Development of Computer Vision Algorithms for Multi-class Waste Segregation and Their Analysis

Classification of waste for recycling has been a focal point scientists interested in the field conservation environment. Recycling consists numerous steps, which one most crucial is segregation recyclables from all other waste. Due to lack safety standards developing countries, collection often done manually by domestic helpers, or "rag-pickers". Such process risks individual and public health. The methods may ultimately cause become non-recyclable due cross-contamination. Literature shows that research this direction focuses on single class detection. proposed work investigates CNN, YOLO, faster RCNN-based multi-class classification detect different types at collecting point. smart dustbin employs these computer vision with Raspberry Pi microcontroller camera module. experimental results show CNN 80% accuracy 60% loss. Whereas YOLO algorithm an 88% loss 40%. But best were obtained RCNN object detection API, 91% 16%. There already existing method making dustbin, so are compared how can be used make dustbin. This Doi: 10.28991/ESJ-2022-06-03-015 Full Text: PDF