Automatic white balance detection system

Abstract

The invention relates to an automatic white balance detection system which adopts unique composition. Characteristics of a camera and characteristics of a black body curve are combined, and an operation of finding out white / gray point-achromatic regions is adopted to replace an existing region algorithm, weight algorithm and the like. Two regions of a warm color light source and a cold color light source are adopted according to light source characteristics; information of R / G and B / G, which is calculated in a changing mode by a variable image partitioning technology, is used as decision elements; a non-closed adjustable mode is adopted for a region of reference points, and a variable maximum and minimum approximate curve mode is determined. According to a detection color temperature range / detector color temperature accuracy, each region is subjected to color temperature calculation; colored regions in color temperature regions are removed and colored regions in brightness are removed, the number of white / gray regions obtained after the colored regions are removed is finally determined, and R and B gains and color temperatures of the determined white / gray achromatic regions are calculated. According to the automatic white balance detection system disclosed by the invention, the problem of influence of light source spectrum characteristics and characteristics of an image sensor, characteristics of an IRCF (Infrared Ray Cut Filter) and characteristics of a lens in a camera module of the image sensor is well solved, and accuracy of white balance is improved.

Description

technical field [0001] The invention relates to an automatic white balance detection system. Background technique [0002] White balance (White balance) is a very important factor in the current digital imaging world, especially in the field of photography and imaging, and has a great influence on the reproducibility of image colors. [0003] The essence of white balance is because of the "Memory color" of the human brain - C.J.Bartleson (1960) memory color. That is, in any lighting environment, or no matter how the lighting changes, it is necessary to correct the "white" seen by the eyes to the white in people's memory color. The eye sees "white" --- it is actually the reflection of the light source by the white object, and the image sensor of the digital camera / camcorder turns the light reflected by the white object into an image. In the absence of white balance processing, the image is a reflection of the real physical quantity and changes due to changes in the light so...

AI Technical Summary

Problems solved by technology

In the 11-2 interval, the upper and lower areas of the D65 reference point are continuous, and it is difficult to distinguish these continuous areas corresponding to white and other colors in the right picture

Although a "correct" judgment can be made by setting limited parameters around the white reference point in R / G and B / G coordinates or adjusting weights such as distance, this "correct" judgment is temporary and unstable. The settings will be quite complicated and need to be debugged repeatedly, seriously affecting the efficiency

Moreover, the light source will change all the time, and every time the scene changes, it will affect the debugging effect again

Images