Research On Road Visibility Measurement And Visibility Early Warning System Based On Optimized Dark Channel Prior

In recent years,low-visibility weathers,such as sand storm,fog and strong light,are causing tremendous risks to road traffic and travel,seriously affecting the daily traffic operation management and the safety of life and property of travelers.In view of this,real-time monitoring of low-visibility road segments and timely warning of corresponding disastrous weather have become urgent tasks to ensure road safety in China.By reusing the surveillance camera equipments densely installed along the road network,this paper designs a monitoring video-based visibility detection method after an optimization of dark channel prior theory,a popular academic research in the field of image processing.Afterwards,with a reference to the industrial standards and regulations of road-side visibility measurement and driving safety,a road-side visibility detection and early warning system is developed based on Java EE technology.The major works of this paper are as follows:(1)Visibility detection method.By combining the priori theory of dark channel in the image processing field with the air visibility calculation model,a monitoring video-based visibility detection method after an optimization of dark original channel was designed.On the basis of the known parameters like actual distance between the camera and the identification target and the offset angle,this method converts the monitoring video frame into a grayscale image,and extracts characteristic variables,such as atmospheric light value and transmittance,hidden in the image.Then,the visibility distance of the road is estimated based on the geometric features of the camera devices and the visibility calculation model.The result of a simulation experiments show that the visibility error measured by the optimized dark channel prior visibility detection algorithm is within 15%,and the fidelity of the processed fog image is acceptable.(2)Visibility early warning strategy and system.with a reference to the industrial standards and regulations of road-side visibility measurement and driving safety,and by using the standard road visibility safety speed limit model,an early warning strategy for drivers and different visibility levels was designed.In view of the application requirement,the frame structure and the working process of a road-side visibility detection and early warning system is designed,and the selection of the components of the lower-end machine is done.An analysis with actual road environments of urban haze and mountainous fog shows that,the system proposed can provide early warning information such as road section direction,fog concentration,driving operation suggestions,maximum driving speed and minimum follow-up distance to drivers,thus being able to satisfy the early warning demand of actual road conditions.(3)System development.This road visibility monitoring and early warning system is mainly composed of two parts: a server system in the road network center and a group of lower end devices on road-side detection sites.The upper end server is implemented by using Java EE technology.It is designed in a hierarchical structure with a presentation layer,a business layer and an entity layer,and provides end users with functionalities like device configuration and road-side visibility warning.The lower end devices are implemented with an embedded IPC program to perform image acquisition and visibility calculation from on-site surveillance videos.A functional and performance test shows that,this system can satisfy the business function requirements of road visibility detection and early warning,and meets a great degree of safety and stability.