Ensemble Tensor Factorization for Background Subtraction in Thermography NDT

This paper proposes an ensemble tensor factorization to extract defects signal of infrared thermography video. The proposed algorithm jointly models the soft weighted low rank and sparse tensor patterns as well as removing the ghosting. In particular, the weak defects information can be separated from the strong interference and the resolution contrast is significantly improved. A multi-layer ensemble iterative decomposition structure is conducted to further enhance the weak information. In order to verify the effectiveness and robustness of the proposed method, experimental studies have been carried out by applying electromagnetic thermal imaging system for cracks detection on samples with different geometry. The results of the experiments have indicated that the proposed method has significantly enhanced the contrast ratio between the defective regions and the non-defective regions.