Technology Of Vehicle Speed Detection Based On Image Sequence Analysis

The video-based traffic information detection technology has become a topic in the field of transportation in recent years. As an important part of video-based traffic information detection, video-based vehicle speed detection has an extensive application prospect. Therefore, it's of great significance to do research on vehicle speed detection.This paper researches on the video-based vehicle speed detection system and the detection algorithm. The one-dimensional and two-dimensional camera calibration methods suitable for vehicle speed detection are designed. The main work is as follows:1. Video-based vehicle speed detection system. This paper studies on the constitution of the system and the network communication between DSP detector and user identification terminal. Taking the communication needs and communication cost into account, this paper adopts a suitable protocol for the communication between different modules. It not only meets the needs of data transmission reliability, but also saves network resource.2. The methods of camera calibration. Extending the one-dimensional calibration algorithm based on linear model to two-dimensional. This improves the accuracy of the camera calibration. As the two-dimensional mapping table takes a lot of storage space and visiting time, so a dynamic mapping table based on the two-dimensional calibration is built.3. Vehicle speed detection algorithm. By analyzing the driving characteristics of vehicles in different traffic conditions, this paper designs two detection algorithms for different scenes. It improves the application flexibility of the video-based vehicle speed detection system.The vehicle speed detection algorithm is tested in urban roads and expressway traffic environment. The results show that the algorithm can timely detect the instantaneous velocity of vehicles with accuracy of more than 90%. The algorithm can meet the requirements of practical applications, and it can provide available reference data for traffic management department.