Research Of Traffic Flow Parameters Detection System Based On Video

Along with improving the construction of the highway network, the use of a more complete video surveillance system for real-time monitoring of road conditions as to improve the quality of the road is an important means of operation. Target detection based on video image analysis and data collection play a greater role in monitoring system to provide the necessary technical support relative to traditional means of monitoring a high degree of intelligence, information diversity, the advantages of high return on investment.Based on digital image processing and computer vision technology, in the video image of target motion model for the analysis and research, through the moving target detection, target tracking and traffic parameter extraction algorithm, completed the target hot spot areas in video surveillance, and target motion behavior of the filtering and classification, completed the statistical analysis of traffic flow parameters. Using this algorithm, the experimental parameters of traffic flow on the section tested, the results show that high detection accuracy, for traffic operation and management to provide the necessary technical support, achieve the purpose of improving the road network operational efficiency.The results of this study, using segmentation of road running lane as monitoring key areas to improve the detection efficiency, suppress noise and reduce the system false alarm rate and the rate of false positives; using Kalman filtering algorithm based on feature area to track the target, enabling fast tracking of moving targets in complex background, increased tracking efficiency, strong robustness. Traffic flow detection system based on this development, video surveillance system to achieve a normal and effective operation, and has been applied in practical engineering, the system is stable, safe, reliable and has high practical value.