My Role

The initial phase involved setting up the data acquisition system to capture images and video streams of vehicles. This data was essential for training and evaluating the vehicle detection and number plate recognition models.

I employed YOLO8, an advanced object detection algorithm, to accurately detect vehicles in the captured images and video streams. This step required fine-tuning YOLO8 to ensure high accuracy in identifying vehicles and their number plates.

The next task was to enhance the detection model specifically for vehicle number plates. This involved training the YOLO8 model with a dataset of annotated number plates, improving its precision in locating and identifying number plates.

To track vehicles and their number plates over time, I implemented efficient sorting algorithms. These algorithms helped in associating detected vehicles with their respective number plates across multiple frames or images.

For extracting characters from the number plates, I utilized EasyOCR, a powerful optical character recognition library. This tool enabled accurate extraction of characters from the detected number plates, turning visual information into actionable data.

Finally, I integrated these components into a cohesive system, ensuring seamless detection, tracking, and recognition of vehicle number plates. The system was thoroughly tested to verify its performance and accuracy before deployment.