Honeycomb sandwich material is a new material widely used in many fields, but
it is easy to produce defects such as delamination and ponding in the
process of manufacturing and service. First, a honeycomb sandwich sample
containing delamination defects and water accumulation was built. Then, a
linear frequency modulated driving halogen lamp is used as the excitation
source. Finally, the surface thermal image sequence of the test sample is
acquired by infrared thermal imager. Image sequences are processed by
inter-frame difference-multi-frame cumulative average method, principal
component analysis, Fourier transform method, and logarithmic polynomial
fitting method, respectively. Define and calculate the signal-to-noise ratio
of the heat map processed by each algorithm. Compared with the other three
algorithms, the principal component analysis method processed the image with
the highest signal-to-noise ratio and high contrast. This algorithm achieves
effective identification of delamination defects and water accumulation in
GFRP/Nomex honeycomb sandwich structure.