Does varifocalnet detect photovoltaic module defects?

The VarifocalNet is an anchor-free detection method and has higher detection accuracy 5. To further improve both the detection accuracy and speed for detecting photovoltaic module defects, a detection method of photovoltaic module defects in EL images with faster detection speed and higher accuracy is proposed based on VarifocalNet.

What is a crystalline silicon PV module?

The crystalline silicon PV module contains glass on the surface, polymers in encapsulant and back sheet foil, aluminum in the frame, silicon in solar cells, copper in interconnectors, silver in contact lines, and other heavy metals such as tin and lead. Another manufacturing of PV is from thin-film technology.

Why do photovoltaic modules have a small feature map?

The photovoltaic module defect is relatively smaller, so the smaller feature map will reduce the detection accuracy. To increase the perceptual field without reducing the feature map size, we design a new bottleneck module that consists of standard convolution and dilated convolution.

How to improve the detection speed of photovoltaic module defects?

Improving detection speed is the focus of the one-stage method, while the two-stage method emphasizes detection accuracy. In the practical detection of photovoltaic module defects, we should consider not only the detection speed but also the detection accuracy. The VarifocalNet is an anchor-free detection method and has higher detection accuracy 5.

Do anomaly PV modules affect power production?

The anomaly PV modules in the dataset affected a total 21,793 kW DC power production with respect to the 16.93% affection of the total anomalies in the real-world findings. This is a useful IR PV module dataset that reflects the real state of anomalies in worldwide PV farms. It is expected to increase about 30% of the total dataset each year .

How does a multi-scale network efficiently classify photovoltaic panel anomalies?

The Multi-scale network efficiently classifies photovoltaic panel anomalies. Oversampling approach overcomes the imbalanced class distribution. Multi-scale branches aim to improve the features extracted by each parallel block. Proposed method is validated with a large dataset collected from six continents.