International Journal of Precision Engineering and Manufacturing
- Volume 1 Issue 1
- /
- Pages.106-110
- /
- 2000
- /
- 2234-7593(pISSN)
- /
- 2005-4602(eISSN)
Ultrasonic Signal Analysis with DSP for the Pattern Recognition of Welding Flaws
-
Kim, Jae-Yeol
(School of Mechanical Engineering, Chosun University, Gwang-Ju, South Korea)
;
- Cho, Gyu-Jae (School of Mechanical Engineering, Chosun University, Gwang-Ju, South Korea) ;
- Kim, Chang-Hyun (School of Precision Mechanical Engineering, Chosun Graduate School, Gwang-Ju, South Korea)
- Published : 2000.06.01
Abstract
The researches classifying the artificial flaws in welding parts are performed using the pattern recognition technology. For this purpose the signal pattern recognition package including user defined function is developed and the total procedure is made up the digital signal processing, feature extraction, feature selection, classfier design. Specially it is composed with and discussed using the ststistical classfier such as the linear discriminant function classfier, the empirical Bayesian classfier.
Keywords
- Pattern recognition;
- digital signal processing;
- feature extraction;
- feature selection;
- bayesian classfier;
- planar and volumetric flaws;
- artificial flaws