In this paper, a license plate recognition (LPR) system that recognizes and extracts data from license plates is designed and implemented for car management systems. The proposed method is to use the deep learning network model to identify the license plate and information from the license plate, thereby determining the information of each different vehicle to include in the management system. By recording a video stream of the scene, the proposed system can detect and track multiple cars appearing in the scene and then recognize their license plates. Although the simulation results of the action recognition model have an accuracy of over 90%, in some cases, accuracy is highly dependent on the amount of training data and their resolution.