Worker safety helmet wearing real-time detection method based on video analysis

Abstract

The invention discloses a worker safety helmet wearing real-time detection method based on video analysis. The method comprises the steps that captured video frames are read in real time from a camera in a monitoring area; a Viola-Jones detector is used to classify the image of the current frame to identify the upper body of a human body; a detected target is located, and the highest point of the detected upper body of the human body is determined; the highest point is used as a reference point to extract a corresponding area; and finally the color of the image area is determined according to color characteristics to determine whether a detected worker wears a helmet; and if a worker who does not wear a helmet is in the detection area, an alarm is given to ensure the safety of workers entering the work area. According to the invention, the influence of the background environment on helmet detection can be reduced; the helmet wearing detection accuracy of workers in the identification area is greatly improved; and the method has the advantages of fast detection speed and good instantaneity.

Description

technical field [0001] The invention relates to the technical field of video monitoring, in particular to a method for real-time detection of personnel helmet wearing based on video analysis. Background technique [0002] Construction sites such as construction sites or maintenance bases have poor operating environments. During construction or maintenance, workers have low awareness of safety precautions and objects are easy to fall, which can easily cause casualties. Wearing safety helmets can effectively reduce injuries to personnel. Therefore, at the entrance of the construction site or maintenance base, real-time detection of workers wearing helmets can effectively prevent potential safety hazards caused by personnel not wearing helmets. By installing a camera at the entrance, the video captured by the camera is processed in real time, so as to judge whether the worker is wearing a helmet and give an alarm in time, which can effectively reduce the labor of manual monitor...

AI Technical Summary

Problems solved by technology

Although the helmet detection method mentioned in Chinese patent CN104504369 can realize the detection of helmets, due to the limited training samples and the actual captured video is affected by the external environment, it is easy to cause false detection or missed detection

Because the background subtraction method used in the detection of moving objects has its own defects, the moving objects are easy to detect errors, so it is not conducive to practical application

Images