The BOP control system can control the BOPs to shut down the well safely and ensure the safety of the drilling platform in case of overflow, well kick and blowout. In order to shut in the well successfully, it is necessary to manually identify the position of the drill pipe so that the tool joint can avoid the semi-closed ram blowout preventer. In case of emergency, operators are under great psychological pressure and prone to misjudgment and misoperation, which will lead to the failure of well shut-in and the expansion of dangerous situation, so it is urgent to use intelligent auxiliary identification and positioning system to ensure the successful implementation of safe shut-in. To solve this problem, this paper proposes a multi-camera perception information fusion method MCPIFM (Multi-camera perception information fusion method) to realize the intelligent recognition and precise positioning of tool joints. Multiple cameras are arranged on the drilling platform to detect the running state of the drill pipe, and the intelligent algorithm is used to identify and track the tool joint and calculate the joint position in real time, respectively. An identification evidence fusion framework is established by combining the YOLO identification algorithm and the D-S evidence theory, and the sensing information of the multiple cameras is fused to improve the accuracy and robustness of the drill pipe state monitoring and joint position calculation results. The system is built on the actual drilling platform and tested under different working conditions, which verifies the accuracy and reliability of the system, and the MCPIFS can be very convenient to achieve cross-platform deployment through automatic information calibration, which lays a solid foundation for the intelligent transformation of the well control safety system.