Campus security incidents in colleges and universities are sudden and urgent in nature. Once they occur, they will have a great negative impact on the safety of college students and social stability. To address this problem, this paper proposes to build a comprehensive safety protection system for college students based on digital twin technology. The system first builds a digital twin model of the campus through 3D modeling technology, and then deploys IoT devices to collect video streams, entry and exit records, and emergency information on the campus. The collected data is transmitted to the big data center and efficiently stored and analyzed using cloud computing technology. On this basis, an intelligent early warning system is developed using support vector machines to predict and warn of potential security risks, while also building an emergency response system. The system also includes virtual safety education and training modules developed using digital twin technology, and regularly organizes emergency drills to test and optimize emergency response processes. Finally, through the suggestion box and online feedback system, we collect opinions and suggestions from students and faculty on the safety assurance system, and optimize and upgrade the safety assurance system based on the feedback and safety incident analysis results. After the comprehensive safety guarantee system for college students was established, the number of safety incidents decreased by about 52.2%, and the average student injury rate dropped from 1.735% to 0.54%. This paper confirms the actual contribution of the system in reducing safety incidents and lowering student injury rates, demonstrates the effective application of digital twin technology in campus safety management, and provides colleges and universities with a more intelligent and systematic safety assurance solution.