Image white balance adjustment method and device

Abstract

The invention discloses an image white balance adjustment method and device, wherein the method includes the following steps: processing the image collected by an optical sensor to generate a corresponding RGB image; The components respectively mark the first-class reference white point and the second-class reference white point of the RGB image, wherein the first-class reference white point includes the second-class reference white point; according to the preset conditions, select the first-class reference white point or the second-class reference white point as the RGB The reference white point of the image; adjust the white balance of the RGB image according to the selected reference white point. This method can more accurately select the reference white point when the image is automatically adjusted for white balance, so that the color temperature can be accurately judged, the white balance adjustment is better realized, and the repetition of debugging is reduced, and it can be done when the color temperature is frequently switched. To more rapid judgment and adjustment.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image white balance adjustment method and an image white balance adjustment device. Background technique [0002] As the color temperature changes, the spectral components also change accordingly. Due to the special adaptability of the human eye, we sometimes do not feel this change; but the optical sensor is different, it will appear a difference according to the change of the spectral composition, which leads to the phenomenon of color imbalance. An algorithm that corrects for this phenomenon is called white balance. [0003] Different optical sensors have different performances on color temperature changes. For Bayer-type optical sensors, the main performance is the difference in the ratio of R, G, and B values. Therefore, the algorithm adopted for such devices is to adjust the ratio of R, G, and B values ​​to make it equal to or close to 1:1:1. Wherein, the bas...

AI Technical Summary

Problems solved by technology

However, for some special scenes where the reference white point cannot be accurately judged, the adjustment coefficient will vary greatly, making the image unable to accurately achieve white balance

For example, some scenes that cover light-colored scenes and skin-colored scenes or scenes in low-light environments are prone to inaccurate corrections, and may even shift the original normal scene, which leads to misadjustment.

Images