System Of Traffic Congestion Identification Based On Multi-feature Fusion

In recent years,with the rapid development of social economy,motor vehicle ownership showed a rapid growth trend,traffic congestion situation is increasingly restricting the development of major cities.Camera monitoring technology is becoming increasingly common,and more powerful.So we can use them to deal with the increasing number of data and video,because simply handing the traffic video analysis by people is almost impossible.Therefore,the intelligent detection of traffic flow status is of great significance to solve traffic congestion.The project uses video security technology to collect and process traffic flow data,which can provide decision-making basis for improving and solving traffic congestion problem.Through the development of road congestion on the PC machine to determine the system to capture the video image processing,you can analysis and get the congestion tips effectively on the camera video.So we can alleviate the excessive output of human resources,to achieve traffic regulation and traffic congestion intelligent judgments.We get the traffic jamming video though using the Gun camera from Ruijie security.Then we rewritten the road extraction algorithm and write it to the embedded system of the camera,based on the multi-feature fusion traffic congestion decision algorithm.And then improved group membership algorithm,we use the real-time video improvement based on the difference and the number of background extraction method so that the camera can realize the real-time acquisition of the video image and return the extracted image of the road background to the PC side.PC side receiving the image and at this time it could detected the moving targets and multi-feature fusion of traffic congestion.Finally we write the system interface by QT language,and achieve a complete road video data acquisition,processing,human-computer interaction process.So the practical road background extraction algorithm and the high accuracy of traffic congestion judgment algorithm can be applied to the actual reality.