Image based identification of Ghanaian timbers using the XyloTron: opportunities, risks and challenges

Abstract

Computer vision systems for wood identifcation have the potential to empower both producer and consumer countries to combat illegal logging if they can be deployed effectively in the feld. In this work, carried out as part of an active international partnership with the support of UNIDO, we constructed and curated a feld-relevant image data set to train a classifer for wood identifcation of 15 commercial Ghanaian woods using the XyloTron system. We tested model performance in the laboratory, and then collected real-world feld performance data across multiple sites using multiple XyloTron devices. We present effcacies of the trained model in the laboratory and in the feld, discuss practical implications and challenges of deploying machine learning wood identifcation models, and conclude that feld testing is a necessary step - and should be considered the gold-standard for validating computer vision wood identifcation systems.

Citation

Ravindran, Prabu; Ebanyenle, Emmanuel; Ebeheakey, Alberta Asi; Abban, Kofi Bonsu; Lambog, Ophilious; Soares, Richard; Costa, Adriana; Wiedenhoeft, Alex C. 2019. Image based identification of Ghanaian timbers using the Xylotron: opportunities, risks and challenges. In: 33rd Conference on Neural Information Processing Systems (NeurIPS 2019), Vancouver, Canada.