Year: 2025
Type: article
Abstract: : — Assessing soil quality is vital for sustainable farming, efficient crop planning, and resource management. Traditionally, soil fertility analysis relies on lab-based chemical testing, which is costly, slow, and often inaccessible for many small and rural farmers. Recent advances in technology, including high-resolution satellite images from platforms like Sentinel-2 and ISRO BHUVAN, along with drone imaging, have revolutionized soil monitoring. The full potential of these rich image datasets is realized through AI, particularly CNN, which automatically detect complex patterns related to soil fertility without requiring manual feature extraction. This project integrates CNNs into an easy-to-use web application that classifies soil fertility levels—high, medium, and low— using satellite and drone images. Users such as farmers and agricultural advisors can quickly upload or capture field images to obtain fast, accurate soil fertility assessments along with practical recommendations. Advanced preprocessing standardizes images from diverse sources to maintain model accuracy across varying conditions, while extensive training on annotated datasets ensures reliable and rapid performance, surpassing many traditional methods. Index Terms: Soil Fertility, Deep Learning, Convolutional Neural Networks, Remote Sensing, Precision Agriculture, Web Application
DOI: 10.55041/ijsrem52552
Authors: Vaishnavi Vaishnavi, Suma Suma
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDGs: Zero hunger, Life on land
Open Access status: bronze
OpenAlex ID: W4414090323