一种色彩校正算法研究与实现

Design and Implementation of a Color Correction Method

数字图像拼接是指将具有重叠区的多幅数字图像或多帧视频,通过配准和融合获得单幅宽视场图像或者动态全景图。在实际应用中,图像因色彩不一致,而利于图像的配准和融合,因此有必要研究图像色彩的校正。文中首先论述了彩色图像的色彩空间以及它们之间的转换关系,然后介绍了常用的色彩校正方法,最后提出了基于Gamma和Linear调整的色彩校正算法。该算法将图像转换到YCbCr色彩空间,对亮度分量进行Gamma校正,对色度分量进行线性校正,实验结果表明,与以往的校正算法相比,校正效果有一定的改善。

The digital image mosaic is to get the single width view image or dynamic panorama by registering the overlap of digital image and fusing the video frame. In practical application, inconsistence in color of the images may cause many disadvantages in image registration and fusion. Therefore, it is necessary to do some research on the image color correction. In this paper, we first discuss the color space of the color images as well as the transformation of the relationship between them. Then, we introduce some commonly used methods of color correction. Finally, we put forward a color correction algorithm based on Gamma and linear adjustment. This algorithm will convert image to YCb C, color space by using the Gamma correction to correct the luminance components and correcting chromaticity components with linear correction. The experimental results show that the calibration results have certain improvement over current correction algorithms.