Vessels Detection Based on Neural Network with Application in Wood Recognition

For the conservation of plant species and the control of timber commerce across the world, tree identification is becoming increasingly crucial. It relies heavily on anatomical traits such as vessels, fibers, parenchyma, and rays. The cell structure of each hardwood species differs greatly across intraspecific species and defines their differences. In this article, we propose a method for the detection of vessels through two steps: feature extraction and classification. They are based on grayscale image entropy and classification using a neural network.