Weakly Supervised Segmentation of Cracks on Solar Cells using Normalized Lp Norm
This work addresses the need for fast and scalable quality management in photovoltaic production and operation, though it appears incremental as it builds on existing weakly supervised segmentation methods.
The paper tackled the problem of segmenting cracks on electroluminescence images of solar cells using only image-level annotations, achieving promising results with a weakly supervised learning strategy based on a modified ResNet-50 and normalized Lp normalization.
Photovoltaic is one of the most important renewable energy sources for dealing with world-wide steadily increasing energy consumption. This raises the demand for fast and scalable automatic quality management during production and operation. However, the detection and segmentation of cracks on electroluminescence (EL) images of mono- or polycrystalline solar modules is a challenging task. In this work, we propose a weakly supervised learning strategy that only uses image-level annotations to obtain a method that is capable of segmenting cracks on EL images of solar cells. We use a modified ResNet-50 to derive a segmentation from network activation maps. We use defect classification as a surrogate task to train the network. To this end, we apply normalized Lp normalization to aggregate the activation maps into single scores for classification. In addition, we provide a study how different parameterizations of the normalized Lp layer affect the segmentation performance. This approach shows promising results for the given task. However, we think that the method has the potential to solve other weakly supervised segmentation problems as well.