Project description:

         The manufacturer is prone to damage conventional modules of monocrystalline silicon solar cell chips during the production and transportation. Serious damages may affect the power generation efficiency and service life, which will be completely reflected in operation. Therefore, EL detection is required in its production, installation and even operation and maintenance. Based on the segmentation, classification and other algorithm modules of Handdle AI and AI technology, DSTEK proposes this EL solution to improve the detection accuracy while reducing the miss rate.

  Defect types:

     Hidden cracks, foggy blackening, sucker prints, graphite boat prints, concentric circles, virtual prints, dirt, black chips, hidden cracks, incomplete solder, edge blackening, black spots, cracked chips, etc.

Original picture                                                                                               Detection picture

           AI deep learning algorithm        Product standardization and low freely built threshold             Continuous optimization of model