Scan-to-CAD alignment of damaged airfoil blade point clouds through geometric dissimilarity assessment

This paper presents a method for accurate alignment of the scanned point clouds of damaged blades with their nominal CAD model, which is an essential task in automated inspection for remanufacturing. The geometric dissimilarity of the underlying surface of the local neighborhoods of each measured data point and its nearest corresponding point on the CAD model is evaluated using a metric combining the average curvature Hausdorff distance and average Euclidean Hausdorff distance. The algorithm eliminates unreliable pairs with high geometric dissimilarity values in damaged regions from the matching process. The effectiveness of the proposed method is verified by experimental tests.