Model classification system and method

Abstract

The invention discloses a model classification system and method, and relates to the technical field of an automobile identification apparatus or method. The classification system comprises a forward snapshot camera, a body scanning camera, a dual-channel video detector and a charging system. The forward snapshot camera and a channel snapshot point are arranged to form an inclination angle of 45 degrees for capturing forward features of a vehicle passing by, and the body scanning camera is arranged at one side of a vehicle channel; and the forward snapshot camera and the body scanning camera are connected with the signal input end of the dual-channel video detector, and video stream vehicle detection analysis software is configured in the dual-channel video detector. The system provided by the invention performs multi-type and multi-dimensional data synchronization, integration and model fitting on information of the same vehicle on the basis of a core layer by use of one high-performance dual-channel video processor through a machine vision algorithm so as to accurately and stably obtain data information needed for accurate vehicle classification in real time and obtain the accurate vehicle classification.

Description

technical field [0001] The invention relates to the technical field of a vehicle identification device or method, in particular to a vehicle type classification system and a method thereof. Background technique [0002] Vehicle classification is an important basis for ETC charges at present. According to the existing national standards, various vehicles are differentiated and priced to ensure the accuracy of the expressway toll system. The current standards are roughly divided into five types of vehicles, and vehicles of different types are charged Different fees ensure the rationality and fairness of expressway tolls. At the same time, the fees charged on expressways also directly reflect the service level, safety and accessibility of expressways. According to the above factors, the accuracy of vehicle classification itself is very important. . [0003] At present, there are basically three types of vehicle classification methods used on expressways: loop coil detection, i...

AI Technical Summary

Problems solved by technology

[0008] The disadvantage is that the accuracy of single-view detection is easily affected by the height and angle of the camera. At the same time, when the vehicle speed is low and the distance between vehicles is small, it is easy to cause front and rear adhesion of vehicles, and it cannot effectively segment multiple vehicles. Multiple vehicles are misjudged as one Vehicle, which will affect the final detection accuracy of the system

Since the system is based on the principle of vehicle forward 2D graphic image processing, it cannot give the key information required for accurate classification such as the number of vehicle axles, wheelbase, number of wheels, and height of the front of the vehicle.

Images