Utilisation of convolutional neural network on deep learning in predicting digital image to tree damage type

Damage is often defined as a condition where there is a change in an object or shape. Damage does not only occur to objects, it can occur to living things, including trees. Damage to trees is seen in the physical shape of the tree that has changed shape, to find out it requires deep learning. One way that can be used is by modelling through computers through artificial intelligence, namely creating a deep learning model that can retrieve image information to recognise objects. This research aims to utilise convolutional neural network algorithm in deep learning to identify tree damage. The research stages carried out are input, feature extraction and classification, and output. The final result obtained is the successful identification of trees in deep learning on the model with an accuracy of 99.06% with a detection error of 0.94%. Detection errors occur due to similarities in terms of patterns, etc. This can be minimised by combining hyperparameters. Copyright © 2024 Inderscience Enterprises Ltd.