METHOD FOR PROCESSING IMAGES OF SEMICONDUCTOR STRUCTURES, AND FOR PROCESS CHARACTERIZATION AND PROCESS OPTIMIZATION BY MEANS OF SEMANTIC DATA COMPRESSION

摘要

The invention to methods for processing images of a semiconductor structure of a solar cell generated using imaging methods. Suitable methods comprise processing the image of the semiconductor structure by means of a trained neural network in order to generate a semantic representation of the image, wherein a data variable of the semantic representation is lower than a data variable of the image. Methods further comprise storing the semantic representation assigned to the semiconductor structure, as an information carrier of relevant information of the image for determining a property of the solar cell and/or for a crystal analysis of a semiconductor structure of the solar cell. Further aspects relate to methods for improving a semiconductor process, to methods for evaluating shaded regions in spatially resolved quality images of semiconductor structures, and to methods for predicting at least one property of a solar cell.