引江济淮工程高低联动智能视频感知技术研究
High-low linkage intelligent video perception technology in Yangtze-to-Huaihe Water Diversion
刘玉敏 1吕喆朋 2王岳飞 1朱献宏 1邵鹏2
作者信息
- 1. 安徽省引江济淮工程有限责任公司,安徽 合肥 230601
- 2. 中电信数智科技有限公司,北京 100036
- 折叠
摘要
引江济淮工程中原有的低点监控视野范围窄、抗干扰能力差、缺乏智能化功能,已无法满足智慧水利建设的要求.利用通信高塔作为挂载点,在引江济淮工程的关键标段植入高点智能化监控设备,从而构建出高低联动智能视频感知体系.其中利用人工智能技术,建立AI模型训练平台,实现了包含区域入侵侦测、水面漂浮物检测在内的多项视频侦测、检测功能.应用实践表明,在多种智能算法的赋能加持下,高低联动智能视频感知体系显著提高了报警准确率和智能识别距离,为数字引江济淮的建设提供了有效助力.
Abstract
The traditional low-point monitoring in the Yangtze-to-Huaihe Water Diversion had a narrow vision field,poor anti-interference ability and lack of intelligent functions,which was unable to meet the requirements of building intelligent water conservancy.In this study,the communication tower was adopted as a carrier,and the intelligent monitoring equipment was em-ployed in the key sections of the Yangtze-to-Huaihe Water Diversion to construct a high-low linkage intelligent video percep-tion system.Among them,a AI model training platform was established by using artificial intelligence technology,and a number of video detection functions including regional intrusion detection and surface floating object detection were realized.The application practice showed that with the support of various intelligent algorithms,the high-low linkage intelligent video perception system significantly improved the alarm accuracy and intelligent recognition distance and provided effective assistance for the construction of the digital Yangtze-to-Huaihe Water Diversion.
关键词
智慧水利/视频监控/高低联动/模型算法/引江济淮Key words
intelligent water conservancy/video monitoring/high-low linkage/model algorithm/Yangtze-Huaihe Water Diversion引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
维普
万方数据