Font Size: a A A

Parking Space Detection System Based On Video Recognition And Laser Positioning

Posted on:2021-04-24Degree:MasterType:Thesis
Country:ChinaCandidate:Y XuFull Text:PDF
GTID:2370330632957803Subject:Engineering
Abstract/Summary:PDF Full Text Request
With the maturity and improvement of automobile production technology,the improvement of people's living standards,and the change of consumption patterns,more and more people choose to travel by car.However,with the continuous growth of car ownership,finding parking spaces during driving has become a major difficulty.Although intelligent parking lot management and navigation technologies continue to be proposed,the high investment costs and vast projects have led to the current parking lot smart system is still limited to automatic identification of license plates and calculation of parking fees when entering and leaving the parking lot.With the development of Internet of Things technology,the monitoring range of parking lots has been greatly improved.Most parking lots are fully covered by cameras,and there is no blind spot for monitoring.But these monitorings are not used for other purposes except for security.Most of the parking lot videos proposed by predecessors envisaged installing new monitoring in each parking space,which is time-consuming and labor-intensive and difficult to implement.With the continuous advancement of video recognition algorithms,the security monitoring of the parking lot can now be used to realize the parking space navigation of the parking lot,and to solve the problem of car searching Therefore,we research and develop a parking space detection system based on video recognition and laser positioning.The system uses single video recognition technology on standard parking lots where surveillance is installed.At the same time,adding laser positioning technology can improve the use range and safety performance of the system.This system uses the latest YOLOV4 technology based on deep learning for target detection and identification of parking spaces.At the same time,it uses binocular ranging algorithm for parking space measurement.At the same time,it uses laser positioning technology to improve the recognition accuracy of parking space detection.Solve the problem of poor accuracy of traditional parking space detection,and improve the safety and robustness of the entire parking space detection system.On the basis of the abovementioned target detection and parking location technology,we designed and implemented a parking detection system based on video recognition and laser positioning.We designed the system architecture,and realized each function module.
Keywords/Search Tags:deep learning, convolution neural network, video recognition, parking lot navigation
PDF Full Text Request
Related items