Apparatus and method for removing image noise

Abstract

The invention provides an image noise elimination device. The image noise elimination device comprises a image noise calculating unit used for performing estimation to the noise variance value of an image, a neighbor generation unit used for establishing the center pixel point neighbor, a variance calculating unit used for estimating the variance value of the center pixel point neighbor, and a comparison and filtering unit used for comparing the noise variance value and the variance value of the center pixel point neighbor and performing noise filtering according to the comparative result. The image noise elimination method comprises the following steps: the image noise variance value is estimated; the noise variance value is compared with the variance value of the center pixel point neighbor, and the noise elimination is performed according to the comparative result. Since the image details are amplified by the method, the image details and the noise can be effectively distinguished,the image details can be retained when the image noise is eliminated, and the definition of the image is not affected.

Description

technical field [0001] The invention relates to the field of image noise removal, in particular to an image noise removal device and method for a camera. Background technique [0002] The basic purpose of image preprocessing is to improve image quality, and removing image noise is an effective way to improve image quality. There are many reasons for the formation of noise, which may be generated during imaging or transmission. There are also many different methods for noise removal in image processing, among which the low-pass filtering method is the most commonly used. [0003] Low-pass filtering is a popular signal processing method, which uses simple neighborhood averaging to reduce noise, that is, replaces the color components of the current center pixel with the average value of the color components of the neighboring pixels. This method can compare Effectively suppress image noise, but there is a large loss of image details, thereby reducing image quality. [0004] ...

AI Technical Summary

Problems solved by technology

Although this method avoids the defect that the previous method performs low-pass filtering on all pixels, resulting in a large amount of detail loss, but this method is difficult to accurately distinguish image details and noise, and a stricter distinction will result in a lot of loss If the image details are wider, the noise will be difficult to remove.

Images