Future directions in industrial machine vision: a case study of semiconductor manufacturing applications

摘要: Machine vision systems are being increasingly used for sophisticated applications such as classification and process control. Vision have access to richer information (colour/texture/depth), employ more powerful hardware than their predecessors. Though there is significant potential the increased deployment of systems, a number important problems be addressed sustain growth in area industrial machine vision. This paper identifies some these future research directions which show promise solving problems. Some examined include non-conventional imaging modes scanning electron microscopy atomic force microscopy, visual defect classification, integration colour, texture depth information, configurability changing field requirements.