Imaging offers a fast and accessible means for spatial characterization of halide perovskite photovoltaic materials,yet extracting optoelectrical properties—such as power conversion efficiency(PCE)—remains challengi...Imaging offers a fast and accessible means for spatial characterization of halide perovskite photovoltaic materials,yet extracting optoelectrical properties—such as power conversion efficiency(PCE)—remains challenging.This study presents a deep learning methodology that correlates optical reflective images of perovskite solar cells with their PCE by focusing on image differences rather than absolute visual features.The approach predicts relative changes in PCE by comparing images of the same device in different states(e.g.,before and after encapsulation)or against a reference image.This comparative technique significantly outperforms traditional methods that attempt to directly infer PCE from a single image.Furthermore,it demonstrates high effectiveness in low-data regimes,using only 115 samples.By leveraging convolutional neural networks(CNNs)trained on small datasets,the method offers an adaptable and scalable solution for device characterization.Overall,the comparative approach enhances the accuracy and applicability of machine vision in perovskite solar cell analysis.展开更多
基金EXC 2089:e-conversion DFG-cluster of excellence-TUM innovation network,Technical University of Munich,funded through the German Excellence Initiative and the state of Bavaria(A.G.,M.H.).Support by the project ProperPhotoMile is also gratefully acknowledged(A.G.,I.V-F.),under the umbrella of SOLAR-ERA.NET Cofund 2 by The Spanish Ministry of Science and Education and the AEI under the project PCI2020-112185 and CDTI project number IDI-20210171the Federal Ministry for Economic Affairs and Energy on the basis of a decision by the German Bundestag project number FKZ03EE1070B and FKZ 03EE1070A+2 种基金and the Israel Ministry of Energy with project number 220-11-031SOLAR-ERA.NET is supportedby the European Commission within the EU Framework Programme for Research and Innovation HORIZON 2020(Cofund ERA-NETAction,N 786483)D.K.K.is grateful for the Blaustein postdoctoral fellowship at BGU.R.K.G.is grateful for the Swiss Inst.of Dryland Environmental and Energy Research postdoctoral fellowship at BGU.The authors are also grateful for partial support by the Israel Ministry of Energy,project number 222-11-081.
文摘Imaging offers a fast and accessible means for spatial characterization of halide perovskite photovoltaic materials,yet extracting optoelectrical properties—such as power conversion efficiency(PCE)—remains challenging.This study presents a deep learning methodology that correlates optical reflective images of perovskite solar cells with their PCE by focusing on image differences rather than absolute visual features.The approach predicts relative changes in PCE by comparing images of the same device in different states(e.g.,before and after encapsulation)or against a reference image.This comparative technique significantly outperforms traditional methods that attempt to directly infer PCE from a single image.Furthermore,it demonstrates high effectiveness in low-data regimes,using only 115 samples.By leveraging convolutional neural networks(CNNs)trained on small datasets,the method offers an adaptable and scalable solution for device characterization.Overall,the comparative approach enhances the accuracy and applicability of machine vision in perovskite solar cell analysis.
