Abstract

Color cast detection is an important part of image quality assessment, which can detect and classify the distortion of the image color in order to correct color cast pointedly. This paper proposes an improved image color cast detection algorithm. The improved algorithm introduces the color cast feature parameter in the height direction of the hue histogram and color cast feature variable quantity in brightness channel into the 2D Lab histogram method. Meanwhile, it makes a more perfect color deviation image classification and optimizes algorithm flow. Compared with the 2D Lab histogram method, the improved algorithm can get higher detection accuracy and efficiency.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call