The Research On Night Video Enhancement Algorithm And Hardware Design

Video technology is one of the most widely used technologies in today's society,which plays an important role in security monitoring,traffic safety,aerospace and other fields.However,in the nighttime environment,due to low light or uneven illumination,the quality of night video is poor and with little detail.So it is necessary to take enhancement on night video for improving the video visual effects and quality.Therefore,this thesis focus on the research of the night video enhancement algorithm,and the hardware design and implementation of embedded night video enhancement system.Firstly,the classical image enhancement algorithm and its application in night image enhancement are studied.Considering the characteristics of night video,the dissertation focuses on the application of histogram equalization;holomomorphic filter and Retinex algorithm in night image,and analyzes the performances of each algorithm by experiments.Secondly,the nighttime image enhancement based on the dark channel prior de-fog algorithm is studied.After comparison and experimental verification,it is found that the histogram distribution of nighttime image and fog image presents a symmetrical characteristic,so the purpose of night image enhancement can be realized by using image de-fog enhancement algorithm.The thesis analyzes the performance of the algorithm by simulation experiments.The problem is found that the dark channel prior algorithm causes halo artifact and requires a lot of computation.To solve this problem,the thesis presents the optimized dark channel prior algorithm,and compared visual effects and objective evaluation index of the optimized algorithm and the other algorithm by experiments.Finally,based on the research of algorithm,the night video enhancement hardware system is designed and implemented.Considering the cost,volume and practicality of hardware system,the night video enhancement system is designed based on Field Programmable Gate Array(FPGA).The thesis introduces the circuit diagram design of the video processing board,and then elaborates the design and implementation of the FPGA functional logic based on Verilog hardware programming language.The modularity and pipeline design are used to realize the FPGA logic modules,including video data reception and send module,color space conversion module and video enhancement processing module.In the end,the night video enhancement system is tested in the actual nighttime environment.The test results show that the embedded night video enhancement system is able to implements the real-time and efficient enhancement of high-definition night video with the resolution of 1920x1080 and frame rate of 30 fps.