Multimodal Image Fusion Method Based on Staged Multiatom Orthogonal Matching Tracking

Abstract

The invention provides a multimode image fusing method based on grading polyatomic orthogonal matching pursuit. The method comprises the steps of dividing a fused source image into image blocks of same size, and converting each image block into a one-dimensional column vector, so as to form a matrix expression way of the source image; carrying out sparse decomposition on each image block, and then fusing the sparse decomposition coefficients of the image blocks of different source images according to a fusing rule to obtain a fused image block coefficient matrix; and finally rebuilding the fused image block coefficient matrix to generate a final fused image; adopting a polyatomic selection strategy at each stage when the sparse decomposition is carried out to image block signals, and then removing redundant atoms through a self-organizing searching and cutting algorithm, so as to reduce the coherence of a support set and ensure the obtaining of the spare expression of original image blocks more effectively. By adopting the image fusing method provided by the invention, the algorithm implementation speed can be greatly improved, and the quality of the fused images can be improved.

Description

technical field [0001] The invention belongs to the technical field of image fusion, and relates to blind source signal separation, matching tracking, signal staged sparse decomposition representation, and specifically relates to the fields of medical image analysis, recovery, multi-mode image fusion and the like. Background technique [0002] In the field of medical imaging, the continuous emergence of various advanced medical imaging equipment provides a more intuitive basis for clinical diagnosis and has become an indispensable part of modern medical methods. However, different imaging equipment presents different multimodal medical Images, provided images of certain human tissues or structures, often have their own advantages and disadvantages. How to use the advantages of various single-mode medical images to obtain more comprehensive and accurate lesion image information is particularly important for the accuracy of doctors' diagnostic results. Based on this, the medic...

AI Technical Summary

Problems solved by technology

[0012] Among the family of greedy tracking algorithms, the Orthogonal Matching Pursuit (OMP) algorithm is the most widely used. The computational complexity of the OMP algorithm is related to the sparseness of the original signal. If the original signal is not so sparse, then the computational performance of the algorithm will decrease. decline

[0013] At present, the OMP algorithm only allows one atom to enter the OMP algorithm at a time, and the calculation complexity is high, which is not suitable for large-scale multi-modal medical image fusion. Therefore, it is urgent to develop an image fusion method that can help reduce the number of iterations and improve the efficiency of algorithm execution.

Images