Title :
Neural decision directed segmentation of silicon defects
Author :
Godbole, Aditi S. ; Tyagi, Kanishka ; Manry, Michael T.
Author_Institution :
Dept. of Electr. Eng., Univ. of Texas at Arlington, Arlington, TX, USA
Abstract :
A system is proposed for recognizing four types of defects present in silicon wafer images. After preprocessing, the system applies four segmentation algorithms, one per defect type. Approximate posterior probabilities from a multilayer perceptron classifier aid in fusing the segmentors and making the final defect classification. Numerical results confirm the feasibility of our approach.
Keywords :
automatic optical inspection; elemental semiconductors; image classification; image recognition; image segmentation; multilayer perceptrons; probability; production engineering computing; silicon; Si; approximate posterior probabilities; final defect classification; multilayer perceptron classifier; neural decision directed segmentation; segmentation algorithms; segmentor fusion; silicon defect recognition; silicon wafer images; Copper; Feature extraction; Filtering algorithms; Image segmentation; Matrix converters; Plastics; Silicon;
Conference_Titel :
Neural Networks (IJCNN), The 2013 International Joint Conference on
Conference_Location :
Dallas, TX
Print_ISBN :
978-1-4673-6128-6
DOI :
10.1109/IJCNN.2013.6706915
