A visual inspection method for delamination extraction and quantification of carbon fiber reinforced plastic (CFRP)

Inspection the delamination of composite components is essential to ensure the safety of aircraft in service. Vision inspection has great potential for the drilling quality inspection of large composite components. However, stable extraction and accurate quantification of delamination are extremely challenging, due to the characteristics of low contrast, varying morphology, and random distribution of delamination. To address this problem, robust delamination extraction algorithm and integrated evaluation method are proposed. The extraction of delamination is achieved by contour extraction, robust circle fitting, and adaptive threshold segmentation. Based on this, the statistical quality description factor Qd_s is designed to quantify multi-dimensional information of delamination. Finally, the effectiveness of the proposed method is demonstrated. The experimental result shows that the proposed method outperforms traditional methods, which is expected to become a general technique for the inspection of delamination in the manufacturing process of CFRP.