Multi-target vehicle trajectory recognition method based on video tracking

Abstract

The invention provides a multi-target vehicle trajectory recognition method based on video tracking, and relates to the technical field of video monitoring. The method comprises the following steps: step 1, acquiring a monitoring video image, and setting related parameters; 2, acquiring all vehicle targets to be tracked in the image by adopting a YOLOv3 algorithm according to the video image, taking the detected N vehicles as tracking targets to obtain a Qth frame target frame set SQ of the N vehicles, and establishing a track set L = {L1, L2, L3,..., LN} by taking the central point of the target vehicle as a track recording point; and 3, carrying out vehicle target detection by adopting an improved YOLO v3 algorithm to obtain a target frame set SQ + 1 of Q + 1 frames, repeating the step until the collected monitoring video images are completely detected, and outputting a final track set L '. The method not only solves the problem of multi-target vehicle trajectory extraction precision, but also has good real-time performance, and also solves the problem that the existing vehicle trajectory extraction precision and speed based on a deep learning method are limited.

Description

technical field [0001] The invention relates to the technical field of video monitoring, in particular to a multi-target vehicle track recognition method based on video tracking. Background technique [0002] With the continuous improvement of transportation facilities and the improvement of people's living standards, there are more and more vehicles on the road, and traffic safety problems also follow. Among many sources of traffic information, the video data based on surveillance cameras has the characteristics of uninterrupted, intuitive, and high reliability. Therefore, the method of vehicle trajectory recognition based on surveillance video is one of the important means to judge whether the dynamic process of vehicles is standardized. one. In addition, vehicle trajectory recognition has been applied in other fields and industries today, such as in automotive assisted driving systems. In harsh environments such as low visibility, vehicle trajectory recognition can remin...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is to provide a multi-target vehicle trajectory recognition method based on video tracking, which not only solves the problem of multi-target vehicle trajectory extraction accuracy, but also has good real-time It solves the problem of limited accuracy and speed of vehicle trajectory extraction based on existing deep learning methods

Images