Mode selection and defect classification based on convolutional neural networks for image synthesis

摘要

Systems and methods are disclosed for classifying defects using a hot scandal and convolutional neural network (CNN). The primary scanning modes are identified by the processor and a hot scan of the wafer is performed. Defects of interest and newsons data are selected and images of these areas are captured using one or more secondary scanning modes. Image sets are collected and divided into subsets. The CNN is trained using image subsets. The ideal secondary scanning mode is determined and a final hot scan is performed. The defects are filtered and classified according to the final hot scan and ideal secondary scanning mode. The disclosed systems to classify defects use a scanning electron microscope as well as an image data acquisition subsystem such as a processor and electronic database.