Intelligent 3D vision measurement systems frequently encounter challenges when handling workpieces with complex and varied shapes, as well as those that are difficult to position. To address these challenges, we propose a global-view 3D reconstruction method that incorporates a cobot-driven independently developed line structured light sensor (LSLS). Initially, the cobot’s hand-eye calibration parameters are established using the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm. By utilizing real-time data interaction between the sensor and the cobot, this method facilitates the real-time automatic stitching of point clouds generated from multiple scans at different cobot orientations. Experimental outcomes demonstrate that the hand-eye calibration achieved an average root mean square error (RMSE) of 0.0950 mm, with an average computation time of 0.0857 ms. Furthermore, the average error for the 3D reconstruction of a standard sphere was 0.0983 mm.