Published Paper Details:

A complex challenge that might be helpful for many forestry-related jobs is identifying tree species using photographs of the bark. In this study, we introduce and make accessible to the public BarNet 1.0, a unique collection containing over 23,000 high-resolution bark photos captured at various tree diameters from 23 different types of trees. With it, we show that it is possible to recognize species using deep learning and pictures of bark. More precisely, we achieved an accuracy of 95.81%. This study aims to investigate different categorization methods for plant leaf identification. A dataset of 250 photos for each 874 plant species was gathered. For classification, four traditional machine learning methods were used: Support Vector Machine (SVM), Multilayer Perceptron (MLP), K-Nearest Neighbors (KNN), Decision Tree (DT), and MLP. In addition to color- and texture-based features, morphological features like physiological length, physiological width, centroid, area, diameter, perimeter, rectangularity, sphericity, aspect ratio, area convexity, and circularity were extracted to assess these classifiers' performance. The main objective was to determine the most efficient and successful machine-learning approach.