Given the exceptional qualities of 3D printed concrete materials, the printed models often suffer from inadequate strength. To overcome this limitation and ensure higher durability, it is crucial to integrate reinforcement measures into the structure. However, manual reinforcement with rebar is not only inefficient but also poses significant safety concerns. To address these challenges, we have devised a novel approach that utilizes modular reinforcement workpieces, placed by depth camera and robotic arms. The approach of YOLOv8 is used for identifying and categorizing the workpieces. By integrating this with a plane grasping network model, the optimal grasping posture for each workpiece can be accurately inferred. The posture results are then transmitted to the robotic arm via the Robot Operating System (ROS) architecture, enabling precise and efficient grasping. Several grasping experiments are carried out and achieved good performance. This validates the feasibility of the proposed approach.