Segmentation-based Vehicle Tracker for Real-Time Parameter Estimation on Smart Camera
Maksim D. Ershov, Sergey Smirnov, Valery V. Strotov · 2020
The article deals with the problem of analyzing traffic flows while observing a road section with a fixed video sensor. An algorithm for tracking vehicles that is moving in the lanes is presented. The algorithm is based on object segments tracking. It is computationally efficient and suitable for implementation on the smart camera platform. Experimental studies show a reduction in the number of false detections and undetected vehicles comparing to the base algorithm. The rate of object classification errors has also decreased, which indicates an increase in the accuracy of estimating the length and speed of detected vehicles.