Efficient Image Segmentation for Detection of Dislocations in High Resolution Light Microscope Images of SiC Wafers

Abstract

The determination of dislocation density and in particular the dislocation distribution in SiC wafers is of particular interest for SiC crystal growth development and production. We present an image recognition tool allowing the wafer analysis with specific needs for SiC. In the first stage of expansion, micropipes are selected and counted from SiC wafers that have been etched by KOH.