Vehicle Speed Detection Using Kalman Filter and DBSCAN
M Naresh Kumar, Gouru Samhitha, Kaitha Sravya · 2022 International Conference on Computer Communication and Informatics (ICCCI) · 2022
Many traffic accidents occur across the planet as a result of rash driving. Because there's no efficient way to regulate or monitor the speed of cars on the road, the population of traffic has risen dramatically. This technique aids within the monitoring of car speeds so as to stop collisions. The footage taken with the help of a camera is employed to determine the vehicle's speed. However, as technology advances, several regulatory organizations are requesting some form of computerized technology to tackle the issues of excessive speeding. In this instance, we design a scheme that can identify vehicles that are exceeding the maximum speed restriction set by the individual roads or highways. The project is divided into three sections: speed detection, picture capturing and transmission, and image analysis. The microwave Doppler radar sensor utilized in the speed detecting gadget operates on the idea of the Doppler effect. The speed is compared to the predetermined threshold, and if the speed restriction is exceeded, the camera is activated. The data is depicted in the form of a clip, and is captured with a camera with a 1280x1024 screen size. Four videos of identical resolution are taken for the data base. After foreground detection and DBS can with the help of bounding boxes speed of particular vehicle is detected. The motion of the vehicle is detected using Kalman filter. We can also theoretically estimate the average velocity using number of video frames.