Smart Garbage Detection System for Sustainable Waste Management using Deep Learning Techniques

Waste management is a significant challenge in India, where the country generates an estimated 62 million tons of waste annually. One of the most visible forms of waste is garbage littered on the roadside, which can have negative environmental, social, and health impacts. Despite the efforts of government authorities to manage waste, many cities in India still lack effective waste management practices. The lack of effective garbage detection and management systems is a major contributor. This paper proposes a smart garbage detection system to improve waste management practices in India. We will use the latest YOLOv8 algorithm to build an object detection model that will identify the garbage littered in the open spaces that have been converted into dumping grounds. The model has been fine-tuned on the TrashNet dataset, an open image dataset comprising over 2500 images of garbage objects. The proposed system has the potential to contribute to sustainable waste management practices in India by detecting the dumping grounds using machine learning models and informing the concerned authorities of the district. This could help improve recycling efforts and reduce the amount of waste in landfills.