Online Program Home
My Program

Abstract Details

Activity Number: 347 - Machine Learning and Applications in Complex Engineering Systems
Type: Contributed
Date/Time: Tuesday, July 31, 2018 : 10:30 AM to 12:20 PM
Sponsor: Section on Physical and Engineering Sciences
Abstract #330014 Presentation
Title: The Identification and Quantification of Pits, Cracks, and Corrosion from Container Material Image Surface Depth Measurements with Subsequent Container Classification
Author(s): James Wendelberger*
Companies: Los Alamos National Laboratory
Keywords: destructive sample; image analysis; corrosion; pitting; cracks; feature statistics

Hazardous materials or conditions may affect the integrity of their container. A sample is taken destructively from the interior surface of a container. Depth measurements on the sample are made using a Laser Conformal Microscope, LCM, to create an image of the surface height or depth over the sample area. The depth information is used to define void features of interest. The features of interest consist of pixels of the image that form contiguous areas with similar depths. A feature is characterized by various statistics. For each feature, the statistics include: number of pixels and the corresponding aggregate pixel area, maximum depth, average depth, volume below the surface, bounding box, major, minor axis and related eccentricity, the orientation of the major axis, the topological "holes" and other specific shape information, such as, exact absolute or relative locations of the forming pixels. The feature statistics may be summarized for all features in the image. The summary statistics of the image may be used to classify the image as interesting, in the sense of requiring further inspection, or uninteresting, in the sense that there are no unacceptable features present.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program