Title :
PCB infrared thermal imaging diagnosis using support vector classifier
Author :
Jiuqing, Wan ; Xingshan, Li
Author_Institution :
Sch. of Autom. Sci. & Electr. Eng., Beijing Univ. of Aeronaut. & Astronaut., China
Abstract :
A printed circuit board (PCB) diagnosis system based on infrared thermal imaging is described. Multi-scale edge detection and nonlinear regression methods are used for heat source recognition and thermal feature extraction respectively. Thermal pattern recognition is implemented by support vector classifier instead of traditional back-propagation feed-forward network classifier and a good experimental result has been achieved.
Keywords :
edge detection; fault diagnosis; feature extraction; infrared imaging; learning automata; pattern classification; printed circuit manufacture; splines (mathematics); PCB infrared thermal imaging diagnosis; heat source recognition; multi-scale edge detection; nonlinear regression methods; printed circuit board; support vector classifier; thermal feature extraction; Automation; Circuit faults; Feature extraction; Image converters; Image edge detection; Infrared imaging; Optical imaging; Pattern recognition; Printed circuits; Temperature distribution;
Conference_Titel :
Intelligent Control and Automation, 2002. Proceedings of the 4th World Congress on
Print_ISBN :
0-7803-7268-9
DOI :
10.1109/WCICA.2002.1020010
