Adaptive automatic solar cell defect detection and classification based on absolute electroluminescence imaging

Current defect inspection methods for photovoltaic (PV) devices based on electroluminescence (EL) imaging technology lack juggling both labor-saving and in-depth understanding of defects, restricting the progress towards yield improvement and higher efficiency. Herein, we propose an adaptive approach for automatic solar cell defect detection and classification based on absolute EL imaging. Specifically, we first develop an unsupervised algorithm to automatically detect defects referring to the defect features in EL images. Then a diagnosis approach is proposed, which statistically classifies the detected defects based on the electrical origin. To the best of our knowledge, the proposed method is the first effort to integrate automatic defect detection with fine-grained classification. Experimental results on multiple types of solar cells show that the proposed method can achieve the average uncertainty of 5.15% at the minimum, with by up to 98.90% optimization ratio compared with two conventional methods. The proposed method is expected to provide more guiding feedback in both practical design and reliability diagnosis of the PV industry. An adaptive approach to automatically detect and classify defects in solar cells is proposed based on absolute electroluminescence (EL) imaging. We integrate the convenient automatic detection algorithm with the effective defect diagnosis solution so that in-depth defect detection and classification becomes feasible. To the best of our knowledge, the proposed method is the first effort to combine automatic defect detection with fine-grained classification based on electrical origin. • An automatic method is proposed for solar cell defect detection and classification. • An unsupervised algorithm is designed for adaptive defect detection. • A standardized diagnosis scheme is developed for statistical defect classification. • Extensive experimental results verify the effectiveness of the proposed method.