Hybrid of Tsr and Seeded Region Growing for Debonding Detection Using Optical Thermography

Qizhi Feng,Bin Gao,Yang,Peng Lu,Jian Zhao,X. Q. Li,Xueshi Qiu,Liangyong Gu,Guiyun Tian
DOI: https://doi.org/10.1109/fendt.2017.8584602
2017-01-01
Abstract:Carbon fiber reinforced polymer (CFRP) are commonly used in the field of aerospace. Under manufacturing or in service procedure, there exists internal defects such as delamination and debonding due to the factors of improper production and environment. In order to guarantee CFRP internal quality and safety, the optical pulsed thermography (OPT) nondestructive testing has been used to detect the internal defects. However, the current OPT related methods has problems of uneven illumination and low resolution of defects detection. In this paper, the hybrid of thermographic signal reconstruction (TSR) and seeded region growing (SRG) algorithm was proposed to deal with the infrared thermal image sequences of the CFRP specimen, which can significantly enhance the detection rate. Finally, the event based F-score is computed to measure the detection results and comparison studies show that the proposed method can improve the performance of the detection.
What problem does this paper attempt to address?