Detección del tizón tardío en folíolos de papa usando imágenes tomadas con dron y técnicas de Deep Learning

J. A. Cairampoma; Delia Gamarra-Gamarra; F. E.  Dionisio

doi:10.17268/sci.agropecu.2026.020

Autores

J. A. Cairampoma Universidad Nacional del Centro del Perú, Huancayo, Perú. https://orcid.org/0000-0002-1998-6178
Delia Gamarra-Gamarra Universidad Nacional del Centro del Perú, Huancayo, Perú. https://orcid.org/0000-0002-6262-237X
F. E. Dionisio Universidad Nacional del Centro del Perú, Huancayo, Perú. https://orcid.org/0000-0002-9769-6782

DOI:

https://doi.org/10.17268/sci.agropecu.2026.020

Palavras-chave:

Detección del tizón tardío, Dron, Red neuronal convolucional, Mask R-CNN, Folíolos de papa

Resumo

Phytophthora infestans causa una de las enfermedades más devastadoras del cultivo de papa, también conocida como tizón tardío. Dado que la identificación temprana de este patógeno es crucial para el control efectivo de la enfermedad, este estudio tuvo como objetivo proponer una metodología automatizada para la identificación de sus lesiones en foliolos de papa, utilizando redes neuronales convolucionales llamadas "Mask R-CNN". Las evaluaciones se llevaron a cabo durante la temporada de lluvias, en cultivos realizados por agricultores en la localidad de Huasahuasi, en los Andes centrales del Perú. Se tomaron cien fotografías (5472 × 3078 píxeles) con un vehículo aéreo no tripulado (UAV) Phantom 4 Pro a una altitud de 3 m en cultivos con una incidencia de tizón tardío entre 2 y 3. Las imágenes se dividieron en cuatro partes y luego pasaron un riguroso control de calidad, dando como resultado 200 fotos (1825 × 1369 píxeles). Del total, el 75% se utilizó para el entrenamiento del modelo y el 25% para su validación. Los modelos se evaluaron en condiciones reales, utilizando métricas como la precisión y la recuperación. Se determinó que la red neuronal Mask R-CNN, basada en la arquitectura de red neuronal profunda ResNet 101, ofrece una precisión y efectividad aceptables (73,5%) en la identificación de lesiones de tizón tardío a nivel de foliolo. Esta metodología constituye una contribución significativa a la agricultura de precisión en los Andes, al validar una herramienta no invasiva capaz de superar las limitaciones topográficas de la zona. Su aplicación práctica optimizaría el uso de fungicidas mediante la detección dirigida, promoviendo así sistemas de producción de papa más sostenibles y rentables para los agricultores locales.

Referências

Abbas, A., Zhang, Z., Zheng, H., Alami, M. M., Alrefaei, A. F., Abbas, Q., Naqvi, S. A. H., Rao, M. J., Mosa, W. F. A., & Abbas, Q. (2023). Drones in plant disease assessment, efficient monitoring, and detection: a way forward to smart agriculture. Agronomy, 13(6), 1524. https://doi.org/10.3390/agronomy13061524

Abdulla, W. (2021). Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow. https://github.com/matterport/Mask_RCNN

Alanazi, R. (2025). A YOLOv10-based Approach for Banana Leaf Disease Detection. Engineering, Technology & Applied Science Research, 15(3), 23522-23526. https://doi.org/10.48084/etasr.11138

Baba, A. (2024). Neural networks from biological to artificial and vice versa. Biosystems, 235, 105110. https://doi.org/https://doi.org/10.1016/j.biosystems.2023.105110

Babli, D., Yadav, P. S., & Sheeren Parveen, V. K. (2022). Efficacy of different eco-friendly methods against late blight of potato, Phytophthora infestans: A review. Pharma Innov. J, 11, 1949-1961.

Bai, X., Wang, X., Liu, X., Liu, Q., Song, J., Sebe, N., & Kim, B. (2021). Explainable deep learning for efficient and robust pattern recognition: A survey of recent developments. Pattern Recognition, 120, 108102. https://doi.org/https://doi.org/10.1016/j.patcog.2021.108102

Berhan, M. (2021). Review on epidemiology, sampling techniques, management strategies of late blight (Phytophthora infestans) of potato and its yield loss. Asian Journal of Advances in Research, 4(1), 199-207. https://doi.org/10.1007/978-3-030-28683-5_1

Bondre, S., y Patil, D. (2024). Crop disease identification segmentation algorithm based on Mask-RCNN. Agronomy Journal, 116(3), 1088-1098. https://doi.org/https://doi.org/10.1002/agj2.21387

Devaux, A., Goffart, J.-P., Petsakos, A., Kromann, P., Gatto, M., Okello, J., Suarez, V., & Hareau, G. (2020). Global food security, contributions from sustainable potato agri-food systems. The potato crop: Its agricultural, nutritional and social contribution to humankind, 3-35. https://doi.org/10.1007/978-3-030-28683-5_1

Ditzler, C., Scheffe, K., & Monger, H. C. (2017). Soil survey manual. USDA Handbook 18. Government Printing Office.

Duarte-Carvajalino, J., Alzate, D., Ramirez, A., Santa-Sepulveda, J., Fajardo-Rojas, A., Soto-Suárez, M., Duarte-Carvajalino, J. M., Alzate, D. F., Ramirez, A. A., Santa-Sepulveda, J. D., Fajardo-Rojas, A. E., & Soto-Suárez, M. (2018). Evaluating Late Blight Severity in Potato Crops Using Unmanned Aerial Vehicles and Machine Learning Algorithms. Remote Sensing, 10(10), 1513. https://doi.org/10.3390/rs10101513

Feng, J., Hou, B., Yu, C., Yang, H., Wang, C., Shi, X., & Hu, Y. (2023). Research and Validation of Potato Late Blight Detection Method Based on Deep Learning. Agronomy, 13(6), 1659. https://doi.org/10.3390/agronomy13061659

Ganesh, P., Volle, K., Burks, T. F., & Mehta, S. S. (2019). Deep Orange: Mask R-CNN based Orange Detection and Segmentation. IFAC-PapersOnLine, 52(30), 70-75. https://doi.org/10.1016/j.ifacol.2019.12.499

Grados, D., García, S., & Schrevens, E. (2020). Assessing the potato yield gap in the Peruvian Central Andes. Agricultural Systems, 181, 102817. https://doi.org/https://doi.org/10.1016/j.agsy.2020.102817

Hindarto, D. (2023). Comparison of Detection with Transfer Learning Architecture RestNet18, RestNet50, RestNet101 on Corn Leaf Disease. Jurnal Teknologi Informasi Universitas Lambung Mangkurat (JTIULM), 8(2), 41-48. https://doi.org/10.20527/jtiulm.v8i2.174

Ivanov, A. A., Ukladov, E. O., & Golubeva, T. S. (2021). Phytophthora infestans: An Overview of Methods and Attempts to Combat Late Blight. Journal of Fungi, 7(12), 1071. https://doi.org/10.3390/jof7121071

Jadhav, P., Kachave, V., Mane, A., & Joshi, K. (2023). Crop detection using satellite image processing. I-Manager’s Journal on Image Processing, 10(2), 50-60. https://doi.org/10.26634/jip.10.2.19800

Kunduracioglu, I., & Pacal, I. (2024). Advancements in deep learning for accurate classification of grape leaves and diagnosis of grape diseases. Journal of Plant Diseases and Protection, 131(3), 1061-1080. https://doi.org/10.1007/s41348-024-00896-z

Lin, P., Zhang, H., Zhao, F., Wang, X., Liu, H., & Chen, Y. (2022). Boosted Mask R-CNN algorithm for accurately detecting strawberry plant canopies in the fields from low-altitude drone images. Food Science and Technology, 42, e95922. https://doi.org/10.1590/fst.95922

Majeed, A., Muhammad, Z., Ullah, Z., Ullah, R., & Ahmad, H. (2017). Late blight of potato (Phytophthora infestans) I: Fungicides application and associated challenges. Turkish Journal of Agriculture-Food Science and Technology, 5(3), 261-266. https://doi.org/10.24925/turjaf.v5i3.261-266.1038

Matsuo, Y., LeCun, Y., Sahani, M., Precup, D., Silver, D., Sugiyama, M., Uchibe, E., & Morimoto, J. (2022). Deep learning, reinforcement learning, and world models. Neural Networks, 152, 267-275. https://doi.org/https://doi.org/10.1016/j.neunet.2022.03.037

MK, D., & Matharasi, P. B. (2025). Leaf Disease Predictions Using Deep Learning Techniques - Potato. International Journal of Information Technology, Research and Applications, 4(3), 33-42. https://doi.org/10.59461/ijitra.v4i3.166

Ortiz, O., Garrett, K. A., Health, J. J., Orrego, R., & Nelson, R. J. (2004). Management of potato late blight in the Peruvian highlands: evaluating the benefits of farmer field schools and farmer participatory research. Plant Disease, 88(5), 565-571. https://doi.org/10.1094/PDIS.2004.88.5.565

Padilla, R., Netto, S. L., & da Silva, E. A. B. (2020). A Survey on Performance Metrics for Object-Detection Algorithms. 2020 International Conference on Systems, Signals and Image Processing (IWSSIP), 237-242. https://doi.org/10.1109/IWSSIP48289.2020.9145130

Perez, W., Forbes, G. A., Arias, R., Pradel, W., Kawarazuka, N., & Andrade-Piedra, J. (2022). Farmer Perceptions Related to Potato Production and Late Blight Management in Two Communities in the Peruvian Andes. Frontiers in Sustainable Food Systems, 6, 873490. https://doi.org/10.3389/fsufs.2022.873490

Phan, Q.-H., Nguyen, V.-T., Lien, C.-H., Duong, T.-P., Hou, M. T.-K., & Le, N.-B. (2023). Classification of tomato fruit using yolov5 and convolutional neural network models. Plants, 12(4), 790. https://doi.org/10.3390/plants12040790

Ren, S., He, K., Girshick, R., & Sun, J. (2016). Faster R-CNN: Towards real-time object detection with region proposal networks. IEEE transactions on pattern analysis and machine intelligence, 39(6), 1137–1149. https://doi.org/10.1109/TPAMI.2016.2577031

Rosebrock, A. (2017). Deep Learning for Computer Vision with Python: ImageNet Bundle. PyImageSearch.

Shelhamer, E., Long, J., & Darrell, T. (2017). Fully convolutional networks for semantic segmentation. IEEE transactions on pattern analysis and machine intelligence, 39(4), 640-651.

Sinamenye, J. H., Chatterjee, A., & Shrestha, R. (2025). Potato plant disease detection: leveraging hybrid deep learning models. BMC Plant Biology, 25(1), 647. https://doi.org/10.1186/s12870-025-06679-4

Wang, F., Wang, C., Song, S., Xie, S., & Kang, F. (2021). Study on starch content detection and visualization of potato based on hyperspectral imaging. Food Science & Nutrition, 9(8), 4420-4430. https://doi.org/https://doi.org/10.1002/fsn3.2415

Yan, J., Wang, H., Yan, M., Diao, W., Sun, X., & Li, H. (2019). IoU-Adaptive Deformable R-CNN: Make Full Use of IoU for Multi-Class Object Detection in Remote Sensing Imagery. Remote Sensing, 11(3). https://doi.org/10.3390/rs11030286

Yu, M., Ma, X., Guan, H., Liu, M., & Zhang, T. (2022). A Recognition Method of Soybean Leaf Diseases Based on an Improved Deep Learning Model. Frontiers in Plant Science, 13. https://doi.org/10.3389/fpls.2022.878834

Yu, Y., Zhang, K., Yang, L., & Zhang, D. (2019). Fruit detection for strawberry harvesting robot in non-structural environment based on Mask-RCNN. Computers and Electronics in Agriculture, 163, 104846. https://doi.org/10.1016/j.compag.2019.06.001

Zand, M., Etemad, A., y Greenspan, M. (2022). Objectbox: From centers to boxes for anchor-free object detection. European Conference on Computer Vision, 390-406.

Zevallos, E., Inga, J., Alvarez, F., Marmolejo, K., Paitan, R., Viza, I., Becerra, D., Rixi, G., & Silva-Diaz, C. (2021). First signs of late blight resistance in traditional native potatoes of Pasco—Peru, a preliminary assay. Agriculture & Food Security, 10(1), 33. https://doi.org/10.1186/s40066-021-00330-9