在多角度图像拼接中利用查找表对图像进行分段线性颜色校正
Piece-wise Linear Color Correction in Multiview Image Stitch with a Look-up Table Technique
颜色校校正被广泛应用在图像拼接中,通常是将目标图像的颜色变换到参考图像中。由于通过不同传感器获得的图像或者是在不同时刻拍摄的图像通常有不同的相机参数,例如曝光时间等,导致图像间的颜色差异较大。为了对图像进行进一步的处理,减少图像拼接中两图像间的颜色差异。本文在图像拼接中配准结果的基础上,对两幅图像间的重合区域建立YUV通道上的颜色查找表,并以该查找表对作为两幅图像间的颜色映射关系,变换目标图像的颜色。最后,本文从颜色相似度相似度与结构相似度两个方面将该方法与其余两种方法进行比较,实验结果证明,本文提出的方法能大幅度减少两幅图像间的差异,在颜色相似性上更优于其余两种颜色校正的算法。
olor correction, which is widely used in image mosaic, is to transfer the color of a source image to that of a target image. Images are acquired by different sensors usually have different colors with different parameters such as exposure time. Thus, in order to reduce differences between the source image and the target image to have a further image process, we proposed a look-up table method to calculate the color mapping and evaluate this method with color similarity and structure similarity. Experimental results show that the proposed look-up table method could reduce the color difference between two images effectively and have a better performance compared with other method in color similarity.
李永、余杭
电子技术应用
计算机图像处理颜色校正查找表色彩相似度结构相似度
omputer image processingColor correctionlook-up tablecolor similaritystructure similarity
李永,余杭.在多角度图像拼接中利用查找表对图像进行分段线性颜色校正[EB/OL].(2016-12-02)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/201612-41.点此复制
评论