In the process of aircraft wing-body docking， the outer surface of the wing-body butt joint needs to be covered with a rectifying skin to maintain the aerodynamic shape of the aircraft. In order to ensure that the gap of the rectifying skin can meet the design requirements， it is necessary to accurately repair the allowance. Therefore， we propose a technology to extract aircraft skin repair allowance based on 3D point cloud processing. Firstly， the scanned wing-body docking data are preprocessed to obtain the key points. Then the histogram feature descriptors are calculated for the key points and model skinning data to find similar descriptors. Finally， the Hough voting algorithm and the iterative closest point algorithm are used to match the data of the model skin and the wing-body docking. When the repair allowance is extracted， the boundary lines of the wing and fuselage skin at the wing body and the end face of the rough skin joint are first detected. Then the boundary lines of the wing and fuselage skin are projected onto the rough skin to calculate its distance from the boundary lines. Hence， the final repair allowance of skin is extracted. This method realizes the automatic registration and repair of the rectifying skin， which greatly reduces the problems of manpower， material resources， and time consumption， and has high application value for actual aircraft assembly.