Intelligent method for detecting traffic accident at night

Abstract

The invention provides an intelligent method for detecting a traffic accident at night. The method comprises the following steps: acquiring a plurality of monitoring video images of a vehicle during driving at night, and processing the monitoring video images so as to acquire a plurality of binary images; judging car light of the vehicle from the binary images, and determining the position of the vehicle according to the car light of the vehicle; establishing a lane model according to the position of the vehicle, and establishing an image coordinate for the vehicle according to the lane model; acquiring the driving track and driving speed of the vehicle according to the image coordinate and the lane model; and judging whether the vehicle violates traffic regulations according to the lane model, and the driving track and driving speed of the vehicle. According to the intelligent method for detecting the traffic accident at night provided by the embodiment of the invention, the working efficiency of existing road video monitoring is improved, the working strength of monitoring personnel is reduced, and meanwhile traffic delay caused by traffic accidents is reduced so as to prevent secondary accidents.

Description

technical field [0001] The invention relates to the technical field of video monitoring, in particular to an intelligent night traffic incident detection method. Background technique [0002] In recent years, with the rapid development of social economy, the traffic volume on many domestic highways is increasing day by day, causing more and more traffic accidents and road congestion, causing huge property losses and casualties, and damaging the public interest. Video surveillance technology has been widely used in the highway operation management system. A large amount of video data is generated in the operation and maintenance management department. It is a labor-intensive task for managers to discover and analyze various abnormal events in time from the massive video data. Therefore, the abnormal event detection technology of expressway based on video surveillance has a good market prospect in expressway safety. By developing a detection algorithm for abnormal video event...

AI Technical Summary

Problems solved by technology

This method has two deficiencies: (1) The algorithm of lane recognition is established by using edge extraction information, but this algorithm cannot automatically establish lanes for suburban roads and expressways with insufficient light, and for situations where lane lines in cities are not obvious or blurred

However, this method has a large amount of calculation and takes a long time to calculate, and because it does not consider the individual behavior of the vehicle, it only judges from the overall characteristics, and cannot give detailed information about the vehicle that occurred in the event. generate false positives

Images