A management method and device for on-street parking

Abstract

Embodiments of the present invention provide a management method and device for on-street parking, the method comprising: acquiring video image information collected by multiple cameras, determining the parking area detection frame in each video frame; detecting whether there is a vehicle in each video frame , if it exists, determine the coordinate information of the berth area detection frame, carry out vehicle feature extraction to each image frame through a multi-layer convolutional neural network, and obtain the first feature representation vector of the vehicle; through a multi-layer convolutional neural network, Perform feature fusion on the first feature representation vector to obtain a second feature representation vector, calculate the motion vector direction and motion distance of the vehicle in each frame of image; analyze the predetermined The direction of the motion vector and the size of the motion distance of the vehicle within the time period are analyzed to obtain the analysis results and determine the parking event of the vehicle. Through the present invention, the accurate and efficient recognition of vehicle entry and exit events is realized.

Description

technical field [0001] The invention relates to the technical field of intelligent parking management, in particular to a management method and device for roadside parking. Background technique [0002] Parking management based on high-level video has become an important topic in the construction and development of smart cities in recent years. The parking management method first collects images and videos of vehicles and parking space information through cameras, and then uses computer vision technology to analyze vehicle information and vehicle behavior. Analyze and understand, so as to realize the monitoring and management of on-street parking. However, the high-level video technology in the existing technology is easily affected by complex scenes, weather, object occlusion and other factors, so that the accuracy rate cannot meet the application requirements of the scene; on the other hand, with the development of deep learning in recent years, based on The deep learning...

AI Technical Summary

Problems solved by technology

However, the high-level video technology in the existing technology is easily affected by complex scenes, weather, object occlusion and other factors, so that the accuracy rate cannot meet the application requirements of the scene; on the other hand, with the development of deep learning in recent years, based on The deep learning of convolutional neural network has made remarkable achievements in the fields of image recognition, image detection and image segmentation. However, in the way of deep learning in the prior art to detect vehicles, it is often detected by marking the detection frame of the vehicle. If When the background of the image captured by the video is complex, it is usually impossible to identify accurate image information, which affects the accuracy of subsequent vehicle recognition or vehicle behavior analysis results

Images