Automated Labeling of Log Features in CT Imagery of Multiple Hardwood Species
Keywords:
Industrial inspection, segmentation, computed tomography, image analysis, log processingAbstract
Before noninvasive scanning, e.g., computed tomography (CT), becomes feasible in industrial sawmill operations, we need a procedure that can automatically interpret scan information in order to provide the saw operator with information necessary to make proper sawing decisions. To this end, we have worked to develop an approach for automatic analysis of CT images of hardwood logs. Our current approach classifies each pixel individually using a feed-forward artifical neural network (ANN) and feature vectors that include a small, local neighborhood of pixels and the distance of the target pixel to the center of the log. Initially, this ANN was able to classify clear wood, bark, decay, knots, and voids in CT images of two species of oak with 95% pixel-wise accuracy. Recently we have investigated other ANN classifiers, comparing 2-D versus 3-D neighborhoods and species-dependent (single species) versus species-independent (multiple species) classifiers using oak (Quercus rubra L. and Q. nigra L.), yellow-poplar (Liriodendron tulipifera L.), and black cherry (Prumus serotina Ehrh.) CT images. When considered individually, the resulting species-dependent classifiers yield similar levels of accuracy (96-98%). 3-D neighborhoods work better for multiple-species classifiers, and 2-D is better for the single-species case. Classifiers combining yellow-poplar and cherry data misclassify many pixels belonging to splits as clear wood, resulting in lower classification rates. If yellow-poplar was not paired with cherry, however, we found no statistical difference in accuracy between the single-and multiple-species classifiers.References
Aune, J. 1995. The development of a log scanner for sawmills in Canada. In O. Lindgren, ed. 2nd International Seminar on Scanning Technology and Image Processing on Wood. Dept. of Wood Technology, Lulcå University, Skellefteå, Sweden.nBenson-Cooper, D. M, R. L. Knowles, F. J. Thompson, and D. J. Cown. 1982. Computed tomographic scanning for the detection of defects within logs. Bull. No. 8 Forest Research Institute, New Zealand Forest Service, Rotorua, NZ. 9 pp.nBirkeland, R., and S. Holoyen. 1987. Industrial methods for internal scanning of log defects: A progress report on an ongoing project in Norway. In R. Szymani, ed. 2nd International Conference on Scanning Technology in Sawmilling, 1-2 October, Oakland/Berkeley Hills, CA. Forest Industries/World Wood, San Francisco, CA.nBurgess, A. E. 1985. Potential applications of medical imaging techniques to wood products. In R. Szymani, ed. 1st International Conference on Scanning Technology in Sawmilling, October 10-12, San Francisco, CA. Forest Industries/World Wood, San Francisco, CA.nCown, D. J., and B. C. Clement. 1983. A wood densitometer using direct scanning with x-rays. Wood Sci. Technol. 17(2):91-99.nDavis, J. R., and P. Wells. 1992. Computed tomography measurements on wood. Ind. Metrology 2(3/4):195-218.nFunt, B. V., and E. C. Bryant. 1987. Detection of internal log defects by automatic interpretation of computer tomography images. Forest Prod. J. 37(1):56-62.nGrönlund, A. 1992. Benefits from knowing the interior of the log. In O. Lindgren, ed. 1st International Seminar on Scanning Technology and Image Processing on Wood, August 30-September 1, Skellefteå, Sweden. Dept. of Wood Technology, Luleå University, Skellefteå Sweden. 7 pp.nGrundberg, S., and A. Grönlund. 1992. Log scanning — Extraction of knot geometry. In O. Lindgren, ed. 1st International Seminar on Scanning Technology and Image Processing on Wood, August 30-September 1; Skellefteå, Sweden. Dept. of Wood Technology, Luleå University, Skellefteá, Sweden. 11 pp.nHarless, T. E. G., F. G. Wagner, P. H. Steele, F. W. Taylor, V. Yadama, and C. W. McMillin. 1991. Methodology for locating defects within hardwood logs and determining their impact on lumber-value yield. Forest Prod J. 41(4):25-30.nHodges, D. G., W. C. Anderson, and C. W. McMillin. 1990. The economic potential of CT scanners for hardwood sawmills. Forest Prod. J. 40(3):65-69.nHopkins, F., I. L. Morgan, H. Ellinger, and R. Klinksiek. 1982. Tomographic image analysis. Mater. Eval. 40(20):1226-1228.nLi, P. 1996. Automatic interpretation of computer tomography (CT) images for hardwood log defect detection. MS thesis, Virginia Tech, Blacksburg, VA.nLi, P., A. L. Abbott, and D. L. Schmoldt. 1996. Automated analysis of CT images for the inspection of hardwood logs. In Proc. 1996 IEEE International Conference on Neural Networks. Institute for Electrical and Electronics Engineers, Inc., Piscataway, NJ.nLi, P., J. He, A. L. Abbott, and D. L. Schmoldt. 1997. Labeling defects in CT images of hardwood logs with species-dependent and species-independent classifiers. Pages 113-126 in A. Prinz and W. Pölzleitner, eds. Proc. International Association of Pattern Recognition Workshop on Machine Perception Applications, 2-3 September 1996, Technical University Graz, Austria. R. Oldenbourg, Vienna, Austria.nLindgren, L. O. 1991. Medical CAT-scanning: X-ray absorption coefficients, CT-numbers and their relation to wood density. Wood Sci. Technol. 25:341-349.nOcceña, L. G. 1991. Computer integrated manufacturing issues related to the hardwood log sawmill. J. Forest Eng. 3(1):39-45.nOcceña, L. G., D. L. Schmoldt, and S. Thawornwong. 1997. Using internal defect information for log breakdown. Pages 63-68 in J. Dennig, ed. ScanPro: Advanced technology for sawmilling. Miller-Freeman, San Francisco, CA.nOnoe, M., J. W. Tsao, H. Yamada, H. Nakamura, J. Kogura, H. Kawamura, and M. Yoshimatsu. 1984. Computed tomography for measuring the annual rings of a live tree. Nuclear Instr. Methods Physics Res. 221(1):213-220.nRichards, D. B., W. K. Adkins, H. Hallock, and E. H. Bulgrin. 1980. Lumber value from computerized simulation of hardwood log sawing. Res. Pap. FPL-356. USDA, Forest Service, Forest Products Lab, Madison WI, 10 pp.nRoder, F. 1989. High speed CT scanning of logs. In R. Szymani, ed. 3rd International Conference on Scanning Technology in Sawmilling, October 5-6, San Francisco, CA. Forest Industries/World Wood, San Francisco, CA.nSchmoldt, D. L. 1996. CT imaging, data reduction, and visualization of hardwood logs. In D. Meyer, ed. Proc. 1996 Hardwood Research Symposium. National Hardwood Lumber Association, Memphis, TN.nSchmoldt, D. L., P. Li, and A. L. Abbott. 1997. Machine vision using artificial neural networks and 3D pixel neighborhoods. Computers and Electronics in Agriculture 16(3):255-271.nSchmoldt, D. L., T. E. G. Harless, P. H. Steele, F. W. Taylor, V. Yadama, and C. W. McMillin. 1990. Potential benefits of internal-log scanning. Pages 77-88 in Proc. Process Control/Production Management of Wood Products: Technology for the 90's, October 30-November 1, Athens GA. The University of Georgia, Athens, GA.nZhu, D., R. W. Conners, F. M. Lamb, D. L. Schmoldt, and P. A. Araman. 1991a. A computer vision system for locating and identifying internal log defects using CT imagery. In R. Szymani, ed. 4th International Conference of Scanning Technology in Sawmilling, Oct. 28-31, San Francisco, CA. Forest Industries/World Wood, San Francisco, CA. 13 pp.nZhu, D., R. W. Conners, L. Schmoldt, and P. A. Araman. 1991b. CT image sequence analysis for object recognition—A rule-based 3-d computer vision system. Pages 173-178 in 1991 IEEE International Conference on Systems, Man, and Cybernetics, October 13-16, Charlottesville, VA.nZhu, D., R. W. Conners, L., and P. A., Araman. 1991c. CT image sequence processing for wood defect recognition, in Proc 23rd Southeast Symposium on System Theory, March, Columbia, SC. 7 pp.nZhu, D., R. W. Conners, L., D. L. Schmoldt, and P. A. Araman. 1996. A prototype vision system for analyzing CT imagery of hardwood logs. IEEE Transactions on Systems, Man, and Cybernetics 26(4):522-532.n
Downloads
Published
Issue
Section
License
The copyright of an article published in Wood and Fiber Science is transferred to the Society of Wood Science and Technology (for U. S. Government employees: to the extent transferable), effective if and when the article is accepted for publication. This transfer grants the Society of Wood Science and Technology permission to republish all or any part of the article in any form, e.g., reprints for sale, microfiche, proceedings, etc. However, the authors reserve the following as set forth in the Copyright Law:
1. All proprietary rights other than copyright, such as patent rights.
2. The right to grant or refuse permission to third parties to republish all or part of the article or translations thereof. In the case of whole articles, such third parties must obtain Society of Wood Science and Technology written permission as well. However, the Society may grant rights with respect to Journal issues as a whole.
3. The right to use all or part of this article in future works of their own, such as lectures, press releases, reviews, text books, or reprint books.
