RiceCare: AI-Powered Rice Crop Decision Support System

Architecture Diagram
Home Page
Yield Prediction Section
Yield Prediction Section in Tamil Language
Crop Health Analysis Section - Nutrient Analysis
Results of Nutrient Analysis
Disease Analysis

🚩 Inspiration

Agriculture plays a vital role in our country, yet many farmers still rely on traditional methods and guesswork for making critical decisions such as predicting yield, identifying crop diseases, and determining nutrient deficiencies.

During our research, we observed:

Difficulty in accurately predicting crop yield Challenges in early disease identification Nutrient deficiencies are hard to detect visually Overuse of fertilizers reduces soil productivity

This inspired us to build a system that simplifies farming decisions using Artificial Intelligence and makes it accessible to real-world users.

💡 What We Built

RiceCare AI is an AI-powered automated farming decision support system designed for rice cultivation.

The system integrates:

Machine Learning for yield prediction Deep Learning for disease and nutrient detection Automation workflows for real-time decision support

It provides:

🌾 Crop yield prediction 🍃 Disease detection from leaf images 🌱 Nutrient deficiency identification 💊 Smart recommendations with product usage guidance 🌍 Multi-language support (English & Tamil) ⚙️ How We Built It 🔹 Data Processing and Input

Users provide:

District Season Area

The system automatically fetches real-time weather data using an external API.

🔹 Yield Prediction Model

We used an XGBoost model trained on environmental and agricultural features such as:

Temperature Rainfall Humidity Solar radiation Soil nutrients (N, P, K, pH)

The model predicts yield as:

Yield= f (Weather, Soil, Area, Season, Location)

🔹 Image-Based Crop Analysis

We implemented Convolutional Neural Networks (CNNs) to analyze rice leaf images.

The system performs:

Disease classification Nutrient deficiency detection 🔹 Automation Workflow

RiceCare AI automates the entire pipeline:

Input collection Real-time data fetching Feature processing Model prediction Recommendation generation

This reduces manual effort and enables faster decision-making.

🔹 Recommendation Engine

The system provides:

Cause of issue Suggested actions Preventive measures Product recommendations with price Proper usage instructions to avoid overuse

🧠 What We Learned

Through this project, we gained experience in:

Machine Learning and Deep Learning API integration Feature engineering Full-stack AI system development

We also understood the importance of building solutions that are practical and accessible to real users.

⚠️ Challenges We Faced Feature mismatch between training and prediction data Deployment issues such as port configuration errors and worker timeouts Integrating multiple AI models into a single system Designing a simple and farmer-friendly user interface

🌍 Impact

RiceCare AI helps to:

Improve crop yield productivity 🌾 Detect crop issues early 🍃 Reduce fertilizer and pesticide misuse Support sustainable agriculture Enable data-driven decision-making

🔮 Future Scope

We plan to:

Extend support to multiple crops Add more regional languages across India Improve model accuracy using larger datasets Build a complete smart farming ecosystem

🚀 Conclusion

RiceCare AI demonstrates how Artificial Intelligence and Automation can solve real-world agricultural challenges.

It combines prediction, analysis, and intelligent recommendations to help farmers make better decisions and improve productivity.

Currently, the system is developed specifically for rice cultivation. In the future, it can be extended to support multiple crops and multiple languages across India, increasing its scalability and real-world impact.