Low Illumination Image Color Estimation Based on Gray Scale Supervision

Aiming at the problem of low illumination image color estimation, a method based on white-grey-black block supervision is proposed. The standard swatch is placed in a low illuminance scene to obtain an image, and the white balance function and the three color gradation transformation function are obtained by supervising the white-gray-black block red, green and blue color values of the grayscale decreasing for the low illuminance image. Object color estimation. The experimental results show that subjective vision can accurately determine the color of the object, and the objective index is better evaluated.