The Vehicle Speed Detection Based On Image Processing In Tracffic Monitor System

In recent years, traffic information detection based on image processing technology hasbecome a hot research field of traffic. As an important part of traffic information detection,and image processing speed detection system has broad application prospects. Therefore, thevehicle speed detection study has important significance.Papers for speed detection systems and vehicle detection algorithm based on imageprocessing were studied design suitable for vehicle speed detection of one-dimensional andtwo-dimensional camera calibration methods. The main contents of the thesis work are thefollowing:1. Video vehicle detection system. Papers for speed detection system carried out adetailed research, analysis now from the difficulties, the paper design for hardware andsoftware structure composed speed detection system.2. Camera calibration methods. The calibration of linear model based on onedimensional extended to two-dimensional calibration, improved camera calibration accuracy.3. Speed detection algorithm. Paper extracted by the image processing of the targetvehicle, to obtain the feature points of the target vehicle, the feature point trajectory fittingcombination camera calibration distance in pixels between each frame will be converted tothe actual distance to calculate the vehicle speed.This paper presents a practical speed detection methods were tested in the city roads, theinstantaneous speed of the vehicle can achieve real-time detection, in the practical applicationmore than95%accuracy. This design can meet the requirements of practical application, thetruth can be applied to urban speed detection, speed and other aspects of the investigation.