No abstract

No abstract

No abstract

No abstract

Non-destructive testing (NDT) methods are commonly used to disclose defective inner structures of materials, where active infrared thermography is a popular NDT technique because of its easy operation, low cost, and the capability of rapid scanning of large areas. However, the thermographic signals often suffer from noise and non-uniform backgrounds, making defect identification difficult. Therefore, an additional processing step of thermographic data is often required to enhance the contrast between defects and their surroundings. In recent years, multivariate statistical analysis methods have been widely adopted to achieve this aim, among which principal component thermography (PCT) is a typical example. Sparse PCT (SPCT) further improves PCT by adding sparsity constraints to the optimization problem. Nevertheless, the performance of SPCT depends on the subjective selection of the tuning parameters. The optimal parameter values are case-dependent and unknown. In this work, an alternative thermographic data analysis method is proposed based on exploratory factor analysis (EFA). By means of factor rotation, EFA minimizes the complexity of factor loadings and makes the results more interpretable. In doing this, EFA provides results similar to those of SPCT, while there is no need for parameter selection. Experimental results illustrate the feasibility of the proposed method.