Research And Implementation Of Image And Video Dehazing Algorithm

The details of images and videos which captured outdoors are affected by the haze weather conditions.Owing to tiny particles in the air absorb and scatter the light,the way of light is changed,which has some influences of the images and videos taken outdoor.Contrast reduction and color distortion have a great influence on image processing operations such as image segmentation,target recognition,and behavior detection.Many researchers attempt to eliminate or reduce the effects of hazy weather on outdoor imaging systems.Some image dehazing algorithms are used to recover the details of images and videos,which is good at our living and production.It is of great theoretical and practical value to study the images and videos dehazing algorithm.Fistly,the research progress of image and video dehazing at home and abroad based on physical model and non-physical model are studied.At the same time,the degrading mechanism of haze images and the evaluation indicators of dehazing algorithm are also analyzed.Sceondly,the dehazing algorithm of dark channel prior was studied in depth,which is one of the most popular algorithm in the image dehazing algorithm field.Aiming at the shortcoming of this algorithm,two improvements are proposed.For one thing,the dark channel map is processed by guided filtering to expect the correct position of the atmospheric light in the image.For another thing,using piecewise function to represent the transmittance of image,which can handle the sky area much better.The experimental results show that the improved algorithm of this paper is better than original algorithm.At the same time,compared with other dehazing algorithms in recent years,the advantages of the improved algorithm in this paper are obvious.Thirdly,the improved algorithm is used for video dehazing.Based on the video's own characteristics and the correlation between the frames and frames,acceleration method of video dehazing algorithm in a fixed background was studied.The same atmospheric light value and universal transmittance method are used to process the foggy video.At the same time,bicubic interpolation algorithm is used to scale the video frames to improve the speed.The results show that the processing speed reaches 0.06 seconds/frame for a size 1280 ?720 video image.The results also show that the video dehazing algorithm proposed in this paper are superior to other defogging algorithms in visual effects,and the objective performance indicators are higher than the dark channel prior dehazing algorithm.Finally,an image and video dehazing system is designed and implemented based on the GUI programming of Matlab.The system is an executable program that can work well on other computers.It has the features of friendly system interface,simple operation and complete functions.It can deal with the foggy images and videos,and can display and save the results.