A video-based car model matching method, device and billing system

Abstract

A video-based car model matching method, including: identifying the license plate information in the target parking space, and recording it as the current license plate information; calculating the similarity between the current license plate information and the first recognized license plate information in this parking event; judging and calculating the obtained Whether the degree of similarity is greater than the first preset threshold; when the calculated similarity is less than the first preset threshold, extract the close-up picture of the vehicle in the target parking space captured this time, and record it as the current close-up picture; extract the previous snapshot The close-range picture of the vehicle in the target parking space that arrives is recorded as the target close-range picture; Whether the similarity of judging the current close-range picture and the target close-range picture is greater than the second threshold; The last collected out-of-position information and in-position information are extracted from the collected video information, the parking event is updated, and the in-position information is used as the updated in-position information of this parking event. Improve the recognition rate of parking time.

Description

technical field [0001] The invention relates to the technical field of image pattern recognition, in particular to a video-based car model matching method, device and billing system. Background technique [0002] With the development of my country's economy, the number of vehicles in the market will continue to increase, and problems such as parking difficulties and effective management of parking spaces will be further aggravated. On-street parking came into being in this historical context. As an important part of intelligent transportation, on-street parking refers to setting up a number of parking spaces on both sides or one side of the road safety red line for temporary parking of passing vehicles. This kind of parking space has the advantages of more flexible setting, fast turnover rate of parking spaces, less space occupation, and low maintenance cost, which can alleviate the current "difficult parking" and other problems to a certain extent. Aiming at parking scene...

AI Technical Summary

Problems solved by technology

During the periodical capture process of the front-end, if the vehicle license plate can be recognized, it can be judged whether to change the car by comparing the license plate information. If the license plate is blocked or there is no license plate and the license plate information cannot be recognized, the only way to realize whether to change the car is through image matching. Restricted by the limited computing power of the front-end capture equipment, the matching is easily affected by the on-site environment, resulting in matching errors, causing one parking event to be divided into multiple parking events, resulting in multiple orders, or multiple parking events being treated as one parking event Events that affect parking management system charges

Images