Enhanced low-rank sparse decomposition model medical CT image denoising method

Abstract

The invention relates to an enhanced low-rank sparse decomposition model medical CT image denoising method. The method comprises the following steps: determining the number and the size of similar blocks in a search window and an image block matrix and the maximum number of iterations during iterative solution according to a calculated noise intensity estimated value, traversing an original image,performing non-local similar block matching, and dividing the original image into a plurality of image block matrixes consisting of non-local similar blocks; carrying out low-rank matrix estimation on the medical CT original image D belonging to Rm*n by adopting a weighted Schatten p norm, and adding a joint constraint L1-2TV regularization item to construct an enhanced low-rank sparse decomposition model; sequentially inputting the image block matrixes into the model, and performing iterative solution by using an alternating direction multiplier method to obtain low-rank matrixes of the corresponding image block matrixes; and aggregating the low-rank matrixes corresponding to all the image block matrixes to obtain a denoised clean image. According to the method, more mixed noise can be separated as much as possible so as to obtain a better medical CT image denoising effect.

Description

technical field [0001] The invention belongs to the field of medical image denoising, in particular to an enhanced low-rank sparse decomposition model medical CT image denoising method, the method mainly uses the weighted Schatten p norm and L 1-2 TV regularization constraints to denoise medical CT images. Background technique [0002] Image noise is the main factor that hinders information understanding and analysis, and image denoising has long been concerned. Computed Tomography (CT) images are one of the important image data of computer-aided medical care. CT images carry a large amount of clinical diagnosis and treatment information, which can effectively assist doctors in disease diagnosis, surgical planning, and postoperative treatment evaluation. However, due to the influence of environment, equipment and other factors in the process of acquisition, compression and transmission of medical CT images, the image signal will inevitably be polluted by noise and the quali...

AI Technical Summary

Problems solved by technology

However, the approximate solution of the minimum rank problem using the nuclear norm is often a suboptimal solution to the rank function minimization problem, which makes the reconstruction effect of the denoised image poor.

In addition, as the noise intensity increases, the image denoised by the RPCA method will often appear the "oil painting" phenomenon of stepped image boundaries, resulting in serious consequences of loss of image edge information

Images