Construction site safety helmet wearing detection method based on lightweight convolutional neural network

Abstract

The invention provides a construction site safety helmet wearing detection method based on a lightweight convolutional neural network, and belongs to the technical field of machine vision, and the method comprises the steps: obtaining a construction site safety helmet picture detection training set; constructing a safety helmet detection lightweight convolutional neural network; pre-processing the construction site safety helmet picture detection training set and inputting the pre-processed construction site safety helmet picture detection training set into a constructed safety helmet detection lightweight convolutional neural network for training; and using the trained safety helmet detection lightweight convolutional neural network for actual construction site safety helmet wearing detection. The lightweight convolutional neural network for safety helmet detection provided by the invention has the advantages that the detection speed is high, the parameter quantity and the calculation quantity are greatly reduced, the accurate detection of safety helmet targets at different distances can be realized, and the real-time requirement of the safety helmet wearing detection on an actual construction site and the problem of being used for embedded equipment are solved.

Description

technical field [0001] The invention belongs to the technical field of machine vision, and in particular relates to a safety helmet wearing detection method at a construction site based on a lightweight convolutional neural network. Background technique [0002] In the construction environment, construction personnel wearing safety helmets can effectively avoid or reduce the injury of safety accidents. It is time-consuming and labor-intensive to detect the wearing of helmets through manual inspection of surveillance video, and it is easy to cause false detection and missed detection. The use of machine vision technology to detect whether the construction site workers wear helmets can effectively replace manual inspections, improve the accuracy of recognition, and avoid safety accidents. [0003] The traditional hard hat detection is mainly by extracting the geometry, color and other features of the target for comparison and recognition. For example, the skin color detection...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the above-mentioned deficiencies in the prior art, the present invention provides a construction site safety helmet wearing detection method based on a lightweight convolutional neural network, which solves the problem that the current safety helmet model based on deep learning has too many parameters and cannot be deployed in embedded problem on a device

Images