SoilMate

Crop Disease Predictor
Fertilizer Predictor
Crop Predictor
Crop Disease Result
Crop Predictor Result
Fertilizer Predictor Result
Crop Disease Result
Home Page

Inspiration

Our project was inspired by the desire to assist farmers and agricultural scientists in finding solutions to various challenges in one centralized platform. We aimed to create a tool that could provide insights and recommendations for crop health, soil management, and disease prevention, all in one place. By leveraging machine learning models and web scraping, we sought to offer comprehensive solutions that would be easily accessible to farmers and researchers alike.

What We Learned

Throughout this project, we learned a great deal about model training, data processing, and how to integrate multiple systems into a functional web platform. We explored advanced techniques in image recognition, crop analysis, and soil condition prediction, which broadened our understanding of agricultural applications in technology. Additionally, we gained valuable insights into web scraping and how to retrieve real-time information to enhance the model's predictions.

How We Built It

We trained three different machine learning models, each designed to tackle a specific task:

Best Crop Prediction: The first model analyzes the city location, soil conditions, and mineral levels to predict the best crop to grow in that environment.
Disease Detection: The second model identifies plant diseases by analyzing uploaded images of diseased crops. It recognizes the plant species, diagnoses the disease, identifies its cause, and provides prevention steps.
Pesticide Recommendations: The third model offers pesticide suggestions based on the soil condition and the type of plant the user is attempting to grow.

To enhance the system, we implemented web scraping to gather additional data for better accuracy and relevancy in predictions.

Blog Section

We also have a blog section where users can find the latest news and updates related to agriculture and farming. This section provides all the most recent developments in one place, helping farmers and scientists stay informed about trends, technologies, and new research in the agricultural field.

Challenges We Faced

One of the main challenges was ensuring that the models provided accurate predictions across various crops and soil types. Training the disease detection model with sufficient accuracy while handling the complexity of real-world images was also a significant hurdle. Additionally, integrating web scraping for real-time data required careful planning to ensure the data was reliable and enhanced the overall functionality of the models.