AI with Clarity: Unraveling Plant Disease through XAI

The health of crops is a very important factor in the entire process of agriculture. The crop health determines the better yield and better productivity. Agricultural crop health monitoring and solution suggestion system monitors the crop distributed over a larger area. The proposed system is an integrated system which consists of a dashboard for farmers, Agricultural experts and pesticide sellers The system enables the farmer to check the status of health of the plants by detection of presence of disease in crops by using deep learning techniques from images of the plants. It also verifies using the Explainable Artificial intelligence(XAI) technique. Deep learning method, Convolutional neural network architecture is used to detect the diseases present in the crops. The comparison of five CNN architectures is made. Among five architectures ResNet50 performed well with accuracy of about 92.9% The XAI method is used to provide explanations that enable humans to trust the results produced by the developed model. Then the pesticides that are necessary to combat the disease are suggested by the system. This System uses Case-Based Reasoning method to recommend the appropriate pesticides. This system also enables the farmer to seek solutions for the unique set of diseases from agricultural experts. The details including the quantity, price of the pesticide is provided by the seller. All these factors can efficiently monitor crops by immediately responding to treat plant disease. This system helps in maintaining the health of crops throughout the entire process of farming.