图像校正中颜色模型的选取方法

Selection of Color Models in Image Correction

在全景图像的拼接中，各子图间颜色不匹配会严重影响全景图像的效果，甚至会产生错误的拼接。选取一种基于统计学的图像颜色校正方法，然后对各种不同失真情况的图像分别在RGB、YUV、HSV几个颜色模型中进行图像颜色校正，并对比所得的实验结果；讨论了针对不同的失真情况，如何选取合适的颜色模型，使图像校正取得更好的效果。结果表明选取合适的颜色空间校正图像，可以显著地提高所生成全景图像的质量。

When the adjacent images are stitched as a panoramic image, the color distorts between the adjacent images would make the panoramic worse. In this paper, we use a color correction method based on statistics, then compare the images correction results in RGB,YUV, HSV color space. At the end of this paper, we discuss how to select the proper color model in image color correction. The results show that correcting the color of images in proper color model can significantly improve the quality of panoramic images.