• 

  Home Page

• 

  Plant Disease Detection

• 

  Crop Recommendation

🚀 Peek Rakshak

🌱 Inspiration

Agriculture is the backbone of our economy, yet farmers face challenges such as crop diseases, inefficient fertilizer use, and unpredictable yields. We wanted to leverage AI/ML to create an intelligent crop monitoring system that empowers farmers with actionable insights.

🛠 How We Built It

We designed Peek Rakshak as a machine learning-powered web platform that integrates multiple AI models for agricultural analysis. The core components include:

• Crop Recommendation System:

  • Trained an ML model using Random Forest on soil and environmental data.
    
  • Suggests the best crops to cultivate based on soil conditions.
    

• Fertilizer Recommendation Model:

  • Developed using decision trees and regression models.
    
  • Provides personalized fertilizer advice based on nutrient levels.
    

• Leaf Disease Detection:

  • Implemented a Resnet (CNN) model in TensorFlow/Keras with 98% accuracy.
    
  • Trained on a dataset of diseased and healthy leaf images.
    
  • Deployed using Flask/FastAPI for real-time predictions.
    

🔥 Challenges We Faced

• Model Optimization: Training deep learning models required computational efficiency, so we optimized architectures using quantization and model pruning.
  
• Integration Integrating the frontend with the ML models running in the backend flask server.
  

🎓 What We Learned

• How to fine-tune deep learning models for agriculture.
  
• Deploying AI models in real-world applications using Flask.
• The importance of user-friendly AI interfaces for non-technical users.
  

🌍 Impact

Peek Rakshak aims to reduce agricultural losses and improve crop yield efficiency by providing AI-driven recommendations to farmers. With real-time insights, we empower farmers to make data-driven decisions for sustainable farming.

Built With

• flask
• machine-learning
• react