Key points are not available for this paper at this time.
Modern intelligent transport systems heavily rely on advanced analysis of road traffic data. This study presents a novel method for detecting vehicle speed that makes use of image and video processing methods. The process analyses video data in real-time without requiring previous camera calibration. Masking and frame subtraction are used to precisely isolate moving cars from the background. To determine speed, two metrics are measured: the time interval between frames and the distance covered by segmented objects. Frame masking also allows for the distinction of different cars inside the same frame. When implemented to different video sequences, the suggested framework shows excellent accuracy in speed recognition, with an average error of only +/- 2 km/h. The approach was put under extensive testing on an active roadway in real-world conditions in order to measure its reliability. Even in difficult circumstances, the experiment's results suggest a very impressive average deviation of only 2.77%. With the solid foundation for optically monitoring vehicle speeds given by this research, traffic management and security in intelligent transportation systems could be significantly improved.
Bhadange et al. (Thu,) studied this question.