Vehicle data processing system and method for multiple parking lots

Abstract

The invention discloses a vehicle data processing system and method for multiple parking lots, and the method comprises the steps: enabling license plate recognition equipment to be connected to a router or a switch of each parking lot, searching and finding a camera unit in a local area network based on an ONVIF equipment discovery protocol, and obtaining a video code stream or a video image through a standard RTSP real-time transmission protocol; analyzing the obtained video code stream or video picture through an artificial neural network and a computer vision technology, and extracting license plate data in the video and a real-time video of a corresponding vehicle reservation time length; obtaining a vehicle identification result according to the license plate data and the real-time video; and finally, recording and storing the obtained license plate data and the real-time video in the vehicle identification equipment through a communication unit of the vehicle identification equipment according to a service strategy, and encrypting and reporting the identification result to the cloud data center. Therefore, a universal license plate data acquisition target which can be quickly copied across manufacturers can be realized by utilizing the camera of the existing vehicle identification equipment, and the development cost and period are reduced.

Description

technical field [0001] The invention relates to the technical field of intelligent parking, in particular to a vehicle data processing system and a corresponding method for multiple parking lots. Background technique [0002] The vehicle license plate information collection system based on automatic snapshot recognition technology has been widely used in parking lot management, but each parking lot is self-collected and built, and the snapshot cameras and management software manufacturers (brands) used are complicated. [0003] Since there is no national standard or industry standard for license plate recognition products, and there is no communication standard, if the competent department wants to supervise and realize the networked collection of license plate data in multiple parking lots, the following two solutions are currently mainly used: [0004] One is to adopt the docking development method of private protocol, that is, to ask each manufacturer of vehicle license p...

AI Technical Summary

Problems solved by technology

[0007] First, if the license plate manufacturer encrypts the data of the license plate recognition result, there is no way to effectively extract and analyze the license plate data without mastering the encryption method;

[0008] Second, many license plate recognition manufacturers have different encryption algorithms and encryption keys, and these encryption methods belong to the license plate manufacturer's technical codes, which will not be disclosed easily, which makes it more difficult to identify and extract license plate data;

[0009] Third, even if the encryption method of a certain license plate manufacturer is mastered by cracking reverse engineering, the license plate manufacturer can easily change the encryption method by upgrading the software, which will completely invalidate the encryption method obtained by cracking and analyzing. The method is time-consuming and labor-intensive and there is a risk of violating the law;

[0010] Fourth, at the network level, there are still abnormal situations such as network data packet loss and network attack, which cause the extracted license plate data to be distorted or lost. data is fake