作者: Ryuichi Funatsu , Hiroshi Miyai , Yasuhiko Nara , Taku Ninomiya
DOI:
关键词: Feature (computer vision) 、 Inspection method 、 Reference image 、 Halftone 、 Cathode ray 、 Artificial intelligence 、 Sample (graphics) 、 Secondary electrons 、 Computer vision 、 Mathematics 、 Optics 、 Pixel
摘要: An inspection method and apparatus irradiates a sample on which pattern is formed with an electron beam, so that image reference can be generated the basis of secondary or reflected emitted by sample. abnormal determined based difference in halftone values each pixel between image. A plurality feature quantities are obtained from pattern, and, distribution range for classifying type designated. Thus, desired defect extracted many defects inspection.