An integrated image processing approach for 3D scanning and micro-defect detection

Abstract

Restoration of high-value components via additive manufacturing requires autonomous surface scanning and defect identification. The 3D free-form surface can be reconstructed with a point cloud obtained from the scanning. Laser line triangulation-based surface scanning is a promising method for generating a 3D point cloud of the component surface. In this article, a robotic defect scanning system developed using py_openshowvar, an open-source cross-platform communication interface is presented. For effective scanning of micro-scale features with minimal noise, it is crucial to optimize the scanning parameters. The scanner parameters such as exposure time and stand-off distance have been optimized for accurate feature detection. After selecting optimal scanning parameters, a generic algorithm is presented for generating a scanning path for automatic scanning of the 3D parts. Surfaces with pre-fabricated micro-defects are automatically scanned using this algorithm, and an integrated image-processing-based defect identification technique is presented. The geometries obtained from the presented technique were validated using focus variation microscopy, and the results are in good agreement with actual defect geometry, and the measurement error is below 9%.