Identifying Optimal Wavelengths as Disease Signatures Using Hyperspectral Sensor and Machine Learning

Hyperspectral sensors combined with machine learning are increasingly utilized in agricultural crop systems for diverse applications, including plant disease detection. This study was designed to identify the most important wavelengths discriminate between healthy and diseased peanut (Arachis hypogaea L.) plants infected Athelia rolfsii, causal agent of stem rot, using in-situ spectroscopy learning. In greenhouse experiments, daily measurements were conducted inspect symptoms visually collect spectral reflectance leaves on lateral stems mock-inoculated inoculated A. rolfsii. Spectrum files categorized into five classes based foliar wilting symptoms. Five feature selection methods compared select top 10 ranked without a custom minimum distance 20 nm. Recursive elimination outperformed chi-square SelectFromModel methods. Adding nm selected improved classification performance. Wavelengths 501–505, 690–694, 763 884 repeatedly by two or more These can be applied designing optical automated rot detection fields. The machine-learning-based methodology adapted signatures other plant-pathogen systems.