In order to enable a fast, low-cost and reliable evaluation of solar cells, we propose an automated defect detection, using a deep convolutional neural network (CNN) for the EL cell image classification.
With the help of transfer learning, the accuracy of solar cell defect detection increases by 11.6%. We propose a ResNet-based micro-crack detection method to detect the micro-cracks on polycrystalline solar cells, including image preprocessing, feature extraction, featu...
Which ML-based techniques are used for surface defect detection of solar cells?
ML-based techniques for surface defect detection of solar cells were reviewed by Rana and Arora, of which were only imaging-based techniques. Similarly, Al-Mashhadani et al., have reviewed DL-based studies that adopted only imaging-based techniques.
How can computer vision and machine learning detect defects in solar cells?
Computer vision and machine learning techniques effectively detect defects in solar cells using EL images automatically. Cracks, inactive regions, and gridline faults have been the focus of statistical techniques, support vector machines (SVMs), and convolutional neural networks (CNNs) for fault detection and localization of various kinds.
What data analysis methods are used for PV system defect detection?
Nevertheless, review papers proposed in the literature need to provide a comprehensive review or investigation of all the existing data analysis methods for PV system defect detection, including imaging-based and electrical testing techniques with greater granularity of each category's different types of techniques.
(BAFPN) for solar defect detection. The BAFPN is an FPN. In their experiments, 3629 images were included, of which 2129 were detectable. The proposed methods have offer a practical solution in solar fault detections. were reported. Du et al. [ 26] proposed a deep CNN to enhance silicon photovoltaic (Si-PV) detection efficienc y.
Are solar cell defects detected by image classifiers?
various solar cell defects. Other image classifier models to detect and classify Si-PV cell faults. Another novel [ 28]. In this work, the short-term features represent denoising auto-encoder (SDAE). In contrast, the CNNs. This work concludes that such a combination of solar cells compared with other methods. and various defects.