Study Of Image Enhancement Methods For CAPF Dealing With Contigencies

Image enhancement is to make the degraded images clearer or to emphasize some characteristics of the images, so as to improve the visual effects or be convenient for other processes. In view of techniques, image enhancement is managed to discard the unnecessary and disturbing information from the images, but highlight or keep necessary information.The purpose of the paper is to meet the actual necessity for the armed police to execute multiple-missions. The basic techniques used in image enhancement are introduced and some representative theories and algorithms just proposed recently have been analyzed. Then image enhancements based on bilateral filtering and image fusion are researched deeply. After understanding the essence of the algorithms, simulations are made according to the relative algorithms, accompanied with low-quality gray images, fog-degraded images and images under complex luminance. Various algorithms are evaluated from both subjective and objective criteria in order to obtain proper image enhancement algorithms suitable for armed police executing multiple-missions.Specifically, the research work of this article mainly includes the following five aspects:1) Conventional image enhancement techniques are analyzed systematically, including some typical methods based on frequency domain and space domain, and some representative image enhancement algorithms in recent years;2) Some basic concepts are expounded systematically, and some basic theories are analyzed deeply, based on the imaging mechanism, digitalized process, mathematical description and basic feature of image;3) The evaluation theories and methods of image enhancement algorithms are explained in detail, especially the analysis of the complexity of algorithm and the evaluation of the effect of image enhancement;4) After the study of the fusion between infrared image and homologous visible image, a method for image enhancement based on bilateral filtering and image fusion is proposed. To evaluate the performance of the method compared with other methods, a large number of simulation experiments are conducted, and related data is collected as well;5) An improved POSHE algorithm for fog image enhancement is proposed. On the basis of POSHE algorithm, by improving the balanced way of each sub image, a novel image histogram algorithm partly overlapped is obtained, which reduces the calculation amount, and meets the requirements of real time image processing.Experimental results show that the enhancement algorithm based on bilateral filtering and image fusion has a good effect for the fusion of infrared image and visible image. The treated image can effectively eliminate noise and highlight the human body, especially the rear condition filtering algorithm, whose processing efficiency is higher, which can detect the human body target hidden in the image accurately, and can meet the needs of the conventional batch processing; compared with the traditional image enhancement methods of fog drop scene quality, the improved POSHE algorithm can not only avoid the block effect after image processing, but also significantly reduce the calculation amount, and the effect of the image detail enhancement is significantly as well.