Research Article
Edge Intelligence Based Garbage Classification Detection Method
@INPROCEEDINGS{10.1007/978-3-031-28990-3_10, author={Ruijia Zhu and Yiwen Liu and Yanxia Gao and Yuanquan Shi and Xiaoning Peng}, title={Edge Intelligence Based Garbage Classification Detection Method}, proceedings={Edge Computing and IoT: Systems, Management and Security. Third EAI International Conference, ICECI 2022, Virtual Event, December 13-14, 2022, Proceedings}, proceedings_a={ICECI}, year={2023}, month={3}, keywords={Edge Intelligence Waste Classification Convolutional Neural Network Smart City}, doi={10.1007/978-3-031-28990-3_10} }- Ruijia Zhu
Yiwen Liu
Yanxia Gao
Yuanquan Shi
Xiaoning Peng
Year: 2023
Edge Intelligence Based Garbage Classification Detection Method
ICECI
Springer
DOI: 10.1007/978-3-031-28990-3_10
Abstract
To address the problem that the classification and cleaning of garbage in city streets is always ineffective nowadays, the paper proposes a garbage detection method based on edge intelligence. The edge intelligence not only reduces the computational load of the cloud and speeds up the data transmission, but also greatly reduces the data transmission cost. First, images of city streets are collected and uploaded to the edge device via mobile devices in various locations in the city. Then, the edge server is used to temporarily store the image information, and the PeleeNet model deployed on it is used to identify and classify various kinds of garbage, and then visualize the information of each street. Finally, the street garbage information is transmitted to the cloud, which provides a detailed picture of the city’s garbage situation and facilitates city management.
In this paper, the PeleeNet model is compared with ResNet, DenseNet and MobileNet models. The results show that the edge devices equipped with PeleeNet model not only have the fastest computation speed and the highest accuracy, but also occupy the least memory. It is fully demonstrated that the method studied in the paper can be applied to the problem of litter detection in urban streets.
