A Framework for Vehicle and License Plate Detection in Pakistani Traffic Environments Using Algorithms Pool

Automation of vehicle and license plate detection plays a vital role in several applications, such as security, traffic management, vehicle monitoring, tool collection, and parking-lot entry/exit control. This paper presents a framework to detect vehicles and license plates for Pakistani traffic environments. To detect vehicles and license plates, the proposed framework uses two pools to locate vehicles and license plates. Each of the pools uses two algorithms to locate vehicles and license plates. The pool of vehicle detection uses the Gaussian Mixture Model (GMM) along with the Faster-RCNN. While the license plate detection pool uses an image processing-based technique along with the HSV segmentation method. Detailed simulations have been performed on our own collected 1142 diverse Pakistani traffic images. Our system achieves a mean accuracy of 90.26% and consumes less than 2 seconds on average to locate vehicles and license plates.