Fairing skin repair method based on measured wing data

A fairing skin repair method based on measured wing data, comprising:

• S1) carrying out a fairing skin registration, comprising:
• S101) obtaining skin point cloud data Q;
• S102) performing denoising and voxel grid filtering on a known point cloud data of docked wing to obtain a data set P1;
• reorganizing the data set P1 through a data reorganization method to obtain a key point set P;
• and calculating a normal line of each key point in the key point set P;
• S103) calculating a histogram feature descriptor in a normal direction of any key point in the key point set P and a histogram feature descriptor in a normal direction of any point in the skin point cloud data Q, respectively;
• S104) calculating a Euclidean distance between feature descriptors of two points through a K-nearest neighbor algorithm;
• searching similar histogram feature descriptors;
• adding points with high similarity to a set M;
• and initially setting the set M as an empty set;
• S105) performing a clustering on the set M using a Hough voting algorithm to obtain a local point cloud set P′ in the key point set P that matches the skin point cloud data Q;
• and
• S106) matching the skin point cloud data Q with the local point cloud set P′ through an iterative closest point algorithm;
• and
• S2) carrying out a fairing skin repair.