Color image denoising method and system thereof

Abstract

The present invention relates to the field of video processing, and disclosed are a denoising method for a color image and a system thereof. In the present invention, cross-channel unchanged information is extracted from a luminance channel image, and used for denoising processing of a chrominance channel image, so as to eliminate false color edges or texture, and reduce blurs at edges, thereby obtaining a better overall denoising effect. The cross-channel unchanged information may be edge information, texture information, noise intensity, and so on. Local data transformation may be performed on the luminance channel image first, the cross-channel unchanged information is extracted in a transform domain, and feature information such as edge texture of local image data may be extracted more effectively by using different characteristics demonstrated by a signal in the transform domain, thereby better performing the denoising processing on the chrominance channel image by using the information.

Description

technical field [0001] The invention relates to the field of video processing, in particular to color image denoising technology. Background technique [0002] Video image denoising (noise reduction) is an important content in the field of video processing. Since the image will inevitably be affected by the interference of the transmission medium, external and environmental optical signals, electrical signals, and mechanical damage during the recording, data compression, storage, and transmission processes, when the video image reaches the display terminal, the video The image content carried by the signal (brightness value and color value of each pixel) changes, and these changes are called image noise because of their spatial and temporal randomness. [0003] Traditional color image denoising methods generally consider that a color image is composed of three independent grayscale images. Therefore, traditional color image denoising methods generally use the same denoisin...

AI Technical Summary

Problems solved by technology

[0004] However, the denoising results of traditional color image denoising methods are often unsatisfactory, and there are shortcomings such as residual color noise, false color edges or textures, edge blurring, etc.

The inventors of the present invention believe that the reason is that color images are generally composed of complementary metal-oxide semiconductor (Complementary Metal-Oxide Semiconductor, "CMOS") or Charge Coupled Device (Charge Coupled Device, "CCD") sensors. Acquisition, there is a great correlation between the RGB or YUV three channels in the color image, and it also shows different noise characteristics. The traditional color image denoising method lacks consideration of these characteristics of the real color image

Images