A Framework for Crop Disease Detection Using Feature Fusion Method

Crop disease detection methods vary from traditional machine learning, which uses Hand-Crafted Features (HCF) to the current deep learning techniques that utilize features. In this study, a hybrid framework is designed for crop using feature fusion. Convolutional Neural Network (CNN) used high level features are fused with HCF. Cepstral coefficients of RGB images presented as one along other popular The proposed model tested on whole leaf and also image patches have individual lesions. experimental results give an enhanced performance classification accuracy 99.93% 99.74% shows significant improvement in comparison state-of-art techniques. improved show prominence fusion establish cepstral pertinent detection.