Digital Imaging Exposure Dynamic Range Expansion Algorithm And Applied Research

Image information is one of the most important branches in the domain of information engineering. It becomes one of the hotspots in the related research areas. The concept of extending the dynamic range of an image is to enlarge the range of recording light intensity while preserving as much detail information as possible in the target image by means of hardware or software. The research works on dynamic range extending have great potentials both technically and commercially. It is especially valuable in the situation where detail information has to be obtained in an image of the high contrast light condition. This technique is also very important and useful in the field of aerial photography, military scout and resource detection.This paper summarizes the academic and experimental research accomplishments on extending dynamic range of digital image. Linear and nonlinear fusion methods on single-image and multi-images are proposed and analyzed. The extending strategies of exposure dynamic range are studied, and the experimental results are presented to validate the proposed method. Detail research works can be described as following: (1) Based on the human visual identify theory, an effective linear algorithm to deal with the actual images is proposed. The methods to apply this linear algorithm on single-image and multi-images are studied in detail. (2) The theory of logarithm image process, one of the nonlinear algorithm for dynamic range extending, is worked over. Experiments have been performed to study the best parameter selections in applications. The significance of using edge detection technique in dynamic range extending processing is also studied and discussed. (3) The theories of human visual system and the human visual basic characteristics are introduced into the field of dynamic range extending, in order to give the direction of the image processing. The subjective and objective image quality assessment ways are also analyzed and synthesized. (4) All the proposed algorithms are put into the hardware simulation and software realization, and the quality assessment system is used to evaluate the output images. The experimental results are given and full discussions on the theory and method are performed. In the end of this paper, conclusions are summarized, and prospective views are profiled.