The system is mainly divided into four parts: interface design, license plate collection, license plate recognition and data management. In the license plate collection, the camera is used to collect the video of the license plate when it is in and out, and the frame analysis of the video stream is completed by calling OpenCV in Python. In terms of license plate recognition, Baidu AI is used. The image recognition method of the platform realizes the recognition of the license plate. In terms of data management, the time of vehicle entering and leaving the parking lot is recorded in the form of a table, and the charging information is automatically generated in combination with the charging rules to achieve automatic charging. The results show that the design can improve the economic benefit of the depot, improve the service quality and enhance the user experience.