Development and Design of Intelligent Garbage Classification System Based on Image Recognition
Download as PDF
AI is the most popular technology science nowadays, and the branch technology-image recognition technology is an important research field. How can machines recognize and distinguish as accurately as humans is a hot research topic at present. In this paper, an intelligent garbage classification system based on image recognition is designed. Raspberry Pi 3B is used as the main control chip of the system, and CCD camera is used as the camera. The model is based on YOLOv4' s garbage detection in complex environment, and an improved ResNet50 network is constructed to train and identify the types of garbage. The experimental results show that the harmful waste classification accuracy of image recognition can reach 100%, and the recovery garbage recognition accuracy can reach 94%. Practice has proved that this identification method is feasible, which helps people to solve the problem of garbage classification, on the other hand, it can reduce the cost of garbage classification and improve the processing efficiency of garbage classification.
Garbage classification; Image recognition; YOLOv4