Waste classification using YOLOv8 and Streamlit offers a powerful solution for improving waste management practices. By leveraging computer vision, deep learning algorithms, and an interactive web interface, we can automate the process of waste classification, enhancing recycling efficiency and resource recovery. The integration of YOLOv8 for accurate object detection, Streamlit for seamless deployment, and other complementary technologies like OpenCV creates a comprehensive waste classification system. This technology stack enables real-time waste classification, user-friendly interactions, and the potential for widespread adoption in various waste management scenarios. By adopting these technologies, we can make significant strides towards sustainable waste management and contribute to a cleaner and greener future
A diverse dataset of waste images including different waste categories such as plastic, metal, paper, glass, cardboard, biodegradable is collected from Roboflow. The dataset has over 6000 images and splitted into 3 sets: Training, Validation, Testing. The dataset is annotated with bounding box labels around the waste objects using in-built annotation features present in Roboflow
You can get the dataset from https://roboflow.com/
Here is how our streamlit webapp is used to detect and classify waste into different categories. The User Interface allows the users to upload waste images, videos and open their camera to detect and classify the waste in real-time.
