Graduation Semester and Year

2011

Language

English

Document Type

Thesis

Degree Name

Master of Science in Electrical Engineering

Department

Electrical Engineering

First Advisor

Michael T Manry

Abstract

Defect segmentation has been a focal point in silicon wafer inspection research and it remains challenging because the defects are complicated by large variations in intensity distribution. An algorithm for silicon wafer defect segmentation is developed using a modified pulse coupled neural network (PCNN). The modified PCNN is simple version of the PCNN in which segmentation depends only on the linking coefficient and initial threshold. The initial threshold and linking coefficient are determined automatically from image statistics using method described in [17] and Otsu's method respectively. The modified PCNN method was found to be simple and efficient for silicon wafer defect segmentation. The performance of the modified PCNN is better than the Otsu's method or a standalone PCNN. Results have been presented for all the four types of silicon defect.

Disciplines

Electrical and Computer Engineering | Engineering

Comments

Degree granted by The University of Texas at Arlington

Share

COinS