Research On The Detection Algorithm Based On Image Processing Of Traffic Flow

Video detection technology has a very broad application prospects in traffic detectionsystems. In a well-functioning traffic parameters and Automatic Incident Detection system,traffic detection is the most critical part. Currently traditional methods are still unsatisfactoryin many respects, and video-based traffic information detection identification law in regard toother detection methods have many advantages, such as simple operation, easy maintenance,testing and information-rich, such as a wider range. With the development of computertechnology, the video image recognition method based more and more attention, has becomethe development trend of intelligent traffic monitoring system.This paper studies a more practical and video traffic detection algorithm optimization,the algorithm uses image processing software to achieve prepared to complete the traffic flowinformation collection. Compared with the previous traditional hardware detection methods,detection technology based video surveillance system to monitor the video stream only needsto load analysis can achieve full multi-lane road monitoring, coverage range, detectionaccuracy, simple equipment, low cost, and easy maintenance and updating.This paper first used digital image processing techniques were analyzed, and thencombined with the specific requirements of the application on this basis, the key algorithmfor the overall design of the system, which is important for the realization of the principlefunctions of the system for a more detailed deduction, while the detection traffic, density andvelocity verified. Software analysis using Matlab as a tool for the field of view of movingtargets using image smoothing, background extraction, image segmentation techniques werestudied to detect traffic problems and the treatment effect was verified completed trafficparameters and event detection system comprehensive verification. Finally, system testresults show that the system is basically designed to meet the design requirements.