A Study On The Detection Of Fog Visibility Based On Image Analysis

With the rapid development of traffic industry,the traffic accidents caused by low visibility of roads are becoming more and more common,which poses a serious threat to people's life and property.In this paper,the fog visibility detection technology based on image analysis is studied.The method of qualitative nighttime fog classification based on street area color feature and the method of quantitative fog visibility detection based on camera calibration results are proposed and implemented.The main contents of this paper include:(1)The research of nighttime fog classification method.Aiming at the elimination of white particle noise,an improved median filtering method is proposed to complete the fast preprocessing of nighttime images.After applying the dark channel and edge detection to the positioning of the light area,we propose the automatic positioning method of the light at night;according to the HIS color model,the luminance attenuation characteristics and saturation characteristics are also proposed for nighttime fog classification.(2)The research of fog visibility detection method.In order to adapt to any shooting angle of the surveillance camera,this paper proposes a camera distance calibration algorithm that is suitable for camera angle offset conditions.It extends the two-dimensional calibration model into a three-dimensional distance calibration model and designs a distance calibration model suitable for camera angle offset conditions.Then,the method for solving the calibration model parameters is obtained;After we finish the camera distance calibration algorithm,this paper study the fog visibility detection method from daytime and nighttime: the daytime visibility detection studies the farthest visual point search method adapt to the deflective roads;the nighttime visibility detection combines the night visibility calculation based on the extinction coefficient and the to the camera distance calibration algorithm.Then,the farthest visual distance detection method based on car taillight.(3)Software design.After clearly defining the methods that need to be implemented and verified,this paper constructs a software-implemented system framework;we perform specific function design and MFC interface design for each function module established,and present the flow charts;the tests ensures the accuracy and stability of the system function,and realizes the study of fog visibility detection based on image analysis.