AgroHub

Landing Page with Typewriter Effect and API integrated as Carousel
Footer and Blogs Page Redirect
Redirected E-Commerce Website
Farmer Uploads a photo of Soil here...
Deployed Models using FastAPI
Working CNN model predicting the soil type.

AGROHUB

Inspiration

Our project is motivated by the imperative to empower farmers with sophisticated, data-driven insights for optimal crop cultivation. By seamlessly integrating price prediction models, soil image classification, and environmental analysis, we aspire to revolutionize traditional farming methods. Our comprehensive approach considers historical minimum support prices, environmental conditions, and soil quality, providing farmers with a holistic decision-making tool to enhance agricultural productivity and sustainability. Additionally, our solution extends to support farmers in accessing eCommerce services, bridging the gap between producers and consumers in the agricultural ecosystem.

What it does

Our project offers a holistic solution for agricultural optimization, leveraging advanced technologies to empower farmers. Through the integration of price prediction models, soil image classification, and environmental analysis, the platform provides farmers with data-driven insights for informed decision-making in crop cultivation. By considering historical minimum support prices, environmental conditions, and soil quality, our system guides farmers in selecting the most suitable crops, optimizing yields, and enabling sustainable agricultural practices. The project also facilitates access to eCommerce services, connecting producers and consumers within the agricultural ecosystem, thereby enhancing overall efficiency and transparency in the agricultural supply chain.

How we built it

We meticulously crafted our comprehensive agricultural solution by incorporating state-of-the-art machine learning models. The Convolutional Neural Network (CNN) employed for soil image classification demonstrated an impressive accuracy of 97%, ensuring precise categorization. For crop classification based on environmental conditions, the RandomForestClassifier achieved a commendable accuracy of 92%. In predicting Minimum Support Prices (MSP) using Linear Regression, the model exhibited an average loss of 29. These robust models were seamlessly transformed into APIs using FastAPI and seamlessly integrated into our React.js-based web application. Additionally, our platform harnessed the power of external APIs for weather forecasting and official government data on MSP prices, enhancing the application's capabilities and providing farmers with accurate, data-driven insights for informed decision-making in agriculture.

Challenges we ran into

During the development of our agricultural solution, we faced challenges in obtaining quality datasets for training our machine learning models, especially for soil image classification, crop classification, and MSP prediction. Designing user-friendly webpages that balance functionality and simplicity presented another hurdle. Identifying official and freely accessible APIs for critical data, like weather forecasting and government MSP prices, proved challenging. The integration of backend and frontend components required careful coordination to ensure seamless communication between FastAPI-based machine learning models and the React.js-based web application. Despite these challenges,we tried our best to implement a User-Friendly Solution for our Farmers.

Accomplishments that we're proud of

We are proud to highlight several key accomplishments in the development of our agricultural solution. The successful training and deployment of our machine learning models, including the CNN for soil image classification, RandomForestClassifier for crop classification, and Linear Regression for MSP prediction, showcase the effectiveness of our predictive capabilities. The seamless integration of these models onto a local server and their successful deployment attest to the reliability of our backend infrastructure. Furthermore, our achievement in designing intuitive and informative webpages using React.js reflects our commitment to delivering a user-friendly experience. These accomplishments collectively demonstrate the successful convergence of cutting-edge technology and user-centric design in our end-to-end agricultural solution.

What we learned

Throughout our project, we acquired valuable insights into the deployment of machine learning models on local servers, emphasizing scalability and efficiency. Utilizing data_files for model incorporation into FastAPI highlighted the importance of streamlined deployment processes. In the agricultural realm, we deepened our understanding of farmers' challenges and the transformative potential of technology. Integrating external APIs, such as weather forecasting and government MSP data, underscored the significance of diverse and reliable data sources. Designing user-friendly webpages reinforced the importance of effective communication and user experience in conveying complex information. These experiences collectively enriched our project, offering a nuanced understanding of both technological and agricultural domains.

What's next for PROJECT

Looking ahead, we're planning to make our project even more useful and user-friendly. We want to provide real-time data access through APIs, keeping information up-to-date for farmers. Using Cronjob for continuous improvements will make our machine learning models stronger and more reliable. We're also working on turning our web app into a mobile app for easier access. To make our platform even better, we're listening to local farmers and adding more features that cater to their specific needs and challenges. Our goal is to keep evolving and staying ahead in the world of agricultural technology.