Assessment of the Sanitary Condition of Trees Using Neural Networks

In this research paper, methods for classifying the health status of trees are explored. To this end, the study analyzes various methods for determining the condition of plants and different machine learning models, including neural networks with convolutional layers. A dataset has been collected for training a neural network that predicts the sanitary condition of trees. Comparisons have also been made between various image classification models. The comparison was done using different metrics, and based on these, a model was selected that best determines the tree's condition from the input image. To work with a neural network, a software application with an interface is developed that accepts an image input and outputs the category to which the tree in the image belongs. The graduate qualification work consists of 59 pages, including 13 illustrations, 9 tables, 51 references and 2 appendices.