Clasificación no destructiva de plantas medicinales peruanas mediante imágenes hiperespectrales NIR (1300–1700 nm) y aprendizaje automático
DOI:
https://doi.org/10.15649/2346075X.5680Palabras clave:
NIR-HSI, aprendizaje automático, plantas medicinales,, clasificación supervisada, control de calidadResumen
Introducción. La identificación de plantas medicinales nativas suele depender de métodos destructivos, lentos o altamente dependientes de experiencia especializada. Este estudio propone el uso de imágenes hiperespectrales en el infrarrojo cercano (NIR-HSI) y aprendizaje automático como herramientas alternativas, no invasivas, para la discriminación rápida de tres especies medicinales del norte del Perú: Peperomia inaequalifolia (congona), Alternanthera sp. (lancetilla) y Teline monspessulana (retama). Objetivos. Evaluar el desempeño de modelos de clasificación multiclase aplicados a espectros NIR-HSI preprocesados, con el fin de desarrollar un sistema confiable para la identificación vegetal y el control de calidad. Materiales y Métodos. Se recolectaron un total de 1467 espectros mediante una cámara NIR-HSI en el rango de 1300–1700 nm. Los datos espectrales fueron preprocesados utilizando suavizado Savitzky–Golay y normalización por variación normal estándar (SNV). Se entrenaron y evaluaron siete clasificadores de aprendizaje automático mediante validación cruzada estratificada de 5 pliegues, incluyendo Random Forest, Gradient Boosting, XGBoost y Ridge Classifier. Resultados. Random Forest alcanzó el mejor desempeño (accuracy = 0.9959), con un ROC-AUC de 1.00. Los demás modelos presentaron valores de exactitud entre 0.9720 y 0.9945, con valores de ROC-AUC cercanos a 1.00, lo que indica una alta capacidad discriminativa. Conclusiones. La combinación de NIR-HSI y modelos supervisados de aprendizaje automático permite una clasificación altamente precisa de plantas medicinales. Este enfoque muestra potencial para aplicaciones en trazabilidad, aseguramiento de la calidad y validación etnobotánica, particularmente en contextos descentralizados o con recursos limitados.
Referencias
1. Valarezo E, Herrera-García M, Astudillo-Dávila P, Rosales-Demera I, Jaramillo-Fierro X, Cartuche L, et al. Study of the chemical composition and biological activity of the essential oil from congona (Peperomia inaequalifolia Ruiz and Pav.). Plants (Basel). 2023;12:1504. Available from: https://doi.org/10.3390/plants12071504
2. Ahmed A, Ghatas Y, Mohamed SM. Evaluation the growth, some chemical constituents and landscape value of Alternanthera dentata plant grown under different planting methods and distances with herbaceous plants. Ann Agric Sci Moshtohor. 2024;62:57-66. Available from: https://doi.org/10.21608/assjm.2024.285941.1279
3. Espinoza-Gavilanes R, Tuza-Roa I, Vásquez-Freytez C, Jaramillo-Loayza K, Noriega-Rivera P. Efecto acaricida y ovicida de los aceites esenciales de Chenopodium ambrosioides L. y Peperomia inaequalifolia Ruiz & Pav. contra Tetranychus urticae en fresa (Fragaria spp.). Polibotanica. 2024;(57). Available from: https://doi.org/10.18387/polibotanica.57.14
4. Islam M, Bijjahalli S, Fahey T, Gardi A, Sabatini R, Lamb DW. Destructive and non-destructive measurement approaches and the application of AI models in precision agriculture: a review. Precis Agric. 2024;25:1127-80. Available from: https://doi.org/10.1007/s11119-024-10112-5
5. Atanassova S, Petrova A, Yorgov D, Mineva R, Veleva P. Visible and near-infrared spectroscopy for investigation of water and nitrogen stress in tomato plants. AgriEngineering. 2025;7:155. Available from: https://doi.org/10.3390/agriengineering7050155
6. Zhao B, Zhang H, Liu X, Dong Q, Zang H. Study of glycated human serum albumin in non-enzymatic glycation process based on MIR/NIR spectroscopy. J Mol Struct. 2025;1335:141928. Available from: https://doi.org/10.1016/j.molstruc.2025.141928
7. Benetti RA, Arantes PR, Oliveira PBR, Santos TB, Filho LJR, Lima UC, et al. Near infrared spectroscopy for the photodiagnosis of osteonecrosis, a future perspective: mini historical review. J Near Infrared Spectrosc. 2025;33:3-9. Available from: https://doi.org/10.1177/09670335251329601
8. Santos YJS, Malegori C, Colnago LA, Vanin FM. Application of infrared spectroscopy for the analysis of total phenolic compounds in fruits. Crit Rev Food Sci Nutr. 2024;64:2906-16. Available from: https://doi.org/10.1080/10408398.2022.2128036
9. Barbinta-Patrascu ME, Bita B, Negut I. From nature to technology: exploring the potential of plant-based materials and modified plants in biomimetics, bionics, and green innovations. Biomimetics (Basel). 2024;9:390. Available from: https://doi.org/10.3390/biomimetics9070390
10. Beć KB, Grabska J, Huck CW. Interpretability in near-infrared (NIR) spectroscopy: current pathways to the long-standing challenge. Trends Anal Chem. 2025;189:118254. Available from: https://doi.org/10.1016/j.trac.2025.118254
11. Wan-Azemin A, Suryati Mohd K, Rao USM, Sasidharan S, Dharmaraj S. Chemometric analysis of attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectra for geographical authentication of Melastoma malabathricum. Res J Pharm Technol. 2024;17:3769-76. Available from: https://doi.org/10.52711/0974-360X.2024.00586
12. Chang B, Li F, Hu Y, Yin H, Feng Z, Zhao L. Application of UAV remote sensing for vegetation identification: a review and meta-analysis. Front Plant Sci. 2025;16:1452053. Available from: https://doi.org/10.3389/fpls.2025.1452053
13. Cioanca O, Lungu II, Mita-Baciu I, Robu S, Burlec AF, Hancianu M, et al. Extraction and purification of catechins from tea leaves: an overview of methods, advantages, and disadvantages. Separations. 2024;11:171. Available from: https://doi.org/10.3390/separations11060171
14. Yu Y, Huang J, Wang L, Liang S. A 1D-Inception-ResNet based global detection model for thin-skinned multifruit spectral quantitative analysis. Food Control. 2025;167:110823. Available from: https://doi.org/10.1016/j.foodcont.2024.110823
15. Singla RK, Dhir V, Madaan R, Kumar D, Singh Bola S, Bansal M, et al. The genus Alternanthera: phytochemical and ethnopharmacological perspectives. Front Pharmacol. 2022;13:769111. Available from: https://doi.org/10.3389/fphar.2022.769111
16. Laranjeira IM, Dias ACP, Pinto-Ribeiro FL. Genista tridentata phytochemical characterization and biological activities: a systematic review. Biology (Basel). 2023;12:1387. Available from: https://doi.org/10.3390/biology12111387
17. Sing D, Banerjee S, Jana SN, Mallik R, Dastidar SG, Majumdar K, et al. Estimation of andrographolides and gradation of Andrographis paniculata leaves using near infrared spectroscopy together with support vector machine. Front Pharmacol. 2021;12:629833. Available from: https://doi.org/10.3389/fphar.2021.629833
18. Kasemsumran S, Apiwatanapiwat W, Ngowsuwan K, Jungtheerapanich S. Rapid selection of Andrographis paniculata medicinal plant materials based on major bioactive using near-infrared spectroscopy. Chem Pap. 2021;75:5633-44. Available from: https://doi.org/10.1007/s11696-021-01746-0
19. Zhang J, Wu X, He C, Wu B, Zhang S, Sun J. Near-infrared spectroscopy combined with fuzzy improved direct linear discriminant analysis for nondestructive discrimination of chrysanthemum tea varieties. Foods. 2024;13:1439. Available from: https://doi.org/10.3390/foods13101439
20. Li G, Li J, Liu H, Wang Y. Rapid and accurate identification of Gastrodia elata Blume species based on FTIR and NIR spectroscopy combined with chemometric methods. Talanta. 2025;281:126910. Available from: https://doi.org/10.1016/j.talanta.2024.126910
21. Jayapal PK, Joshi R, Sathasivam R, Van Nguyen B, Faqeerzada MA, Park SU, et al. Non-destructive measurement of total phenolic compounds in Arabidopsis under various stress conditions. Front Plant Sci. 2022;13:982247. Available from: https://doi.org/10.3389/fpls.2022.982247
22. Chen Z, Xue X, Wu H, Gao H, Wang G, Ni G, et al. Visible/near-infrared hyperspectral imaging combined with machine learning for identification of ten Dalbergia species. Front Plant Sci. 2024;15:1413215. Available from: https://doi.org/10.3389/fpls.2024.1413215
23. Liu X, Wu Z, Zhao Q, Yu Y, Li Z. Using near-infrared hyperspectral imaging combined with machine learning to predict the components and the origin of Radix Paeoniae Rubra. Anal Methods. 2025;17:1334-44. Available from: https://doi.org/10.1039/D4AY01977F
24. Ahmed MT, Monjur O, Kamruzzaman M. Deep learning-based hyperspectral image reconstruction for quality assessment of agro-product. J Food Eng. 2024;382:112223. Available from: https://doi.org/10.1016/j.jfoodeng.2024.112223
25. Du Z, You S, Cheng C, Wei S. Automatic spectral calibration of hyperspectral images: method, dataset and benchmark. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2025. p. 28081-90. Available from: https://doi.org/10.1109/CVPR52734.2025.02615
26. Rodríguez-León A, Oblitas J, Quevedo-Olaya JL, Vera W, Quispe-Santivañez GW, Salvador-Reyes R. Non-destructive detection of Elasmopalpus lignosellus infestation in fresh asparagus using VIS-NIR hyperspectral imaging and machine learning. Foods. 2026;15:355. Available from: https://doi.org/10.3390/foods15020355
27. Dalal N, Sáiz MJ, Caporale AG, Baldini F, Babayan SA, Adamo P. Fishy forensics: FT-NIR and machine learning based authentication of Mediterranean anchovies (Engraulis encrasicolus). J Food Compos Anal. 2024;136:106847. Available from: https://doi.org/10.1016/j.jfca.2024.106847
28. Krishnamoorthi S, Urano D. Hyperspectral reflectance imaging and spectral component analysis techniques to reveal distinct color patterns on plant leaves. STAR Protoc. 2025;6:103854. Available from: https://doi.org/10.1016/j.xpro.2025.103854
29. Sutliff BP, Beaucage PA, Audus DJ, Orski SV, Martin TB. Sorting polyolefins with near-infrared spectroscopy: identification of optimal data analysis pipelines and machine learning classifiers. Digit Discov. 2024;3:2341-55. Available from: https://doi.org/10.1039/D4DD00235K
30. Guo K, Shen Y, Gonzalez-Montiel GA, Huang Y, Zhou Y, Surve M, et al. Artificial intelligence in spectroscopy: advancing chemistry from prediction to generation and beyond. In: Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. 2025. p. 10445-54. Available from: https://doi.org/10.24963/ijcai.2025/1160
31. Beć KB. A simple guide to complex world of overtone and combination bands: theoretical simulation and interpretation of NIR spectra—summary of the workshop at NIR-2021 Beijing Conference. NIR News. 2021;32:15-8. Available from: https://doi.org/10.1177/09603360211060966
32. Türker-Kaya S, Huck C. A review of mid-infrared and near-infrared imaging: principles, concepts and applications in plant tissue analysis. Molecules. 2017;22:168. Available from: https://doi.org/10.3390/molecules22010168
33. Beć KB, Grabska J, Huck CW. Principles and applications of miniaturized near-infrared (NIR) spectrometers. Chemistry. 2021;27:1514-32. Available from: https://doi.org/10.1002/chem.202002838
34. Birenboim M, Brikenstein N, Kenigsbuch D, Shimshoni JA. Aquaphotomics study of fresh cannabis inflorescence: near infrared spectral analysis of water matrix structures. Anal Bioanal Chem. 2025;417:747-60. Available from: https://doi.org/10.1007/s00216-024-05685-z
35. Ma L, Peng Y, Pei Y, Zeng J, Shen H, Cao J, et al. Systematic discovery about NIR spectral assignment from chemical structural property to natural chemical compounds. Sci Rep. 2019;9:9503. Available from: https://doi.org/10.1038/s41598-019-45945-y
36. Shah N, Meng Q, Zou Z, Zhang X. Systematic analysis on the horse-shoe-like effect in PCA plots of scRNA-seq data. Bioinform Adv. 2024;4:vbae109. Available from: https://doi.org/10.1093/bioadv/vbae109
37. Cao L, Sun M, Yang Z, Jiang D, Yin D, Duan Y. A novel Transformer-CNN approach for predicting soil properties from LUCAS Vis-NIR spectral data. Agronomy. 2024;14:1998. Available from: https://doi.org/10.3390/agronomy14091998
38. Peng X, Yu X, Lu L, Ye X, Zhong L, Hu W, et al. Application of handheld near infrared spectrometer in quality control of traditional Chinese medicine: rapid screening and quantitative analysis of Lonicerae Japonicae Flos adulteration. Spectrochim Acta A Mol Biomol Spectrosc. 2025;326:125215. Available from: https://doi.org/10.1016/j.saa.2024.125215
39. Chen R, Liu F, Zhang C, Wang W, Yang R, Zhao Y, et al. Trends in digital detection for the quality and safety of herbs using infrared and Raman spectroscopy. Front Plant Sci. 2023;14:1128300. Available from: https://doi.org/10.3389/fpls.2023.1128300
40. Hajaj S, El Harti A, Pour AB, Khandouch Y, Üstüner M, Amiri MM. Balancing hyperspectral dimensionality reduction and information preservation for machine learning-based lithological classification using EnMAP hyperspectral imagery. Remote Sens Appl Soc Environ. 2025;38:101618. Available from: https://doi.org/10.1016/j.rsase.2025.101618
41. Yang Y, Wang S, Zhu Q, Qin Y, Zhai D, Lian F, et al. Non-destructive geographical traceability of American ginseng using near-infrared spectroscopy combined with a novel deep learning model. J Food Compos Anal. 2024;136:106736. Available from: https://doi.org/10.1016/j.jfca.2024.106736
42. Zheng C, Li J, Liu H, Wang Y. Rapid and non-invasive estimation of total phenol content and species identification in dried wild edible bolete using FT-NIR spectroscopy. Arab J Chem. 2024;17:106011. Available from: https://doi.org/10.1016/j.arabjc.2024.106011
43. Li G, Li J, Liu H, Wang Y. Rapid and accurate identification of Gastrodia elata Blume species based on FTIR and NIR spectroscopy combined with chemometric methods. Talanta. 2025;281:126910. Available from: https://doi.org/10.1016/j.talanta.2024.126910
44. Altieri G, Laveglia S, Rashvand M, Genovese F, Matera A, Mininni AN, et al. Portable NIR spectroscopy combined with machine learning for kiwi ripeness classification: an approach to precision farming. Appl Sci (Basel). 2025;15:6233. Available from: https://doi.org/10.3390/app15116233
45. Yuan M, Ding L, Bai R, Yang J, Zhan Z, Zhao Z, et al. Feature-level hyperspectral data fusion with CNN modeling for non-destructive authentication of “Weilian” from different origins. Microchem J. 2025;215:114201. Available from: https://doi.org/10.1016/j.microc.2025.114201
46. Chen Y, Li S, Jia J, Sun C, Cui E, Xu Y, et al. FT-NIR combined with machine learning was used to rapidly detect the adulteration of Pericarpium Citri Reticulatae (Chenpi) and predict the adulteration concentration. Food Chem X. 2024;24:101798. Available from: https://doi.org/10.1016/j.fochx.2024.101798
Descargas
Publicado
Cómo citar
Descargas
Número
Sección
Licencia
Derechos de autor 2026 Innovaciencia

Esta obra está bajo una licencia internacional Creative Commons Atribución-NoComercial 4.0.
Todos los artículos publicados en esta revista científica están protegidos por los derechos de autor. Los autores retienen los derechos de autor y conceden a la revista el derecho de primera publicación con el trabajo simultáneamente licenciado bajo una Licencia Creative Commons Atribución-NoComercial 4.0 Internacional (CC BY-NC 4.0) que permite compartir el trabajo con reconocimiento de autoría y sin fines comerciales.
Los lectores pueden copiar y distribuir el material de este número de la revista para fines no comerciales en cualquier medio, siempre que se cite el trabajo original y se den crédito a los autores y a la revista.
Cualquier uso comercial del material de esta revista está estrictamente prohibido sin el permiso por escrito del titular de los derechos de autor.
Para obtener más información sobre los derechos de autor de la revista y las políticas de acceso abierto, por favor visite nuestro sitio web.









