A Practical Approach for License Plate Recognition for Non-Helmeted Motorcycle Riders

To protect motorcycle riders, it is now required to detect license plates while they are not wearing helmets. This research uses the real-time object detector YOLOv10 to demonstrate the real time detection of License plate for helmetless motorcyclists. And, a single convolutional neural network was used in the suggested method to automatically identify a non- helmeted motorcyclist’s License Plate from the video stream. Moreover, the false positive caused by the helmeted rider leaving the video frames was removed using the centroid tracking approach and a horizontal reference line. $\mathbf{9 8. 5 2 \%}$ of all License Plate were detected overall.