Literature Review: Garbage Classification by Robotic Arm Based on Image Recognition

Abstract

To address the problems of low efficiency and high cost of traditional manual garbage classification and promote the intelligent transformation of garbage classification, robotic arm garbage classification technology based on image recognition has become a research hotspot, which is of great significance for improving sorting efficiency and facilitating resource recycling. Adopting the literature research method and case analysis method, this paper systematically combbs the core system, domestic and foreign research progress and application status of this technology, aiming to clarify the current technical bottlenecks and propose future development directions. The research finds that image recognition algorithms (such as YOLO series and Faster R-CNN) and robotic arm control technologies (such as PID control and visual servo control) have been partially applied in garbage sorting. Foreign countries take the lead in multi-sensor fusion and reinforcement learning, while domestic research focuses on the technical optimization adapted to local classification standards. However, there are still key challenges including insufficient recognition accuracy of small targets, high confusion rate of similar garbage categories, poor adaptability to grabbing irregular garbage, and low system collaboration efficiency. Future research should focus on breakthroughs in multi-modal fusion recognition, flexible robotic arm grasping and intelligent operation and maintenance, so as to promote the low-cost and large-scale application of this technology.