3DForEcoTech Workshop

The laser scanners use cameras and images to colorize point clouds. Photogrammetry is primarily using the images to create point clouds and then also to colorize them. We believe these images can be used for various classifications and enrich the information that we can get from the collected data. That is why we are focusing on tree species classification based on tree bark using RGB images. The first experiment considers four tree species. In this experiment, we want to show the potential of the classification. We have considered the following methods: Random forest classifier, K-nearest Neighbours (KNN), Decision Tree, Gradient Boost, Support Vector Machine (SVM), Gaussian Naive Bayes, and Convolutional Neural Networks (CNN), Multilayer Perceptron (MLP).