Interaction Behavior Measurement of Electrical Workers Based on Gravitational Field Model

Effective interaction among workers at electrical work sites can help prevent accidents and improve work efficiency. The existing interaction measurement methods for electrical workers mainly use visual observation, which is inefficient, slow, and easily influenced by subjectivity, resulting in low measurement accuracy. To improve the measurement speed and accuracy, in this paper, a method for measuring the interaction behavior of electrical workers based on the gravitational field model is proposed by using video surveillance and computer vision technology. Firstly, based on this technology, four features, namely interaction distance, interaction emotion, posture openness, and interaction duration, are extracted. Then, a measurement model for interaction between the four features and workers is established using the gravitational field in physics. Finally, a video dataset of interaction behavior is constructed to verify the accuracy of the model. The results show that the average relative error of this method is $0.199\pm 0.0129$ . This article can provide a new idea for enriching interaction behavior measurement methods. It also provides new technologies for feature extraction.