【摘要】 针对多视点视频系统中视点间图像颜色差异的问题,提出了一种基于SIFT特征提取的多视点颜色校正算法。该方法首先通过SIFT算法提取参考图像和待校正图像的特征点;然后根据匹配算法得到对应的匹配点;再通过匹配点的特征得到R、G、B分量的校正因子,并校正得到最终序列。校正在RGB颜色空间中完成,需要进行YUV和RGB之间的转换。实验结果表明,本文方法对校正图像的主观质量和客观质量均有提高,尤其是对纹理细节比较多的序列效果更明显,相比于偏色校正方法,PSNR提高了0.96dB。更多还原

【Abstract】 In order to solve the non-consistence color appearance between different viewpoint images,a multi-view color correction based on SIFT algorithm is proposed.In this paper,firstly feature points of all images are extracted by SIFT algorithm.Secondly,corresponding matching points are acquired by match algorithm.Finally,the color correction coefficients of R,G and B channels are computed by the matching points’ features,and the sequence can be corrected.The correction is complete in RGB color space,and conversion between YUV and RGB is needed.The experimental results have shown that the proposed color correction method improve both subjective quality and objective quality,especially for more texture details.Compared with the method of color cast correction,PSNR has gained 0.96 dB.更多还原