Artificial Intelligence Based Plant Disease Detection

Abstract

This paper analyses the advanced Neural Network (NN) techniques for hyperspectral data processing, with a particular focus on detecting plant disease. To initiate, we will discuss the NN mechanism, types, models, and classifiers that are used to process the hyperspectral data using different algorithms. Following that, this project discusses the current state of imaging and non-imaging hyperspectral data for early screening and diagnosis of Disease. The NN-hyperspectral hybridization method has been shown to be a great resource for disease detection and diagnosis. The SDI is defined as the proportion of different spectral bands in pure disease spectra. Following that, we'll go over NN techniques for rapid SDI developments. In addition, we discuss current hyperspectral data challenges and future trends. The trained model used by us achieves an accuracy on a held-out test set, showing the approach's practicability. Overall, training deep learning models on increasingly large and publicly available image datasets indicates a clear path toward widespread global smartphone-assisted crop disease detection.