Automated pattern inspection system for PCB photomasks using design pattern comparison method

Author

Hamada, Toshimitsu ; Nakahata, Kozo ; Nomoto, Mineo ; Nakagawa, Yasuo ; Hashimoto, Yutaka ; Karasaki, Kouichi

fYear

1990

Firstpage

780

Abstract

An automated pattern inspection system for printed circuit board (PCB) photomasks has been developed. This system is a model-based system in which detected patterns are compared with design patterns to achieve a highly reliable inspection. Detected images are converted into binary patterns by selecting one of three thresholds. Defects are detected by a two-step image processing inspection. The first step is coarse alignment between the detected patterns and the design patterns. The second step is a local pattern comparison method in which small defects up to 1.5 pixel size can be detected without being influenced by pattern registration errors and sampling errors. A defect detection and pattern generation speed of 0.1 μs/pixel has been achieved. The minimum size of detectable defects is 15 μm, and it takes 14 min to inspect a photomask measuring 500×600 mm when pixel size is 10 μm

Keywords

automatic optical inspection; computerised picture processing; masks; printed circuit manufacture; 14 min; 15 micron; PCB photomasks; automated pattern inspection system; binary patterns; coarse alignment; defect detection; design pattern comparison method; local pattern comparison method; model-based system; pattern registration errors; sampling errors; two-step image processing; Design engineering; Design methodology; Feature extraction; Hardware; Humans; Image converters; Image edge detection; Inspection; Printed circuits; Reliability engineering;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial Electronics Society, 1990. IECON '90., 16th Annual Conference of IEEE

Conference_Location

Pacific Grove, CA

Print_ISBN

0-87942-600-4

Type

conf

DOI

10.1109/IECON.1990.149239

Filename

149239