Nowadays, many farmers face critical challenges in traditional agriculture due to a lack of precise knowledge regarding crop suitability and soil health. Driven by guesswork, the unregulated application of fertilizers often leads to soil degradation, reduced crop productivity, and financial losses. To mitigate this issue, this paper proposes AgriScan Intelligence, a secure and comprehensive cloud-based web application designed to support data-driven agricultural decisions. The system accepts crucial environmental and soil parameters as input, including Nitrogen (N), Phosphorus (P), Potassium (K), soil pH, temperature, and rainfall. Advanced Machine Learning models, specifically Random Forest and Logistic Regression, process this real-time data to accurately predict the most suitable crop and recommend the precise fertilizer mixture required for optimal soil health. The system's backend is engineered using Python with the Flask framework, and user data is securely managed via an Aiven Cloud MySQL database, achieving efficient query response times of under 200 ms. To enhance platform security, the Mailtrap API is integrated for secure OTP-based user authentication. Furthermore, the complete application features full regional language support (Marathi and Hindi) and has been successfully deployed on the Render cloud platform. Empirical results demonstrate that the Random Forest Classifier outperforms traditional models, delivering a predictive accuracy of 95-97%. Ultimately, AgriScan Intelligence provides an accessible, high-speed, and cost-effective smart farming solution, effectively bridging the gap between predictive machine learning analytics and small-scale farmers. Keywords: Precision Agriculture, Machine Learning, Flask, Cloud Database, Smart Farming, Random Forest.
Bodare et al. (Tue,) studied this question.