A Framework for Constructing Large-Scale Dynamic Datasets for Water Conservancy Image Recognition Using Multi-Role Collaboration and Intelligent Annotation
Source abstract
Abstract
The construction of large-scale, dynamic datasets for specialized domain models often suffers with problems of low efficiency and poor consistency. This paper proposes a method that integrates multi-role collaboration with automated annotation to address these issues. The framework introduces two new roles, data augmentation specialists and automatic annotation operators, to establish a closed-loop process that includes dynamic classification adjustment, data augmentation, and intelligent annotation. Two supporting tools were developed: an image classification modification tool that automatically adapts to changes in categories and an automatic annotation tool with rotation-angle perception based on the rotation matrix algorithm. Experimental results show that this method increases annotation efficiency by 40% compared to traditional approaches, while achieving 100% annotation consistency after classification modifications. The method’s effectiveness was validated using the WATER-DET dataset, a collection of 1500 annotated images from the water conservancy engineering field. A model trained on this dataset achieved an F1-score of 0.9 for identifying water environment problems in rivers and lakes. This research offers an efficient framework for dynamic dataset construction, and the developed methods and tools are expected to promote the application of artificial intelligence in specialized domains.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Hydrological Forecasting Using AI · Hydrology and Watershed Management Studies · Flood Risk Assessment and Management
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Civil Engineering and Architecture, Xi’an University of Technology, Xi’an 710048, China
Apparel & Art Design College, Xi’an Polytechnic University, Xi’an 710048, China
School of Water Resources and Hydro-Electric Engineering, Xi’an University of Technology, Xi’an 710048, China
References (31)
Show source-provided references
1. Wang, J., and Song, X. (2022). A Review of the Application Research of Computer Vision Technology. J. Comput. Era, 10.
2. Lung Nodule Detection Method Based on Improved YOLO V3
3. PID Target Detection Based on Improved YOLO V3
4. Embedded Dorsal Vein Detection Based on Improved YOLO Nano
5. A Risk Prediction Model for Post–operative Healthcare–Associated Infections Based on Deep Learning
6. Research on Intelligent Analysis and Verification Algorithms for Medical Data Based on Deep Learning
7. ECARA–YOLO: An Improved Printed Circuit Board Defect Detection Algorithm
8. Research on a Tobacco Leaf Debris Removal System Based on an Improved YOLO Model
9. Li, J., Qiu, D., Yang, H., and Liu, K. (2020). Research on the Workpiece Recognition Method Based on Improved YOLO v3. Modul. Mach. Tool Autom. Manuf. Tech., 8.
10. Lightweight DCN–YOLO for Steel Strip Surface Defect Detection in Complex Environments
11. (2020). Research and Design of Intelligent Security System for the Middle Route Project of South–to–North Water Diversion. Yellow River, 12, 119–134.
12. Analysis of the Application of Remote Security Intelligent Monitoring Platform in the Energy Industry
13. An Intelligent Monitoring System for Laboratory Security
14. The Intelligent Monitoring System Builds a Security Fortress for the Xi’an International Horticultural Exposition
15. Intelligent Monitoring May Become the Next Inflection Point of the Security Industry
16. Shao, W., Wang, R., Dong, H., and Jia, X. (2022, January 25). The Impact of Chronic Diseases on the Health–Promoting Lifestyle of Community–Dwelling Elderly. Proceedings of the 12th National Sports Science Congress, Rizhao, China.
17. COVID–19 Medical Image Datasets and Research Progress
18. Enlightenment of the Upgrading of the Minimum Data Set of the American Long–Term Care Assessment System to the Elderly Nursing Assessment in China
19. Revision Research on the Basic Dataset for Disease Control
20. Rui, A. (2023). Prediction of miRNA-Disease Association Relationships by Multiview Convolutional Neural Networks. [Master’s Thesis, Southwest Minzu University].
21. Exploration: Intelligent Monitoring Technology Achieves Full–Scale Security Automation
22. Asynchronous Update Reinforcement Learning Decision–Making Algorithm for Operating Indexes of Uncertain Industrial Processes
23. Shen, T. (2022). Research and Application of the Process Object Data Update Method Based on Dynamic Clustering. [Master’s Thesis, University of Jinan].
24. Cheng, X. (2022). Research on Multi–Fault Diagnosis Methods for Industrial Processes with Class–Imbalanced Data. [Master’s Thesis, Beijing University of Chemical Technology].
25. Incremental Update Algorithm for (p,k)–Anonymous Datasets
26. An Auxiliary Annotation Method Based on Labelimg
27. Defect Detection for Industrial Neutron Radiographic Images Based on Modified YOLO Network10.1016/j.nima.2023.168694 ↗
28. Deng, X. (2023). Research on the UAV Image Vehicle Target Detection Algorithm Based on Improved YOLOv5s Model. [Master’s Thesis, East China University of Technology].
29. Design and Implementation of the LabelMe Annotation Verification System
30. Research on the Update Method of the Whole–Set Water Network Dataset
31. Fish Detection in Underwater Environments Using Deep Learning10.1007/s40009-023-01265-4 ↗