Study On System Design And Algorithm Of Video Dehazing System Based On FPGA

In recent years,the industrial technology of China has been rapidly increasing,resulting in increasingly serious pollution and increasing fog and haze.In foggy conditions,visibility in the atmosphere is reduced and images and video taken by outdoor vision equipment are severely degraded.With the rapid popular ization of smart city,there are more and more monitoring equipments in the streets of the city.So,it is of great realistic value and signif icance to study the real-time defogging of video.With the development of microelectronic technology,the research on video defogging has been no longer confined to PC.In this dissertation,the defogging algorithm and FPGA design have been studied in depth,and a video de-fogging system based on FPGA has been designed.First of all,this article has analyzed the signif icance and value of defogging research,has introduced the current research situation in the field at home and abroad.In view of the demand of de-fogging system and the characteristics of various core processors,this paper has selected FPGA as the core processing unit of the system and designed the architecture of video de-fogging system.And then,this paper has also analysed and designed video acquisition,frame storage,video display module and other subsystems in detail.Secondly,this paper has emphasisly analysed the dark channel prior theory and its advantages and disadvantages.Then,some methods have been comprehensively applied to improve the algor ithm for the design goals of this paper.The complexity of the algorithm has been reduced under the premise of ensuring the de-fogging effect.On this basis,in order to achieve real-time video defogging and eliminate flicker,the calculation of atmospheric light and dark channel have been optimized.When calculating the atmospheric light,according to the correlation between adjacent frames in the video,the atmospheric light of a frame has been extended to several frames thereafter.And from the perspective of ensuring the display effect,the condition that atmospheric light needs to be recalculated has been determined to solve the sudden change of atmospheric light.And a calculation method of atmospheric light has been proposed.By calculating the maximum,minimum,and average gray of all pixels in the dark channel image,have found the threshold,and then have filtered the brightest 0.1% of the pixels through the threshold,thereby avoiding the sorting operation and improving the efficiency of the algorithm.In the process of calculating dark channel images,the dark channel image of one frame has been extended to the next several frames according to the correlation between adjacent frames in the video.When performing de-fogging calculation on a certain frame,whether it is necessary to recalculate the dark channel image is judged according to the relationship between its pixel gray level and the original dark channel image.If no recalculation is needed,the original dark channel image is corrected according to the new frame.This method has reduced the amount of calculation greatly on the premise of ensuring the effect of defogging.Finally,this paper has implemented the designed video de-fogging algorithm on the FPGA-based platform,including solving dark channel diagrams,division calculation optimization and enhancing images.In summary,this article has designed hardware platform and software algorithms of de-fogging system.Through theoretical analys is and experimental verif ication,it can be proved that the video de-fogging system designed in this paper has good de-fogging effect and meets the real-time requirements.