Image denoising method based on Contourlet transformation self-adaptation direction threshold value

Abstract

The invention discloses an image denoising method based on a Contourlet transformation self-adaptation direction threshold value, and belongs to the technical field of images. An elliptical direction neighborhood consistent with a sub-band energy gathering direction is adopted to calculate the self-adaptation direction threshold value. The method comprises the steps that (1), an image with noise I is input; (2), Contourlet transformation is carried out on the I; (3), local signal direction energy of a coefficient is estimated through the elliptical direction neighborhood; (4), a direction energy factor of the coefficient is calculated according to the local signal direction energy and dimension signal average energy; (5), based on the Bayes criterion, a Bayes threshold value is corrected through the direction energy factor to obtain the self-adaptation direction threshold value; (6), based on soft threshold value rules, threshold value shrinkage is carried out on the coefficient with noise through the self-adaptation direction threshold value; (7), Contourlet inverse transformation is carried out through the shrunk coefficient to obtain a denoised image. The method can effectively remove the noise, and clear image details can be kept.

Description

(1) Technical field [0001] The invention relates to the technical field of image processing, in particular to an image denoising method based on Contourlet transform adaptive direction threshold. (2) Background technology [0002] In today's information age, digital images have become an important data source for computer processing. In reality, digital images are often affected by imaging equipment and external environmental noise interference during digitization and transmission. For example, medical images, microscopic images, remote sensing images, and aerial images all contain strong noise. For the accuracy of subsequent target recognition and analysis, a large number of images need to be denoised, which increases the urgent need for image denoising technology. [0003] The so-called image denoising refers to a basic operation that reduces the noise level in a digital image and maintains image detail information as much as possible. As a basic image preprocessing pro...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to provide an image denoising method based on Contourlet transform adaptive direction threshold, to improve the artificial artifacts such as blurred details and artifacts that are prone to occur when the existing method is used for image denoising, to achieve the removal of noise, edge and the purpose of keeping the texture clear

Images