Performance Evaluation of Imaging System of Electroluminescent Defect Detector for Photovoltaic
In order to improve the accuracy and repeatability of line pair image discrimination in the performance evaluation process of the electroluminescent defect detector imaging system.Refer to relevant standards,a special algorithm is used to realize automatic discrimination of line pair images,and the results are compared with manual discrimination by human eyes.At the same time,a model is established,and the method of using Monte Carlo method to evaluate the uncertainty of line pair image is studied,and the influence of three factors on algorithm discrimination results are analyzed,including the machining accuracy of the spatial resolution test board,the image tilt angle,and the uniformity of image.The results show that the accuracy and repeatability of the algorithm discrimination are better than manual discrimination by human eyes,with the accuracy of 98%and the repeatability of 100%.For the example given,the uncertainty of the algorithm discrimination is 0.0387 lp/mm.In general,compared with human eye discrimination,algorithms discrimination can greatly improve the accuracy and repeatability of line pair image discrimination in the performance evaluation process of the electroluminescent defect detector imaging system,and the uncertainty of the discrimination results is low.
PhotovoltaicElectroluminescent defect detectorSpatial resolutionMonte carlo method