Adaptive Human Interaction Detection Algorithm for Oilfield Security System

This paper proposes an adaptive human interaction detection algorithm based on the Vision Transformer structure and domain adaptive algorithm for the security system in oil fields. The algorithm extracts features of individuals in the image using the Vision Transformer structure and adapts to different oil field scenes using domain adaptive algorithm. Experimental results show that the proposed algorithm has high accuracy and robustness in the application of security systems in oil fields.