A Fast and Differentiated Localization Method for Complex Surfaces Inspection

Although significant progress has been made in free-form surface inspection, it remains a difficult problem in some tough situations. For a complex part with multiple freeform surfaces, different quality requirements are often assigned to different surface regions to guarantee the part functionality and/or assembly relations. In this case, the existing localization methods, which treat each individual surface without discrimination, are prone to over rejection of some qualified parts. In this paper, a new localization method is proposed to tackle this problem. It features in that a novel dynamic weighting strategy is introduced in the iterative optimization process to make the localization in favor of the surfaces with higher quality requirements. This strategy can avoid improper scrap of parts in inspection and proves to be more reasonable for the quality evaluation of complex parts. In addition, a distinct fast algorithm is proposed to solve the localization optimization problem. It needs only to solve a 4-order linear system to determine the rigid transformation in each iteration. This solving algorithm is much faster than the popular quaternion-based algorithm for the weighted optimization model. Experiments on a simulated example and a real integral impeller are included to verify the validity and practicability of the proposed algorithm.