Car Type and License Plate Detection Based on YOLOv4 with Darknet Framework (CTLPD)

Abstract

AbstractVehicle and license plate detection are important functions in an intelligent transportation system. This research paper proposes automatic vehicle identification and categorization, and also detects and recognizes of license plates and characters on it. The state-of-the-art You Only Look Once (YOLO)-Darknet deep learning framework is used to address these tasks. Two different YOLO-Darknet frameworks are applied. The first one detects and categorizes the cars. The other framework detects the license plates of the identified car accurately. Experiment results show that the proposed technique is efficient and outperforms the traditional machine learning and deep learning techniques like SVM-classifier, traditional-CNN, and fast R-CNN networks for object detection and categorization and character recognition.KeywordsCharacter segmentationDeep learningDarknetSliding windowVehicle detectionVehicle type detectionYOLOv4