基于形态学变换和FFCM聚类的灰度图像颜色迁移算法

Colour transfer for grayscale images based on morphology transformation and FFCM cluster

针对传统颜色迁移算法计算量大、对图像无法准确进行颜色迁移的问题,提出一种基于形态学变换和快速模糊C均值聚类(FFCM)的灰度图像颜色迁移算法。首先对目标图像进行腐蚀膨胀运算,消除亮度不均匀的区域,通过FFCM聚类算法对目标图像进行准确聚类,然后在目标图像与源图像中选取对应样本块,完成样本块的颜色迁移,并以已上色的样本块为参考,完成图像的全局颜色迁移。实验结果表明:与Welsh和FCM算法相比较,本文算法处理时间分别缩短64.29%和54.25%,结果图像在类间交界处的颜色过渡更加自然,证明了算法的有效性。

To deal with the problem that the traditional color transfer algorithms can not trans- fer the image colour accurately and have large amount calculation, we proposed a colour trans- fer algorithm for colorizing grayscale images based on morphology transformation and fast fuzz- y C-means （FFCM） cluster. We used morphology transformation to process the target image for removing the non-uniform luminance region first, and adopted the FFCM algorithm to clus- ter the target image accurately. Then we established the correspondence between the swatches of source and target images to complete the color transfer of swatches which were as the refer- ences to complete the target image＇s colorization. The experimental results showed that, com- pared with Welsh and fuzzy C-means （FCM）, the elapsed time decreased by 64. 2~ and 54.25~ respectively and the result image was more natural in clusters＇ junction which proved the validity of our algorithm proposed.