Research Of Image Enhancement Method Under Severe Weather For Intelligent Transportation

Intelligent transportation system to improve the utilization of roads to ensure,the safety of traffic and improve the driver's comfort,it has become the future direction of urban transport development.Vehicle monitoring system and traffic monitoring system have been a major component of intelligent transportation system,the system's safety and stability have a higher demand,but the outdoor weather is complex,the fog,haze,rain,snow,dust and other weather have affected the imaging of vehicle monitoring system and traffic monitoring system,complex weather have limited the application environment of intelligent monitoring system.The goal of this paper is enhancing the processing of image in bad weather and reducing the impact of bad weather on the intelligent monitoring system.Intelligent monitoring system get images in bad weather conditions,intuitively,images of contrast and sharpness have dropped,the image of objective parameters like variance and definition have decreased,it has seriously affected the system in the target recognition,traffic monitoring and other aspects of the use.In this paper,mainly focused on the images of haze conditions,proposed several haze weather image enhancement methods,analyzed the physical characteristics of haze weather and the objective parameters of haze weather images,and established the haze weather scattering model.The proposed method includes the image enhancement method based on the image HSL color pattern theory,the image enhancement method based on the theory of light propagation independence,the image enhancement method based on the atmospheric haze scattering model and the luminance model image theory,and the image enhancement method based on fast discretization curvelet transform.The motion model and the optical model of raindrops are established for the digital image of rainy days.The principle and steps of image enhancement method based on fuzzy C mean theory are introduced in detail.This paper has presented a variety of methods using Matlab and Visual Studio 2013 software programming,Visual Studio 2013 software using OpenCV computer vision library.Through the program to obtain the enhanced digital image,the enhanced image from the subjective and objective aspects of the analysis,to discuss the advantages and disadvantages of the method.Contrast before and after processing image quality of the image.Evaluation of image quality,used the variance,clarity and entropy objective parameters to confirm the haze image enhancement method.The rainy image enhancement method can effectively remove the raindrops in the image,and the enhanced images retain more image information.The image clarity is obviously improved,and the application scene of the intelligent traffic monitoring system is extended.