The acceleration of China's urbanization process, the passenger flow of urban rail transit is continuously increasing, causing the safety of subway operation increasingly severe challenges. The Video Surveillance System (VSS) shoulders functions such as safety prevention and passenger abnormal behavior monitoring. VSS can assist station staff in quickly locating and handling abnormal events. Video recording data is also used as an auxiliary sentence for handling subway public security cases. The increasing security requirements make traditional video surveillance methods difficult to meet development needs. Meanwhile, the massive amount of video data also brings enormous storage pressure. New technologies such as cloud computing, big data, and artificial intelligence (AI) have given rise to intelligent video surveillance systems. The development and application of new technologies can comprehensively enhance the disaster prevention and reduction capabilities of stations and improve safety management levels. This article will start from the current construction status of Shenzhen's rail transit and elaborate in detail on the intelligent application of VSS, including video analysis applications based on image recognition and cloud storage construction models based on line network cloud platforms. By analyzing the problems in the construction of intelligent video surveillance systems, this paper proposes thoughts and suggestions for the future construction of urban rail transit video surveillance systems.