Adaptive Nighttime Image Enhancement Algorithm Based On FPGA

With continuous innovation and development of technology, video systems as an important source of information and exploration of human knowledge has been around in various disciplines and fields. However, image form video information systems have impatient large, low contrast, gray dim and gray dynamic range contraction and other shortcomings in low-light environments. Through the introduction of image enhancement technology, the video information can be restored and improved. Specific targeted image enhancement processing, can effectively improve the quality of the video image, facilitate observation with the human eye and can be used to extract information. But the real-time video image processing has demanded large amount of information, signal conversion timing characteristics of complex hardware video systems, and it is difficult to rely solely on software to meet the real-time processing of video images. That FPGA Field Programmable Gate Arrays, parallel computing, mainly in terms of speed and performance are unmatched MCU and DSP. As a user programmable digital integrated circuits, high stability and low cost, it makes the image processing being used widely because of its speed, strong performance.Paper studies the theoretical basis of image enhancement algorithms, through the principle of Retinex and Gamma correction has been understanding and explored in-depth, considers and analyzes separately for each image algorithm and FPGA features systematically, in order to design a real-time nighttime image enhancement system which is based on the luminance evaluation of Retinex under adaptive Gamma Correction.The traditional image enhancement system can not meet special scenarios which were too dark or the uneven light brightens. Unlike the traditional image enhancement system, the algorithm used in this paper have adaptive capacity. For the actual scene of different brightness, the brightness of the algorithm first assessed by Retinex principle, according to the brightness of assessment to self-regulating manner Gamma parameters of image pixels to achieve efficient Gamma Correction enhancement, and ultimately to significantly improve image readability goals.In the process of implementation, since there are some limitations in processing exponential power of the Gamma correction of FPGA. In this paper, we consume at the expense of FPGA internal resources to make the lookup table secondary mapping which replaces the gamma index of arbitration and index exponentiation based Gamma Correction, in order to optimize the image processing module deeply which is designed to improve the overall speed of the system operation. Finally.we verify the real-time image processing system can effectively improve the video image quality at night,by means of FPGA-based hardware development platform in order to achieve overall system simulation, debugging and running.