Empowering Precision Agriculture: A Novel ResNet50 based PDICNet for Automated Apple Leaf Disease Detection

Abstract

Inspection of plant leaves through the naked eye is difficult and does not guarantee accurate assessment which results in economic loss to the farmers. Biotic or abiotic stress develop lesions on plant leaves and reduce crop quality and yield. This paper presents a novel model that uses RestNet50-based Deep Learning Convolutional Neural Network (DLCNN) classifier. This approach combines the optimization power of ACO with the robustness of ResNet50 to enable effective feature extraction and selection. We verify the efficacy of the strategy by extensive testing on an indigenously built dataset demonstrating its superior performance over other approaches. Our work advances automated methods for apple leaf disease detection and provides a dependable and useful solution for precision agriculture in apple orchards.