A Study on Garbage Classification Detection Employing the Faster R-CNN Framework

Abstract

Since the comprehensive implementation of mandatory garbage classification in Shanghai in 2019, accompanied by the introduction of relevant management regulations, garbage classification has not only repeatedly captured public attention but has also emerged as a "new social norm." Nevertheless, classification accuracy rates remaining below 50% have accelerated the expansion of related markets and catalyzed technological innovation in this domain. In response to the dual challenges of continuously increasing waste volumes and persistently low disposal accuracy, this paper proposes the application of the Faster R-CNN algorithm for garbage classification detection. By obtaining garbage categories through object detection, the proposed approach provides technical support for subsequent waste disposal operations, thereby reducing the workload associated with environmental governance and enhancing the efficiency of garbage recycling processes.