Multi-exposure Based High Dynamic Range Color Image Composition

Dynamic Range in Digital Image Processing refers to the ratio of maximal value to minimal value. In sRGB color space, 8-bit integer is used to represent each color channel of traditional low dynamic range, with at most 256 of luminance level. Different from ordinary image, HDR image is designed using float value to fully demonstrate the total range of real world scene, ranging from 1 to 100000000. HDR Image has been widely used in several aspects such as movie, human vision system as well as remote sense.Among the research of high dynamic range composition, grey-scale image composition technique has been researched a lot and is mature, while color image composition is much more complicated and far from being fully researched. This paper focuses on the color performance in high dynamic range image composition based on multi-exposure composition, and experiment results in different color space (RGB, HSV) combined with analysis on performance are given. We proposed a YUV model based algorithm for HDR composition. The experiment results show that, Composition in YUV space works fine in terms of efficiency, SNR, color performance as well as real color.