Feature Extraction and Classification in Automated Inspection of NDE Images
Date
Authors
Major Professor
Advisor
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Authors
Research Projects
Organizational Units
Journal Issue
Is Version Of
Versions
Series
Begun in 1973, the Review of Progress in Quantitative Nondestructive Evaluation (QNDE) is the premier international NDE meeting designed to provide an interface between research and early engineering through the presentation of current ideas and results focused on facilitating a rapid transfer to engineering development.
This site provides free, public access to papers presented at the annual QNDE conference between 1983 and 1999, and abstracts for papers presented at the conference since 2001.
Department
Abstract
Industry currently relies on manual inspection of X-Ray images, which is expensive, time consuming and subject to inconsistent results by various inspectors. Automatic inspection of the industrial NDE images using high speed computers would make the flaw detection task more consistent and efficient. However, since the objects to be inspected usually have complex geometric structures, it is very difficult to separate automatically flaws from the complex geometric background of an image. Most of the automated inspection systems developed so far are customized packages which are tailored for particular types of applications, such as the inspections of aluminum wheel castings [1], the welds of space shuttle fuel tanks [2], and nodules in chest X-rays [3]. Customized packages have the advantages of relatively fast speed and good liability, but they can not be used in other applications without substantial changes in computer programs.