摘要
研制了基于嵌入式Linux操作的边缘计算终端,推动系统由自动化向自主化发展,研究国内领先的AI图像识别技术,通过水尺自动定位算法进行水位识别,白天、夜晚水尺刻度清晰可见,根据视频图像法测水位在水库、灌区等应用中的特点,不断优化完善算法。
The edge computing terminal based on embedded Linux operation is developed to promote the development of the system from automation to autonomy and to study the leading AI image recognition technology in China.The water level is recognized through the automatic positioning algorithm of the water gauge.The scale of the water gauge is clearly visible during the day and at night.According to the characteristics of the application of the video image method to measure water level in reservoirs and irrigation areas,the algorithm is constantly optimized and improved.
作者
吴婷
褚泽帆
孙勇
丁超
WU Ting;CHU Zefan;SUN Yong;DING Chao(Nanjing Automation Institute of Water Conservancy and Hydrology,Nanjing 210012,China;Rearch Center on Hydrology&Water Resources Monitoring,Nanjing 210012,China;Water Conservancy Bureau of Gaoyou City,Yangzhou 225600,China;Jiangsu Hongyuan Bidding Acting Co.,Ltd.,Nanjing 210019,China)
出处
《江苏水利》
2023年第8期55-60,共6页
Jiangsu Water Resources
基金
江苏省水利科技项目(2021070)。
关键词
边缘计算
嵌入式开发
形态学处理
HOUGH变换
畸变校正
edge computing
embedded development
morphological processing
Hough transform
distortion correction
