Exploring Spectral Patterns of Tomato Phytophthora Infections using the VMRNMAF Deep Learning Framework

Abstract

Identifying and mitigating diseases affecting tomato crops pose significant challenges for researchers and agricultural producers alike. Effective disease detection is crucial for optimizing yields and implementing targeted treatments. This study presents a comprehensive framework, Vector Machine Recurrent Neural with Manual Activation Function (VMRNMAF), for early prediction of Phytophthora infestans and other common tomato diseases. Leveraging image preprocessing techniques and feature identification processes, VMRNMAF accurately distinguishes between healthy and infected tomato segments. By integrating Recurrent Neural Networks (RNNs) and manual activation functions in image segmentation, this framework enhances disease identification accuracy. Throughout deep learning methodologies, VMRNMAF streamlines the identification of bacterial, viral infections, offering a promising solution for improving tomato crop management practices. Experimental results demonstrate the efficacy of VMRNMAF in accurately identifying diseased areas within tomato crops, underscoring its potential for enhancing agricultural sustainability and productivity.