A model for detecting safety hazards in key electrical sites based on hybrid attention mechanisms and lightweight Mobilenet

The emergence of electrical energy has brought great changes to people’s lives, and with the development of science and technology the demand for electricity is increasing and the number of electrical places is not increasing, with this brings more and more complex safety hazards. This paper proposes a hybrid attention mechanism and lightweight Mobilenet-based safety hazard detection model, which is less expensive to train, faster and more effective than existing model algorithms, and can detect safety hazards in key power sites as early as possible to reduce casualties and property damage.