Helmet Detection under the Power Construction Scene Based on Image Analysis

Abstract

In this paper, an intelligent safety surveillance system is designed for the electric power construction system to detect whether workers wear helmets correctly. The YOLOv3 objective detection algorithm is used for fine-tuning the datasets for the electric power construction scene which is made by ourselves. The main targets for detection are the helmet and the head. The helmet and the head are tested to detect whether the worker wears a helmet. After our experiment, the accuracy of the helmet detection is over 90%. This method replaces the manual method with an automated way which saves a lot of human and material resources. This method can be applied to actual production.