Inteligencia artificial para la conservación forestal: identificación de especies maderables mediante MobileNetV3 y mecanismos de atención en Ecuador
DOI:
https://doi.org/10.18800/kawsaypacha.202602.A014Palabras clave:
Redes neuronales convolucionales, Computación en el borde, Aprendizaje supervisado, Aplicación de la ley, Botánica, EcologíaResumen
La tala ilegal y la deforestación —tráfico de madera— representan amenazas críticas para la sostenibilidad de los ecosistemas forestales en el sur de Ecuador, lo que exige herramientas de control más eficientes y accesibles. Este trabajo presenta un modelo basado en inteligencia artificial (IA) que utiliza la arquitectura MobileNetV3 —especializada en la detección de patrones en imágenes y optimizada para dispositivos de bajo poder de cómputo— para la identificación anatómica de cuatro especies maderables: cedro (Cedrela odorata), nogal (Juglans neotropica), faique (Prosopis pallida) y guayacán (Handroanthus chrysanthus). Los resultados evidenciaron un desempeño adecuado y estable en diferentes escenarios de validación, incluyendo condiciones controladas de laboratorio y entornos con mayor variabilidad y complejidad, alcanzando una precisión del 81% y del 80%, respectivamente, lo que demuestra que el modelo aprendió a discriminar las especies a partir de características anatómicas como vasos, radios y anillos. Se concluye que la solución basada en IA constituye una herramienta tecnológica para apoyar la lucha contra el tráfico de madera y, al mismo tiempo, un valioso recurso didáctico para la formación de futuros profesionales en computación, gestión forestal y conservación ambiental.
Descargas
Referencias
Ajay, A.; Sandosh, S.; Saji, A. & Agarwal, H. (2025). An Explainable Deep Learning Framework for Sorghum Weed Classification Using Multi-Scale Feature Enhanced DenseNet. IEEE Access, 13, 26973-26990. https://doi.org/10.1109/ACCESS.2025.3538937
Alhichri, H.; Alswayed, A. S.; Bazi, Y.; Ammour, N. & Alajlan, N. A. (2021). Classification of Remote Sensing Images Using EfficientNet-B3 CNN Model With Attention. IEEE Access, 9, 14078-14094. https://doi.org/10.1109/ACCESS.2021.3051085
Datta, D.; Simeone, J. C.; Meadows, A. Outhwaite, W., Keong Chen, H., Self, N., Walker, L., & Ramakrishnan, N. (2025). Combating trade in illegal wood and forest products with machine learning. PLOS ONE, 20(1), pp. 1-24. https://doi.org/10.1371/journal.pone.0311982
De Andrade, B. G.; Basso, V. M. & De Figueiredo Latorraca, J. V. (2020). Machine vision for field-level wood identification. IAWA Journal, 41(4), pp. 681-698. https://doi.org/10.1163/22941932-bja10001
Devore, J. L. (2008). Probabilidad y estadística para Ingeniería y ciencias (7.a ed.). Cengage Learning Editores, S.A. de C.V., una compañía de Cengage Learning, Inc.
Diaz-de-Arcaya, J.; Torre-Bastida, A. I.; Zárate, G., Miñón, R. & Almeida, A. (2024). A Joint Study of the Challenges, Opportunities, and Roadmap of MLOps and AIOps: A Systematic Survey. ACM Computing Surveys, 56(4), pp. 1-30. https://doi.org/10.1145/3625289
Fei, Y.; Li, Z.; Zhu, T. & Ni, C. (2023). A Lightweight Attention-Based Convolutional Neural Networks for Fresh-Cut Flower Classification. IEEE Access, 11, 17283-17293. https://doi.org/10.1109/ACCESS.2023.3244386
García-Cox, W.; López-Tobar, R.; Herrera-Feijoo, R. J.; Tapia, A.; Heredia-R, M;, Toulkeridis, T. & Torres, B. (2023). Floristic Composition, Structure, and Aboveground Biomass of the Moraceae Family in an Evergreen Andean Amazon Forest, Ecuador. Forests, 14(7), pp. 1-16. https://doi.org/10.3390/f14071406
Gayathri, R.; Manohar, S.; Dayana, R.; Mathivanan, S. K. & Deshpande, N. M. (2025). An Optimized Algorithmic Frameworks for Transfer Learning to Enhance Efficiency and Accuracy in Pretrained Model Adaptation. 2025 4th OPJU International Technology Conference (OTCON) on Smart Computing for Innovation and Advancement in Industry 5.0, Raigarh, India, pp. 1-5. https://doi.org/10.1109/OTCON65728.2025.11070524
Harika, M. R.; Reddy, P. K.; Jannu, C.; Parisae, V.; Krishna, C. V. M. & Madhumati, G. L. (2024). Real Time Speech Enhancement Using Triple Attention U-Net. 2024 4th International Conference on Artificial Intelligence and Signal Processing (AISP), Vijayawada, India, 1-5. https://doi.org/10.1109/AISP61711.2024.10870819
Herrera-Feijoo, R. J.; Torres, B.; López-Tobar, R.; Tipán-Torres, C.; Toulkeridis, T.; Heredia-R. M. & Mateo, R. G. (2023). Modelling Climatically Suitable Areas for Mahogany (Swietenia macrophylla King) and Their Shifts across Neotropics: The Role of Protected Areas. Forests, 14(2), pp. 1-23. https://doi.org/10.3390/f14020385
Herrera-Poyatos, D.; Poyatos, A. H.; Soldado, R. M.; De Palacios, P.; Esteban, L. G.; Iruela, A. G.; Fernández, F. G. & Herrera, F. (2024). Deep Learning methodology for the identification of wood species using high-resolution macroscopic images. 2024 International Joint Conference on Neural Networks (IJCNN), Yokohama, Japan, 1-8. https://doi.org/10.1109/IJCNN60899.2024.10650590
Howard, A.; Sandler, M.; Chen, B.; Wang, W.; Chen, L.-C.; Tan, M.; Chu, G.; Vasudevan, V.; Zhu, Y.; Pang, R.; Adam, H. & Le, Q. (2019). Searching for MobileNetV3. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 1314-1324. https://doi.org/10.1109/ICCV.2019.00140
International Association of Wood Anatomists (1989). IAWA list of microscopic features for hardwood identification with an appendix on non-anatomical information (Boletín No. 10; pp. 219-332). http://bit.ly/3MavfcC
Islam, M.; Aloraini, M.; Habib, S.; Alanazi, M. D.; Khan, I. & Khan, A. (2024). Optimal Features Driven Attention Network With Medium-Scale Benchmark for Wheat Diseases Recognition. IEEE Access, 12, 150739-150753. https://doi.org/10.1109/ACCESS.2024.3434575
Jiao, R.; Xu, Z. & Zhang, Z. (2024). Improving YOLOv8 via Triple Attention for Pedestrian Detection. 2024 5th International Conference on Computer Vision, Image and Deep Learning (CVIDL), 1517-1521. https://doi.org/10.1109/CVIDL62147.2024.10604206
Kaleab Taye, A.; Park, J.-M.; Cho, H.-J.; Kang, J.-T. & Seo, Y.-O. (2025). ResTreeNet: A Height-Aware LiDAR Tree Classification Model With Explainable AI for Forestry Applications. IEEE Access, 13, 81392-81405. https://doi.org/10.1109/ACCESS.2025.3567042
Karnik, A. S.; Nair, N.; Sagili, Y. & Shanthi, P. B. (2025). Multi-Scale Venation Pattern Analysis for Medicinal Plant Species Recognition. IEEE Access, 13, 125526-125536. https://doi.org/10.1109/ACCESS.2025.3589278
Kreuzberger, D.; Kühl, N. & Hirschl, S. (2023). Machine Learning Operations (MLOps): Overview, Definition, and Architecture. IEEE Access, 11, 31866-31879. https://doi.org/10.1109/ACCESS.2023.3262138
Krishnamoorthy, M.; Sjoding, M. & Wiens, J. (2025). Cross-Validation for Longitudinal Datasets with Unstable Correlations. Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.2, New York, NY, USA, 1219-1228. https://doi.org/10.1145/3711896.3736898
Kumar, B.; Sinha, R. S.; Burman, R. K.; Alam, Md. S.; Kumari, A. & Sah, R. D. (2024). TreeIDNet: Species Identification through Video Analysis using Detectron2-Based Computer Vision. 2024 5th International Conference for Emerging Technology (INCET), Belgaum, India, 1-7. https://doi.org/10.1109/INCET61516.2024.10592969
Li, H.; Li, N.; Wang, W.; Yang, C.; Chen, N. & Deng, F. (2024). Convolution Self-Guided Transformer for Diagnosis and Recognition of Crop Disease in Different Environments. IEEE Access, 12, 165903-165917. https://doi.org/10.1109/ACCESS.2024.3495529
Li, W.; Ma, L. & Sciarrone, A. (2025). Robust Multi-Agent Collaborative Perception via Triple-Attention and Dynamic Gating. 2025 IEEE 101st Vehicular Technology Conference (VTC2025-Spring), 1-5. https://doi.org/10.1109/VTC2025-Spring65109.2025.11174503
Liu, S.; Zheng, C.; Wang, J.; Lu, Y.; Yao, J.; Zou, Z.; Yin, Y. & He, T. (2024). How to discriminate wood of CITES-listed tree species from their look-alikes: Using an attention mechanism with the ResNet model on an enhanced macroscopic image dataset. Frontiers in Plant Science, 15, pp. 1-14. https://doi.org/10.3389/fpls.2024.1368885
Mascarello, M.; Amalfi, M.; Asselman, P.; Smets, E.; Hardy, O. J.; Beeckman, H. & Janssens, S. B. (2021). Genome skimming reveals novel plastid markers for the molecular identification of illegally logged African timber species. PLOS ONE, 16(6), pp. 1-18. https://doi.org/10.1371/journal.pone.0251655
Ministerio del Ambiente, Agua y Transición Ecológica (MAATE) (2018). Estadísticas del patrimonio natural del Ecuador continental. https://doi.org/10.5281/zenodo.18158663
Ministerio del Ambiente, Agua y Transición Ecológica (MAATE) (2023). Rendición de cuentas. https://doi.org/10.5281/zenodo.17866986
Mou, X.; Chen, H.; Yu, X.; Chen, L.; Peng, Z. & Wang, R. (2024). Lightweight Segmentation Method for Wood Panel Images Based on Improved DeepLabV3+. Electronics, 13(23), pp. 1-13. https://doi.org/10.3390/electronics13234658
Nguyen, P. Q.; Truong, D. T. M.; Huynh, T. V.; Nguyen, T. H. N. & Le, Q.-H. (2023). Design and Build a System for Illegal Logging Detection Based on Acoustic Signal and Automatic Warning Using LoRA Network. 2023 International Conference on System Science and Engineering (ICSSE), Ho Chi Minh, Vietnam, 265-270. https://doi.org/10.1109/ICSSE58758.2023.10227201
Ojeda Luna, T.; Eguiguren, P.; Günter, S.; Torres, B. & Dieter, M. (2020). What Drives Household Deforestation Decisions? Insights from the Ecuadorian Lowland Rainforests. Forests, 11(11), pp. 1-20. https://doi.org/10.3390/f11111131
Prommakhot, A.; Onshaunjit, J.; Ooppakaew, W.; Samseemoung, G. & Srinonchat, J. (2025). Hybrid CNN and Transformer-Based Sequential Learning Techniques for Plant Disease Classification. IEEE Access, 13, 122876-122887. https://doi.org/10.1109/ACCESS.2025.3586285
Raju, R. R.; Naidu, S. N.; Moulali, U. & Basha, S. R. (2025). Hardwood Classification using ResNet18 and Macroscopic Image Analysis. 2025 4th International Conference on Innovative Mechanisms for Industry Applications (ICIMIA), Tirupur, India, 1920-1927. https://doi.org/10.1109/ICIMIA67127.2025.11200564
Ravindran, P. & Wiedenhoeft, A. C. (2022). Caveat emptor: On the Need for Baseline Quality Standards in Computer Vision Wood Identification. Forests, 13(4), pp. 1-19. https://doi.org/10.3390/f13040632
Riesco Muñoz, G.; Imaña Encinas, J. & De Paula, J. E. (2019). Densidad de la madera de 59 especies del orden Sapindales procedentes de bosques naturales brasileños. Madera y Bosques, 25(2), pp. 1-12. https://doi.org/10.21829/myb.2019.2521817
Senthilvadivu, S. & David, A. (2025). Vision Transformers with Contrastive Learning for Robust Wood Species Identification. 2025 3rd International Conference on Self Sustainable Artificial Intelligence Systems (ICSSAS), Erode, India, 1085-1090. https://doi.org/10.1109/ICSSAS66150.2025.11081278
Silva, J.; Bordalo, R.; Pissarra, J. & De Palacios, P. (2022). Computer Vision-Based Wood Identification: A Review. Forests, 13(12), pp. 1-26. https://doi.org/10.3390/f13122041
Souza, D. V.; Santos, J. X.; Vieira, H. C.; Naide, T. L.; Nisgoski, S. & Oliveira, L. E. S. (2020). An automatic recognition system of Brazilian flora species based on textural features of macroscopic images of wood. Wood Science and Technology, 54(4), pp. 1065-1090. https://doi.org/10.1007/s00226-020-01196-z
Srinivasan, S.; Prabin, S. M.; Mathivanan, S. K.; Rajadurai, H.; Kulandaivelu, S. & Shah, M. A. (2025). Sugarcane leaf disease classification using deep neural network approach. BMC Plant Biology, 25(1), pp. 1-20. https://doi.org/10.1186/s12870-025-06289-0
Sugiarto, B.; Suaib, N. M.; Mohd Rahim, M. S.; Prakasa, E.; Fairos Ismail, N. A. & Albakri Mustafa Albakri, I. F. (2024). Wood Species Classification Based on Wood Porosity Distribution of Macroscopic Image. 2024 IEEE International Symposium on Consumer Technology (ISCT), Kuta, Bali, Indonesia, 230-235. https://doi.org/10.1109/ISCT62336.2024.10791236
Tang, F.; Porle, R. R.; Tung Yew, H. & Wong, F. (2025). Identification of Maize Diseases Based on Dynamic Convolution and Tri-Attention Mechanism. IEEE Access, 13, 6834-6844. https://doi.org/10.1109/ACCESS.2025.3525661
Venugopal, D.; Sandhiya, R.; Swetha, S.; Sredhar, R.; Rohit, N. & Finney Daniel, S. (2024). Machine Learning-based Algorithm for Type of Wood Identification. 2024 International Conference on Emerging Research in Computational Science (ICERCS), Coimbatore, India, 1-6. https://doi.org/10.1109/ICERCS63125.2024.10895161
Wang, D. & Zhang, Z. (2025). A Comparison Study of Image Data Augmentation Techniques in Computer Vision. 2025 5th International Conference on Computer Vision, Application and Algorithm (CVAA), Ningbo, China, 410-414. https://doi.org/10.1109/CVAA66438.2025.11193215
Wang, G.; Chen, J.; Mo, L.; Wu, P. & Yi, X. (2024). Border-Enhanced Triple Attention Mechanism for High-Resolution Remote Sensing Images and Application to Land Cover Classification. Remote Sensing, 16(15), pp. 1-21. https://doi.org/10.3390/rs16152814
Xie, Y. & Ling, J. (2023). Wood defect classification based on lightweight convolutional neural networks. BioResources, 18(4), 7663-7680. https://doi.org/10.15376/biores.18.4.7663-7680
Xie, Y.; Zheng, J.; Gou, A.; Sattar, F. & Liao, L. (2025). Log End Face Feature Extraction and Matching Method Based on Swin Transformer V2. Forests, 16(1), pp. 1-17. https://doi.org/10.3390/f16010124
Zhang, Z.; Liu, T.; Gao, J.; Yang, M.; Luo, W. & Lin, F. (2025). TDR-Model: Tomato Disease Recognition Based on Image Dehazing and Improved MobileNetV3 Model. IEEE Access, 13, pp. 852-865. https://doi.org/10.1109/ACCESS.2024.3522101
Descargas
Publicado
Número
Sección
Licencia
Derechos de autor 2026 Jorge Daniel Ortega Alburqueque, Luis Antonio Chamba Eras
Esta obra está bajo una licencia internacional Creative Commons Atribución 4.0.
