Color image noise channel extraction method based on bounding box

Abstract

The invention belongs to an image processing technology, and particularly relates to a color image noise channel extraction method based on a bounding box. The method comprises the following steps of: selecting a representative noise pixel region from an image interface, and obtaining a set (key patch) of the selected noise pixel points; computing a transformation matrix M of a minimum cuboid bounding box of the key patch in a RGB (Red, Green, Blue) three-dimensional color space; obtaining an axis of the longest edge of three-dimensional edges of the cuboid bounding box, and taking the axis as a noise characteristic channel of the set of the noise pixels distributed in the color space; removing noise in the noise characteristic channel; and carrying out inverse transformation matrix operation on the noise characteristic channel together with two unprocessed channels corresponding to the other two axes of the bounding box, thereby obtaining a noise-removed image in the RGB space. Compared with the prior art, the method provided by the invention has a better noise-removing effect and is easier to optimize.

Description

technical field [0001] The invention belongs to image processing technology, in particular to a method for extracting noise channels of color images based on bounding boxes. Background technique [0002] In image / video processing software, it is a very wide demand to remove the noise existing in the image. There are many reasons for the formation of image noise, which may be generated during the imaging process or during the transmission process. There are also many different methods for removing noise in image processing. [0003] For common RGB color images, commonly used denoising algorithms are performed in the R, G, and B channels. RGB color images are obtained by changing the three color channels of red (R), green (G), and blue (B). And their mutual superposition to get a variety of colors. However, due to the variability of the distribution and shape of different image pixel sets in the color space, it is difficult to effectively represent the noise in the image wit...

AI Technical Summary

Problems solved by technology

However, due to the variability of the distribution and shape of different image pixel sets in the color space, it is difficult to effectively represent the noise in the image with a fixed RGB channel, and it is also difficult for the corresponding denoising algorithm to effectively remove the noise in the image.

Images