An Innovative Hybrid Feature Extraction Method for the Diagnosis of Coconut Leaf Diseases

In South India, cultivating coconuts is a major agricultural activity, but the trees are challenged by unfavorable weather conditions and environmental factors. These difficulties comprise pest infestations and a variety of leaf diseases. Because of the coconut trees' enormous foliage and shadowing, it might be challenging to identify and locate these problems. The study proposes a feature extraction strategy which integrates the feature vectors of deep learning (DL) and handcrafted features into a single, cohesive representation for detecting the coconut leaf diseases. The selected features were fed into the Machine learning (ML) classifier like Support Vector Machine (SVM), Random Forest (RF) and XGBoost. The experimental results show that the proposed technique outperforms with an accuracy of 97.4% showing the importance of feature extraction techniques in identifying the coconut leaf diseases.