Abstract: ［Objective］ In response to issues in existing urban rail transit video surveillance systems—such as insufficient continuity assurance for video recording, data security risks, limited storage capacity scalability, and complex resource management—it is urgently necessary to implement a cloud storage-centered expansion and renovation plan. ［Method］ The definition, basic structural model, and application advantages of cloud storage are briefly outlined. Based on the network architecture of existing urban rail transit video surveillance systems, a comparative analysis of distributed and centralized cloud storage architectures is conducted in terms of storage space demand, equipment space demand, power consumption demand, bandwidth requirements. On this basis, the architecture of the cloud storage renovation plan is proposed. The implementation process of the cloud storage renovation plan is described in three stages: construction of the cloud storage management system, backup of existing recordings and establishment of cloud storage nodes, and establishment of backup cloud storage nodes at the operation control center. ［Result & Conclusion］ The expansion and renovation plan for the video surveillance system based on distributed cloud storage technology adopts a direct data block storage approach. This solution achieves high reliability, high performance, high IO (input/output), scalability, and manageability. At the same time, it effectively reuses existing equipment resources to reduce renovation costs, and addresses technical challenges such as rapid hardware iteration, high failure risks, and long data retention periods.