A Study on the Terahertz Nondestructive Testing of Foam Sandwich Structure Composites Based on Morphological Filtering

Zhilong Li,Jian Zuo,Yuanmeng Zhao,Zhongde Han,Zhihao Xu,Yunzhang Zhao,Chao Yang,Weidong Hu,Cunlin Zhang
DOI: https://doi.org/10.1117/12.2606883
2021-01-01
Abstract:A new method for defect recognition of foam sandwich structure composites based on morphological filtering is proposed in this paper. Firstly, the terahertz frequency modulated radar imaging system is used to detect the foam sandwich structure composites to obtain the original detection signal, and the original detection image is obtained by using amplitude imaging method. Secondly, the original detection images are transformed to binary images by setting threshold values. Thirdly, based on the morphological filtering theory we carry out closing operation to fill the inner tiny holes in the binary images, and connect the discontinuous adjacent targets. Finally, the opening operation is used to remove the redundant small objects in the image, and obtain the final detection result image. In this paper, the imaging detection of polymethacrylimide foam sandwich structure composites is carried out, and the original detection signal is processed by using the above method, the resulting images with the prefabricated defect effectively recognized is obtained.The accuracy of the proposed method for defect recognition is evaluated by calculating and comparing the error ratio of the defect area in the processed image and the original detection image relative to the actual defect area.The results show that the proposed method can effectively recognized the prefabricated defects in foam sandwich structure composites and optimize the size of the defects to be close to the actual defects, which is conducive to the further application of terahertz nondestructive testing technology in foam sandwich structure composites.
What problem does this paper attempt to address?