Stereoscopic video color correction based on decorrelation color space

Abstract

In the production of stereoscopic video, there are some color differences between different views. In this paper, a color correction method based on decorrelation color space is proposed. Image features from the reference image and original image are first extracted by Harris algorithm, and Normalized Cross-Correlation algorithm (NCC) is used to find the corresponding match points of the images. Then the RGB color space is converted into decorrelation color space in which colors are corrected with a formula which includes mean and variance. After that, images are converted back into the RGB color space. Experiment results show that the proposed method can produce a better correction result than histogram match.