The traditional way of river sewage outlets management highly relies on artificial inspection, which is time-consuming and labor-consuming, with limited covered areas. The drawbacks still exist when UAV inspection is now using for sewage outlets management, due to the lack of ability that automatically recognizes the river sewage outlets on UAV aerial images. To solve this problem, this research studies the sewage outlet recognition method based on deep learning. After the image preparation work such as image preprocessing, registration, and mosaic, a YOLOv3 model is selected among various common models, and adjusted to improve model performance. The research help management departments improve their working efficiency on river sewage outlets management, and also make a solid foundation for a long-term mechanism for the management.
Recognition Method of River Sewage Outlets in UAV Aerial Images Based on Deep Learning
Lecture Notes on Data Engineering and Communications Technologies
International conference on Big Data Analytics for Cyber-Physical-Systems ; 2021 ; Shanghai, China December 28, 2021 - December 29, 2021
2021 International Conference on Big Data Analytics for Cyber-Physical System in Smart City ; Chapter : 115 ; 1037-1045
2021-12-10
9 pages
Article/Chapter (Book)
Electronic Resource
English
Red River Waterway Revised Outlets for Red River Locks: Hydraulic Model Investigation
HENRY – Federal Waterways Engineering and Research Institute (BAW) | 1990
|Face recognition in images based on deep learning
IEEE | 2024
|Outlets capacity analysis of the Złotniki storage reservoir on the Kwisa River
BASE | 2006
|