In this study, a safety monitoring system based on intelligent video analysis is designed and realized for the safety management needs in the high-risk operation environment of power grid construction sites. The system constructs a complete technology chain from front-end data acquisition to edge intelligent analysis to cloud decision support by integrating multimodal perception technology and deep learning algorithms. At the video monitoring level, the system adopts highdefinition video acquisition and multi-source sensor data fusion program, which breaks through the limitations of traditional monitoring means in complex construction scenarios. The intelligent analysis algorithm layer integrates target detection, behavior recognition, and anomaly detection technologies and realizes real-time accurate identification of helmet/work uniform compliance, high-risk behaviors, and environmental risks through lightweight model deployment and model compression technology. The system architecture adopts a three-layer design mode, including perception layer, edge layer, and platform layer, in which the intelligent analysis gateway supports dynamic rule configuration and multi-algorithm collaborative reasoning, the data convergence platform realizes unified access and storage management of multi-protocol video streams, and the visualization interface realizes three-dimensional reconstruction of the construction scene and safety situational awareness through digital twin technology. In terms of key technology implementation, the study proposes a safety equipment detection method that integrates color features and human posture estimation, as well as an early warning mechanism for high-risk behaviors that combines geo-fencing and trajectory prediction, which significantly improves the level of intelligence and real-time response capability of power grid construction safety control. This study provides an endto-end solution for safety monitoring in the electric power industry, which is of great theoretical value and practical significance for reducing the risk of construction safety accidents and improving the effectiveness of safety management.