CleanLoop is a project focused on detecting trash in images and estimating the cleaning time required for the identified areas. This feature leverages YOLOv8, along with Cloudinary and Imagga APIs, to process, recognize, and tag trash in sample images, offering a visual representation of detected areas alongside an estimation of cleaning time.
This project was developed at the SFSCON Hackathon (Nov 2024), Bolzano, Italy 🇮🇹. Learn more about the event here.
- Trash Detection: Identifies and tags trash in images using the YOLOv8 model, optimized for real-time object detection.
- Cleaning Time Estimation: Calculates an estimated cleaning time based on detected trash areas.
- Cloud Storage and Management: Utilizes Cloudinary for image hosting and management.
Original Image
Tagged Trashes
The project uses an algorithm designed to estimate cleaning time by analyzing the tagged trash areas within the image.
- Model: YOLOv8 - An object detection model, optimized for identifying trash in real-time.
- Cloudinary: For image storage, transformation, and optimization.
- Imagga: For image tagging and analysis to enhance trash detection.
To run this feature locally:
- Clone the repository:
git clone https://github.com/pouyasattari/trash-detection-cleanloop-project.git cd trash-detection-cleanloop-project/trash-detection