Improved canny edge detection based non-local means MRI (magnetic resonance image) denoising method

Abstract

The invention discloses an improved canny edge detection based non-local means MRI (magnetic resonance image) denoising method. The method includes: (1) subjecting an MRI to amplitude square processing to obtain an amplitude square image; (2) performing noise estimation; (3) adopting a canny operator to extract edges of the amplitude square image; (4) computing edge / intensity similarity measure between the edges of two similar blocks; (5) computing weight parameters; (6) processing via a non-local means method; (7) correcting deviation. The method has the advantages that due to the fact that edge similarities between the similar blocks are taken into consideration at the same time, the improved canny edge detection technology is adopted, drawbacks caused by the fact that the distance between the similar blocks is subjected to weight parameter computation by only relaying on single parameter of pixel intensity in a traditional NLM method are overcome, similarity measure is more accurate, and further effect of image denoising is improved; non-local means filter replaces Gaussian filter, the edges are well kept, part of noise is removed, and edge detection is more accurate.

Description

technical field [0001] The invention relates to an image denoising method, in particular to a non-local mean MRI image denoising method based on improved canny edge detection, and belongs to the technical field of image processing. Background technique [0002] In recent years, with the development of magnetic resonance imaging (magnetic resonance images, MRI) technology, the ability of human beings to study the brain has been unprecedentedly improved, and at the same time a large number of MRI images have been produced. The acquisition process of MRI images is often disturbed by different noises. It is found that the noise in MRI images mainly presents Rician distribution. [0003] At present, MRI images have been widely used in medical diagnosis, and there are many corresponding processing techniques for the noise existing in MRI images, among which the anisotropic diffusion denoising method, total variation denoising method, non-local mean denoising, etc. [0004] The c...

AI Technical Summary

Problems solved by technology

When the image is polluted by strong noise, the gradient caused by the noise is likely to be greater than the gradient of the edge of the image detail. At this time, the model cannot correctly distinguish the noise and the edge of the detail, so it cannot remove the noise well.

[0005] The total variation denoising method finds an equilibrium state between the TV norm and the loyalty item of the image, that is, the minimum value of the energy functional, but when the parameter λ in the energy functional is small, images such as textures are small detail features will be destroyed

[0006] The non-local mean denoising method has the following disadvantages: (1) The complexity of the algorithm is large and the calculation time is long; (2) The similarity measurement function of the similar window only considers the intensity similarity of pixels, and the accuracy of pixel weight distribution needs to be improved ; (3) The selection of some parameters is too sensitive and lacks perfect theoretical guidance; (4) There are deviations in non-local mean denoising, which should be corrected in practical applications

Images