Plant personnel abnormal behavior detection method based on deep learning

Abstract

The invention provides a plant personnel abnormal behavior detection method based on deep learning. The method comprises the steps: training a plant personnel model based on a YOLOv4 network; settinga capture frame rate and a detection interval; acquiring a real-time video stream of a production area through a camera, capturing images according to a set image capturing frame rate and a set detection interval, and preprocessing the images; loading the trained factory personnel model, and carrying out factory personnel detection on the preprocessed image to obtain a target box of one or more factory personnel areas; performing target tracking on the target frame by using a Sort multi-target tracking algorithm to obtain a target area image; cutting the target area image, performing gray scale transformation on the cut image, and converting the cut image into a gray scale image; utilizing a Horn-Schunck optical flow method to calculate the change amplitude of the grayscale image in the same frame between two times of anomaly detection; and judging whether the personnel behavior is abnormal or not according to the change amplitude. According to the invention, the judgment result is accurate, and the monitoring efficiency and reliability can be improved.

Description

Technical field[0001]The present invention relates to the field of intelligent monitoring, and in particular, an abnormal behavior detection method based on deep learning plant personnel.Background technique[0002]The security of the company, ordered production is the key to the continued development of the national economy and the key to enterprises. How to effectively do the safety monitoring of production places, do safety precautions, eliminate safety hazards, avoid accidents, and become safe The topic should have been concerned in production.[0003]For production lines, operating workers cause physical injuries due to misunderstandings, or hire, go, idle, watching mobile phones, causing the production of abnormal interruptions. For more equipment, production workshops in the production process, there are many closed areas. In this area, when only one person is working in the field and the accident is lost, it may be delayed due to unreasonable discovery and delay in the treatment...

AI Technical Summary

Problems solved by technology

[0003] For the production line, it often happens that operators are injured due to misoperation, or abnormal production is interrupted due to behaviors such as dozing off, distracted, chatting, and looking at mobile phones.

For production workshops with many equipment and complex production processes, there are many closed areas

In this area, when there is only one person working on the site and an unexpected situation occurs where unconsciousness or immobility (such as fainting) occurs, treatment may be delayed and the condition may worsen due to unnoticed

[0004] In order to ensure high-efficiency and safe production, manual inspections are mainly used at present. Post workers have to inspect all production lines and closed areas 7*24 hours, which is inefficient and requires a lot of manpower, material resources, and financial resources.

Images