Funded Research Projects: SRM

Machine Vision for Classification of Spruce, Pine, and Fir

Research Sponsored By: US Forest Service
Principal Investigator: Indroneil Ganguly
Project Description
Expertise in wood species identification is an that is declining in the United States. However, the need for identification has grown in recent years. The Xylo Tron developed out this need for wood species identification. To date, the Xylo Tron has targeted angiosperms from Central and South America. This research seeks to investigate the efficacy of the Xylo Tron to classify gymnosperms from North America. Specifically, the objective is to determine the ability to separate spruce, pine, and fir; a common commercial grouping. The tasks are to prepare the end grain, image, seek features from the images, and then train the machine learning algorithms. Wood samples from the Forest Products Laboratory (MADw and SJRw) and Burke Museum collection (WTUFw) wood collections will be used as the imaging samples. The outcome of this research will be a publication on the efficacy of the Xylo Tron to classify North American spruce, pine, and fir and recommendations of alterations to optimize the Xylo Tron for gymnosperms.