作者: Matthew M. Nowell , Stuart I. Wright
DOI: 10.1016/J.ULTRAMIC.2004.11.012
关键词: Orientation (computer vision) 、 Artificial intelligence 、 Pattern recognition 、 Search engine indexing 、 Identification (information) 、 Crystallography 、 Computer science 、 Electron backscatter diffraction
摘要: Automated Electron Backscatter Diffraction (EBSD) has become a well-accepted technique for characterizing the crystallographic orientation aspects of polycrystalline microstructures. At advent this technique, it was observed that patterns obtained from grains in certain orientations were more difficult automated indexing algorithms to accurately identify than other orientations. The origin problem is often similarities between EBSD pattern correct and or phases. While practical solutions have been found implemented, identification these generally occurs only after running an scan, as are readily apparent resulting maps. However, such approach finds those present scan area. It would be advantageous all regions space may problems prior initiating minimize through optimization acquisition parameters. This work presents new methods identifying where reliability suspect performing scan. methodology used characterize impact various parameters on algorithm.