Automatic color correction matrix algorithm based on the least square method

Abstract

The invention provides an automatic color correction matrix algorithm based on the least square method. The algorithm performs sampling statistics on two images with differences, outputs respective standard color correction matrixes, and fits a polynomial matrix satisfying the difference of sensors according to the basic principle of the least square method. The problem of visual error caused by insufficient precision of the sensor, the optical filter, the lens and other hardware, and such factors as sensitometric curve discreteness is solved.

Description

technical field [0001] The invention belongs to the field of image processing, in particular to an automatic color correction matrix algorithm based on the least square method. Background technique [0002] With the in-depth development of image fusion technology, more and more fields have begun to use image fusion technology, but due to the lack of hardware precision such as sensors, filters, lenses and other factors such as the discreteness of the sensitivity curve, the spliced ​​image and There will be differences in color and brightness between images, which will affect the image effect of the entire splicing and fusion. [0003] The existing color correction matrix is ​​to restore the deviation between the RGB components and the human eye's response to the spectrum due to the corresponding difference of the CMOS and other image sensors for the spectrum. Usually, a color correction matrix is ​​used to correct the cross-response and response intensity of the spectral res...

AI Technical Summary

Problems solved by technology

However, there are problems such as visual errors caused by factors such as insufficient hardware precision of sensors, optical filters, and lenses, and discreteness of the sensitivity curve.

Images