Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Mangrove Tree Species Classification Based on Leaf, Stem, and Seed Characteristics Using Convolutional Neural Networks with K-Folds Cross Validation Optimalization

Rektorat UDINUS 2023 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2023
Goals
Record
45fb66ee-7171-4d6a-9161-93c224953134
Published
2026-10-30

Universitas Dian Nuswantoro conducted an activity on mangrove tree species classification using convolutional neural networks and K-Folds cross-validation. The activity involved Farhan, Fadillah; Sari, Christy Atika; Rachmawanto, Eko Hari; and Cahyo, Nur Ryan Dwi as participants. The study used a meticulously curated dataset comprising diverse mangrove tree images standardized to 250 x 250 pixels. The images captured nuances of various mangrove species and provided the basis for applying deep learning techniques to classification.

The activity compared models developed without K-Folds and with K-Folds. The model without K-Folds achieved 99.23% accuracy, while the model with K-Folds achieved 99.78% accuracy. The models showed recall, precision, and F1-Score metrics exceeding 99%. Rigorous testing covered 10 experiments. In every classification scenario, both the K-Folds and non-K-Folds methods achieved 100% accuracy, with True Positives present throughout the testing. These results describe the reported consistency and robustness of the algorithms for classifying mangrove species.

The source presents the classification approach as a tool for ecological research and conservation initiatives. It also describes practical relevance for environmentalists, conservationists, and policymakers involved in mangrove preservation. Accurate species classification is stated to support understanding of biodiversity, targeted conservation efforts, and sustainable management of coastal ecosystems. Within the scope of the reported research, UDINUS frames this activity as institutional research addressing mangrove classification and its relevance to environmental monitoring, conservation, and ecological research.

Reference https://www.scopus.com/pages/publications/85199810964?origin=resultslist