Seasonal

• 

• 

• 

  See recipes specific to the ripeness of the fruit

💡 Inspiration

Discussing a variety of topics food scarcity and insecurity was a big topic of discussion especially during the winter months, not only is it hard to produce domestic fruits, but it also becomes more difficult to import them. Transportation times increase due to poor weather conditions and as a result, there is less supply of fruits which decreases fruit affordability and accessability. There our idea for an app was born! The app scans the fruit and provides the user with the ripeness of the fruit allowing users to purchase fruits that are more nutrient-dense and reduce food waste by providing recipes when the fruit is overripe.

🍎 What it does

Our trained AI model provides the top three fruits which accurately depict the image from the camera and provide the user with the number one match. The model will analyze the ripeness of the fruit and depict the most closely matching “ripeness” state.

⚙️ How we built it

Our front end was developed on Figma as a prototype and then implemented via React. The AI model was taught through inputting hundreds of images of various fruits and their corresponding “ripeness” state.

🥊 Challenges we ran into

One of our main challenges was implementing the OpenAI API key to prompt chatGPT for feedback on the fruits. As well as implementing MongoDB data that would provide the seasonality of the fruit and connect this to our backend.

🏆 Accomplishments that we're proud of

Leveraging advanced technology and curated data to determine the ripeness of the fruit. Storing Mongo database with information related to the national value, seasonality and recipe or the fruits. Training the model to accurately differentiate between models.

📚 What we learned

How to train a computer vision, using image-based machine learning then extracting that model and implementing it in our react app.

→ What's next for Seasonal

We aim to further refine the Seasonal ripeness detection system by providing the machine learning algorithm and image recognition technology with more images and data We plan to expand Seasonal’s fruit database to include a wide variety of fruits and vegetables from different regions worldwide.

Built With

• javascript
• mongodb
• openai
• opencv
• react
• tailwind