An Integrated Hyperspectral Imaging Approach for Early Detection of Black Sigatoka Disease in Musa acuminata Varieties

Advances in Agriculture and Agricultural Sciences | Vol. 12, No. 2, February 2026 | pp. 33–38

DOI: 10.46882/2026.AAAS.120207

Research Article

Title: An Integrated Hyperspectral Imaging Approach for Early Detection of Black Sigatoka Disease in Musa acuminata Varieties

Names of Authors: Lopez, A. R.¹, Gomez, F. E.¹, and Martinez, J. L.²

Authors’ Affiliations: ¹Centro de Investigacion Cientifica de Yucatan, Merida, Mexico. ²Department of Plant Pathology, University of Florida, Homestead, USA.

Abstract: Black Sigatoka caused by Mycosphaerella fijiensis remains a highly destructive foliar disease affecting banana cultivars (Musa acuminata). Visible symptoms typically emerge after significant internal cellular degradation has occurred, limiting the performance of protective fungicides. This research established a non-destructive early diagnostic model using hyperspectral imaging across the 400–1000 nm spectral spectrum. Reflectance indices were extracted from healthy, latently infected, and symptom-bearing leaves from susceptible and resistant banana cultivars. Machine learning models using Support Vector Machine (SVM) algorithms classified asymptomatic leaf zones with an overall accuracy of 92.5%. The primary spectral variations occurred within the green peak (550 nm) and red-edge (710–740 nm) zones, matching early reductions in chlorophyll content and structural cell wall modifications. Photochemical Reflectance Index (PRI) values dropped significantly 6 days before visual necrotic spots appeared. This remote sensing technique allows targeted fungicide applications, reducing environmental chemical exposure and input costs.

Keywords: Remote sensing; Spectral signature; Pathogen detection; Banana cultivation; Machine learning; Precision crop protection.

Manuscript Timeline: Received November 02, 2025; Revised December 27, 2025; Accepted January 24, 2026; Published February 26, 2026.

Citation: Lopez, A. R., Gomez, F. E., and Martinez, J. L. (2026). An Integrated Hyperspectral Imaging Approach for Early Detection of Black Sigatoka Disease in Musa acuminata Varieties. Advances in Agriculture and Agricultural Sciences, 12(2), pp. 33–38.

Subscribe to read the full article: https://mail.internationalscholarsjournals.org/subscribe-to-read