Multi-target tracking and behavior analysis detection method

Abstract

The invention provides a multi-target tracking and behavior analysis detection method, which is applied to the technical field of automatic target tracking, and comprises the steps of decomposing based on a to-be-tracked figure image to obtain a plurality of target figure head portraits; inputting the head portrait of the target person and the to-be-analyzed image into a network tracking model toobtain a confidence coefficient value and a coordinate box, the network tracking model being a tracking model composed of a CNN and an RPN; selecting a first target number of coordinate boxes as firstcoordinate boxes according to the confidence coefficient value and the coordinate boxes; obtaining a tracking result according to the historical tracking value and the first coordinate box; and outputting detected candidate human head target information according to the tracking result, the human head target information comprising the scale of each target and the total number of candidate human head targets. By applying the embodiment of the invention, the problems of false alarm, missing alarm, difficulty in tracking and the like easily caused by the reasons that a person detection and tracking method based on face detection, head and shoulder detection or whole body detection is easily shielded, targets are difficult to classify and the like in the prior art are solved.

Description

Technical field [0001] The invention relates to the technical field of target tracking, in particular to a method for multi-target tracking and behavior analysis and detection. Background technique [0002] With the rapid development of social economy, the application of video surveillance systems in various industries is gradually mature, and the number of cameras built and put into use is rapidly increasing. As the surveillance system collects a large number of videos, real-time monitoring of these videos has gradually become a major problem. It is difficult for users to keep their vigilance all the time while staring at the screen, and subsequent data loss cannot cause analysis. Therefore, it is necessary to adopt an intelligent analysis system to analyze the video while collecting the video. [0003] In the prior art, the current intelligent analysis technology is based on human face detection, head and shoulder detection and whole body detection. The accuracy of the analysis ...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to overcome the defects of the prior art, and provide a multi-target tracking and behavior analysis detection method, aiming to solve the personnel detection and tracking methods based on face detection, head shoulder detection or whole body detection in the prior art, False positives, missed negatives, and difficult tracking are likely to occur due to reasons such as easy occlusion and difficult target classification.

Images