Attack behavior early warning method and system

Abstract

The invention relates to an attack behavior early warning method and system. According to the technical scheme, the system comprises a monitoring end and a data processing end. The monitoring end is arranged at the top of a monitored space and is used for shooting pedestrian images in a monitoring range, identifying whether a current image contains pedestrians or not by utilizing a built-in lightweight neural network, tracking trunk organs which possibly have attack behaviors when the image contains the pedestrians, and transmitting real-time shot pictures to the system data processing end; the data processing end calculates the target behavior action characteristics, calculates the movement distance by using the trunk part movement path intercepted in the real-time camera shooting picture, calculates the behavior characteristics by combining the time consumed by the behaviors, compares the behavior characteristics with the attack behaviors of the array, and gives an alarm when the dangerous behavior alarm condition is met. And alarm information is sent to public security management personnel in the current area. The method has the advantages that group abnormal behaviors are found in time through intelligent algorithm comparison, and system monitoring management personnel are effectively assisted to conduct scientific early warning in advance.

Description

technical field [0001] The invention belongs to the field of attack behavior early warning systems, and relates to an attack behavior early warning method and system. Background technique [0002] Human motion recognition technology is widely used in monitoring recognition and other fields. In the environment of receiving video surveillance, the surveillance system uses deep learning technology to analyze the abnormal behavior of the crowd, so as to achieve the purpose of early warning to avoid risks. The existing single-stage action recognition technology is difficult to predict in terms of obtaining pedestrian information, and cannot make accurate judgments and timely warnings for abnormal crowd behaviors. [0003] Therefore, the present invention provides an attack behavior early warning method and system, which constructs an attack behavior array, compares it with the suspected attack behavior monitored in real time, and combines the action characteristics and cut-off c...

AI Technical Summary

Problems solved by technology

The existing single-stage action recognition technology has the problem of difficult prediction in obtaining pedestrian information, and it is impossible to make accurate judgments and timely early warnings of abnormal crowd behaviors.

Images