Aircraft Inlet Duct Reconstruction and Measurement Methods Based on Point Cloud Splicing Optimization

The accurate reconstruction and measurement of aircraft inlet duct is extremely difficulty, due to its narrow structure, few surface features and large shape differences. To address these issues, inlet duct reconstruction and measurement methods based on point cloud splicing optimization is proposed in this paper, serving as a comprehensive full-process solution for industrial settings. Firstly, to complete the acquisition and coarse splicing of point clouds from the inner surface of the inlet duct, a point cloud coarse splicing system is designed using two unmanned vehicles as carriers. Subsequently, to optimize the splicing accuracy of point clouds, an iterative closest point registration method based on dynamically selecting overlapping domains (DSOD-ICP) is improved, and re-registration is achieved through the incorporation of adjacency relations detection. Finally, to mitigate the impact of point spacing on measurement results, a deviation measurement method based on the solution of projection distance from the centroid in normal cylindrical space (PDC-NCS) is proposed, which independently calculates the machining deviations in both positive and negative directions for each cross-sectional group. The experimental results demonstrate that, the absolute error of the measurement methods proposed in this paper is less than 0.3 mm compared with the actual model reconstructed by the T-Scan system. And the measurement accuracy and efficiency is improved by about 70% and 67% respectively compared with traditional manual measurement methods, which can provide a reliable solution for the measurement of machining deviations in aircraft inlet duct.