Abstract: To address the issues of large initial positioning errors and varying allowances for critical machined surfaces in the rough machining of large, complex parts, an allowance allocation method with process constraints and an adaptive positioning method based on feature-based positioning and contour matching are proposed. Firstly, the point cloud alignment is realized by sample consensus initial alignment (SAC-IA) algorithm based on fast point feature histogram (FPFH) features and iterative closest point (ICP) algorithm to complete the model pre-alignment. Secondly, the key process constraints of machining surface are integrated into the variance minimization of variance (VMM) model to optimize the margin distribution. Finally, the adaptive positioning parameters are generated based on the optimized position, and adaptive positioning is realized by the blank feature matching and contour positioning technology to form the the closed-loop process of “offline 3D scanning-point cloud alignment-allowance allocation-adaptive positioning”. Tests show that this method can meet the different machining surfaces of the blank differentiated margin constraints, while reducing the blank positioning manual intervention.