Image noise reduction method and device

Abstract

The invention is applicable to the technical field of image processing, and provides an image noise reduction method and a device. The method comprises steps: a pixel point set is extracted from an input image, a preset parameter value k of a pixel in the pixel point set is no less than 1 and no more than K, and the pixel (x, y) in the pixel point set is marked to be (xk, yk); based on the input image, K first images Ik(x, y) are generated; based on the input image, K second images Ck(x, y) are generated; first filter processing and second filter processing are respectively carried out on each first image, and an image after noise reduction processing is obtained. According to the image noise reduction scheme provided by the invention, the image can be segmented according to attributes such as a brightness value of the image pixel, computational complexity of the overall scheme is reduced, and the software cost and the hardware cost for scheme realization are greatly reduced. While the image noise is reduced, edges of an object in a fuzzy image can be avoided.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a method and device for image noise reduction. Background technique [0002] During the process of digitization and transmission, the image is easily affected by the interference of imaging equipment and external environmental noise, which makes the picture quality of the image degrade. Therefore, in the process of image processing, image noise reduction is particularly important. Among them, smoothing filter is the key to One of the commonly used methods for image noise reduction, it can average the pixel values ​​in a certain area of ​​the image, but it will also cause the loss of image edge information, making the output image blurred. [0003] In the prior art, nonlinear filters such as Bilateral and Guided can better solve the problem of blurred edges generated in the smoothing filtering process. However, such nonlinear filters are complex in operation and difficult...

AI Technical Summary

Problems solved by technology

[0002] During the process of digitization and transmission, the image is easily affected by the interference of imaging equipment and external environmental noise, which makes the picture quality of the image degrade. Therefore, in the process of image processing, image noise reduction is particularly important. Among them, smoothing filter is the key to One of the commonly used methods for image noise reduction, which can average the pixel values ​​​​in a certain area of ​​​​the image, but it will also cause the loss of image edge information, making the output image blurred

[0003] In the prior art, nonlinear filters such as Bilateral and Guided can better solve the problem of blurred edges generated in the smoothing filtering process. However, such nonlinear filters are complex in operation and difficult to implement, which virtually improves the Software and hardware cost of image noise reduction

Images