Publication Details

A graphical automated detection system to locate hardwood log surface defects using high-resolution three-dimensional laser scan data

Publication Toolbox

  • Download PDF (921412)
  • This publication is available only online.
Thomas, Liya; Thomas, R. Edward

Year Published

2011

Publication

In: Fei, Songlin; Lhotka, John M.; Stringer, Jeffrey W.; Gottschalk, Kurt W.; Miller, Gary W., eds. Proceedings, 17th central hardwood forest conference; 2010 April 5-7; Lexington, KY; Gen. Tech. Rep. NRS-P-78. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station: 92-101.

Abstract

We have developed an automated defect detection system and a state-of-the-art Graphic User Interface (GUI) for hardwood logs. The algorithm identifies defects at least 0.5 inch high and at least 3 inches in diameter on barked hardwood log and stem surfaces. To summarize defect features and to build a knowledge base, hundreds of defects were measured, photographed, and categorized. Our cost-effective, reliable, and robust statistical method locates surface defects using a commercially available high-resolution laser scanner. The scanned data capture three-dimensional (3-D) images that illustrate log shapes—protrusions and depressions—and portray bark patterns. Most available optimization systems were designed to determine gross external characteristics (e.g., shape, diameter, sweep) of softwood logs. They then better position the log with respect to the saw and improve the sawyer’s decision-making ability. Adding external defect information to the optimization process is a natural extension of current technology. Our method applies decision rules obtained from the knowledge base and identifies severe defects. Using the defect detection GUI, we tested our system. Using 14 randomly selected logs, we detected 53 severe defects out of 64. Nine nondefective regions were falsely identified. When defects were calculated as areas, 90.2 percent of defects were located.

Note: This article is part of a larger document. View the larger document

Citation

Thomas, Liya; Thomas, R. Edward. 2011. A graphical automated detection system to locate hardwood log surface defects using high-resolution three-dimensional laser scan data. In: Fei, Songlin; Lhotka, John M.; Stringer, Jeffrey W.; Gottschalk, Kurt W.; Miller, Gary W., eds. Proceedings, 17th central hardwood forest conference; 2010 April 5-7; Lexington, KY; Gen. Tech. Rep. NRS-P-78. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station: 92-101.

Last updated on: June 14, 2011