Colorizing Biomedical Images Based on Color Transfer
Yuanmeng Zhao, Lingxue Wang, Weiqi Jin, Shiming Shi · 2007
Biomedical images with single color are dyed usually in the means of manual pigmentation. In this paper, we develop a novel interdisciplinary method, which aims to adding colors to the grayscale biomedical images more efficiently and effectively than ever before. To achieve this goal, we use computer processing within imaging pipeline. More specifically, we need to select a most appropriate chromatic image for reference at first. In the next step, both of the grayscale and chromatic images are transformed into a decorrelated color space (Ialphabeta space). Third, a best matching reference pixel is determined for each pixel of the grayscale image, and then chromaticity values are transferred from the best matching reference pixel to corresponding grayscale pixel. By doing so, we can obtain resulting biomedical images with natural and sightly color appearance which looks like the reference images', and meanwhile the whole procedure is quite convenient and efficient.