Street view flow information acquisition method based on computer vision

Abstract

The invention discloses a street view flow information acquisition method based on computer vision. The method comprises the following steps: detecting and identifying objects in a street view for each frame of video by adopting a target detection algorithm YOLOv5; extracting appearance features of the detected object to assist matching of the detection frame and the prediction frame; predicting the position of each detected target appearing in the next frame by using a Kalman filtering algorithm; using a Hungary algorithm to calculate a cost matrix by using the extracted appearance features and motion features, and realizing cascade matching of detection frames to allocate a tracking target sequence number for an identified object; maintaining appearance features and tracking serial numbers of the detected object by using data structures such as a prototype library, and judging whether the detected object appears in the video for the first time or not according to the appearance features and the tracking serial numbers; and intercepting a small image of the object detected for the first time, transferring the small image to a specified path, counting the number of different objects appearing in each category, and displaying the motion trail of the object in the video. According to the invention, image information of common objects and counting statistical information of different types of objects in the street scene are collected in real time.

Description

technical field [0001] The invention belongs to the technical field of multi-target tracking, and in particular relates to a method for collecting street view flow information based on computer vision. Background technique [0002] Video target tracking is an important task in computer vision. It refers to the process of continuously inferring the target state in the video sequence. Its task is to generate the trajectory of the target by locating the target in each frame of the video, and Provides the complete target area where the tracking target appears in the video at each moment. Video tracking technology has a very wide range of applications in the field of computer vision. The present invention combines and improves the target tracking technology and detection technology, and optimizes the problems of slow reasoning speed and tracking loss of commonly used street view flow information collection technology. [0003] The target detection algorithm commonly used in str...

AI Technical Summary

Problems solved by technology

[0004] At the same time, the tracking algorithm commonly used in street view traffic information collection is the sort algorithm, which fails to make full use of image information, resulting in track-id switching after the tracking target is temporarily blocked

Images