Medical image fusion method based on dictionary learning and low-rank representation

Abstract

The invention provides a medical image fusion method based on dictionary learning and low-rank representation, and the method comprises the following steps: firstly segmenting an image through employing a window sliding technology, and carrying out the classification of a multi-source image sample according to a gradient histogram feature (HOG); secondly, converting the image into a vector, performing dictionary learning, and training a dictionary through a multi-iteration singular value decomposition (K-SVD) method; then obtaining a coefficient which is represented by a low rank by using a low-rank representation (LRR) method; obtaining a preliminary fusion image through a 1-norm maximum principle and a fusion rule; and finally, obtaining a final fusion image through image compensation. According to the method, a good effect is achieved in the aspects of subjective vision and objective indexes of the image, and the fused image with good quality is obtained.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a medical image fusion method based on dictionary learning and low-rank representation. Background technique [0002] Medical images play an important role in clinical diagnosis and surgical navigation. However, due to differences in imaging mechanisms, different medical images differ in the expression of tissue and organ information. For example, computed tomography (CT) imaging can accurately detect dense structures such as bones and implants. Magnetic resonance imaging provides high-resolution anatomical information for soft tissues, but the diagnosis of fractures is less sensitive than CT. The imaging of the same human organ tissue through a single modality medical image can only reflect limited structure, morphology and information. In order to obtain sufficient diagnostic information, doctors need to extract information from images of different modalities. Obviously, thi...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the defects or deficiencies in the above-mentioned prior art, the present invention proposes a medical image fusion method based on dictionary learning and low-rank representation to solve the problems of global structure, local structure capture and detail preservation in the existing medical image fusion, including steps as follows:

Images