This project is a deep learning-based image classification app that predicts whether a fruit is fresh or rotten based on a user-uploaded image. It’s built using TensorFlow/Keras for model training and Streamlit for deployment.
Food quality assessment is a crucial task in the food industry. This project offers a smart and lightweight tool to classify fruit freshness from images, helping reduce food waste and automate sorting processes.
Manual sorting of fruits based on freshness is time-consuming and inconsistent. There’s a growing need for a fast, automated solution that ensures fruit quality is monitored efficiently without human error.
- Developed a CNN model trained on fruit images (Fresh vs Rotten)
- Deployed a web app using Streamlit
- Enables real-time image upload and freshness prediction
- Upload any image of fruit and get instant prediction
- Clean, user-friendly web interface
- Efficient and fast classification using a lightweight model
Extend to more fruit types (bananas, apples, etc.)
Multi-class classification (Fresh, Slightly Spoiled, Rotten)
Add real-time camera input for live predictions
Mobile-friendly UI
This project combines deep learning, image processing, and real-world impact. It demonstrates how AI can be applied in agriculture and retail to ensure quality and efficiency.
- Python
- TensorFlow / Keras
- OpenCV / PIL
- Streamlit (Frontend + Deployment)
git clone https://github.com/your-username/fruit-freshness-classification-app.git
cd fruit-freshness-classification-app
pip install -r requirements.txt
streamlit run app.py