Time Littering Detection—Using Pose Estimation and Object Detection
R. Kaladevi, K. Ganesh, P. Kishore, Raghavan V. Vijaya · 2024
Public littering is a huge problem in Developing nations currently. Even though the world governments have implemented various schemes and awareness programs to clean our surroundings, there are not many ways to prevent people from littering. We see every other street littered with garbage and there is no proper system in place to monitor and hold people accountable for littering. Hence this research aims to solve this problem by developing a real time Litter detection system which uses YOLO, Faster-RCNN and Human Action Recognition methods to identify if a person litters in real time using live footage from IP cameras and captures the face of the litterer if visible and flags them so we can identify if they repeat their actions in the future. This data can be used by the authorities to hold the people who litter accountable for their actions, and it can also help us identify which places require a cleanup.