Deep Learning based Mango Leaves Disease Detection for Smart Agriculture

Several diseases impact the mango crop, lowering yields and lowering fruit quality. Because of economic hardship and inadequate medical facilities, many nations still struggle with the diagnosis of these illnesses. Several strategies for the automated identification of mango illnesses have been developed by researchers in the recent decade. The suggested system for detecting mango leaf diseases uses improved deep learning methods to identify the root causes of the diseases and provide recommendations for their treatment. To detect diseases in images of mango leaves, the suggested study uses Tensor Flow. The proposed method employs the identification of items API tensor flow to independently diagnose illnesses; it is trained using a convolutional neural network (CNN). This research study has developed advanced transfer learning models using the VGG-19, ResNet101, and InceptionV1 topologies to detect mango diseases from pictures of healthy and diseased leaves. The proposed approach outperforms its predecessors with a 95.2% accuracy rate in illness leaf identification.