Using polynomials to correct non-uniform backgrounds in thermal images caused by uneven heating

In order to correct non-uniform backgrounds in thermal images of carbon fiber-reinforced polymers (CFRP), we propose a processing and analysis procedure in this paper. We apply thermographic signal reconstruction (TSR) using time-direction information to reduce noise in thermal images, followed by Whittaker smoothing (WS) to further reduce noise in two-dimensional image space. Finally, polynomial background correction (PBC) is used to correct non-uniform backgrounds retained in the reconstructed thermal images. We tested a CFRP specimen with seven defective regions using the proposed method. The results showed that with the aid of PBC, the non-uniform background caused by uneven heating was successfully corrected and the defective regions were easily identified.