Research Article
Design and implementation of Garbage Detection in Water Area Based on Yolov5 Algorithm
@INPROCEEDINGS{10.4108/eai.15-12-2023.2345400, author={Jiajia Meng and Liuling Lang and Zhenzhan Lu and Xiling Tang and Huimin He}, title={Design and implementation of Garbage Detection in Water Area Based on Yolov5 Algorithm}, proceedings={Proceedings of the 3rd International Conference on Public Management and Big Data Analysis, PMBDA 2023, December 15--17, 2023, Nanjing, China}, publisher={EAI}, proceedings_a={PMBDA}, year={2024}, month={5}, keywords={object detection; yolov5; water litter; pyqt5}, doi={10.4108/eai.15-12-2023.2345400} }- Jiajia Meng
Liuling Lang
Zhenzhan Lu
Xiling Tang
Huimin He
Year: 2024
Design and implementation of Garbage Detection in Water Area Based on Yolov5 Algorithm
PMBDA
EAI
DOI: 10.4108/eai.15-12-2023.2345400
Abstract
The target detection of water waste helps to timely discover and warn the floating garbage in the water area, realize the 24-hour uninterrupted monitoring of water waste, provide data support for the corresponding garbage cleaning, improve the cleaning efficiency, and optimize the management of water environment.In this paper, the YOLOv5 algorithm is combined with floating garbage in waters, and the detection interface designed by PyQt5 is used to realize the convenient detection of images, videos and cameras of water garbage.The YOLOv5 model used in this study is trained on a new dataset, including 4591 images and 6622 bounding boxes of three types of common garbage. Five models of YOLOv5 are trained, and the optimal model under the same experimental conditions is selected. The model achieves 99.5% and 82.5% average precision values (mAP@0.5 and mAP@0.5:0.95).
