Research On Construction And Visualization Of High Dynamic Range Image

High Dynamic Range Image (HDRI) is a kind of image that represents accurately the wide range of luminosity levels in real scenes. Now, HDRI is widely used in many domains, such as digital photography, image-based illumination technologies, image-based modeling, and virtual reality. However, two problems prevent its development in application. First, the specialized hardware which produce HDRI is too expensive for everyday use; Second, most display devices just support Low Dynamic Range (LDR) output. We can solve these issues through software synthesis and tone reproduction. The technology of tone reproduction is used to compress the dynamic range of HDRI without losing image quality. This paper studies the synthesis and tone reproduction technology.In this thesis, first, we summarize the situation of the technology of generating HDRI. Up to now, there are algorithms can fuse the multiple photographs into a single HDRI, one of which reconstructs HDRI through recovering response curve of camera imaging system. It is possible to get HDRI through common digital camera using this algorithm. We implement this algorithm, make researches on effects of several important parameters to the synthesis results, and get a synthesis method for images missing the information of exposure time. We improve the automation of algorithm through random sampling; then, we analyze the HDRI tone reproduction algorithms, which can be divided into two categories: spatially uniform and spatially varying. The first kind of algorithm has the advantage of easy and quick. An improved algorithm is presented based on the existed algorithms in this paper. We use this method to process HDRI and get good performance.