Thesis Archive
Diode And Fet Defect Localization Using Wavelet Transform (2007)
ENRIQUE FLORES, YSMAEL PUZON, JOSEPH VALDEZ, MARVIN SANTIAGO, advised by DR. FELICITO CALUYO
Abstract:
-This paper describes the localization and classification of the defects in diode and FET images with the use of wavelet analysis, independent component analysis and Euclidean distance measure implemented in the Matlab environment. Our program is composed of several modules that extract the feature vectors of the reference and sample image to be analyzed. The feature vector then serves as data for the whole process of the analysis. Further in this paper, the different modules will be discussed and some of the limitation of the program will be presented on the conclusion. The defects considered in our study are only limited to ESD (Electrostatic Discharge) and EOS (Electrical Over-Stress) and the differences between the two will be discussed later. In the end, the results will be presented for each defect classification for the purpose of analysis.
Correspondence:
Dr. Felicito Caluyo
[email protected]