FIELDVISION: ENHANCING AGRICULTURE THROUGH CONVOLUTIONAL NEURAL NETWORKS

Abstract

Modern agricultural approaches now include precision farming as a crucial element, with smart agriculture devices providing essential data on crop disease, soil nutrient levels, and recommended fertilizer applications.This research presents a convolutional neural network (CNN) based smart agriculture system for nutrient analysis and crop suggestion based on soil photographs of five different types of soil: cinder, black, peat, yellow, and laterite. The suggested system consists of a nutrient analysis and crop recommendation model based on CNN, an image-capturing device, and a user interface. In order to accurately predict the best crop and fertilizer for a given soil type, the system uses an image-capturing device to take soil images. The CNN-based model is then used to process the images. This model is trained on a large dataset of soil images along with associated crop types, nutrient content, and recommended fertilizers. Farmers can interact with the system through the user interface to receive crop recommendations, nutrient analyses, and fertilizer recommendations based on the images of the soil that were collected.