightness information, color distortion and poor anti-noise ability in the prior art

Images