Implementing Thermal Image Processing Techniques for Enhancing the Detectability of Defects in Thermography of Additive Manufacturing Components

Nowadays, as the application of additive manufactured equipment is increasing in the industry, an appropriate inspection method for identifying defects of these products has become a pressing need. In this contribution, a study on inspection of the artificial defects of an additive manufactured specimen via thermography was carried out. A projector with 2KW in power was utilized as the heating source. The temperature of the sample was recorded by a thermal camera. Moreover, the camera kept recording the sample’s temperature for a while after that heating source was shut down. The best frame of raw thermal data was selected. To enhance the thermal raw data in case of the contrast between defective and sound regions and the number of detectable defects, two well-known thermal image processing methods, namely, Pulsed Phase Thermography (PPT) and Principle Component Analysis (PCA), were applied to the initial data. It was illustrated that all defects could be detected through processed images, whereas only 18 defects out of 20 could be revealed in the best frame of raw thermal data. Furthermore, for evaluating the ability of each technique to improve the contrast, the SNR parameter was adopted. According to the concluded data, the processed image via PCA with SNR average equal to 14.75 had the highest amount. This amount was almost three times higher than that of the best frame of initial thermal data.