基于嵌入式Linux的水位视频在线监测系统设计

陈城1,2，孙峰3，曲金秋1,2，吴婷1,2，褚泽帆1,2

（1.水利部南京水利水文自动化研究所，江苏 南京 210012；

2.水利部水文水资源监控工程技术研究中心，江苏 南京 210012；

3.江苏南水科技有限公司，江苏 南京 210012）

　　摘 要：为有效提高水位数据在线监测技术的自动化、智能化，针对当前传统水文仪器监测设备的微控制器(C51，MSP430及STM32等)主频低、智能性、鲁棒性差以及水位监测数据的不精确、实时性差以及成本高等问题，提出采用嵌入式Linux系统、图像处理技术以及无线传输技术等实现水位数据在线自动监测。首先，利用嵌入式Linux系统(bootloader、Linux内核以及rootfs文件系统)定时控制网络摄像机获取高清图像，并在嵌入式Linux系统进行图像识别处理，然后将识别的水位信息和采集的图像通过无线模块传输至控制中心，以期实现水位视频在线监测与图像查询。该方法相比传统微控制器，监测水位数据处理速度更快、智能化程度更高、水位数据精度高、成本低实时性强等优点，并能够进一步推动水文仪器现代化建设发展。

　　关键词：水位监测；图像识别；嵌入式Linux；无线传输

Design of video on-line monitoring system of water level based on embedded Linux

CHEN Cheng1,2, SUN Feng3, QU Jinqiu1,2, WU Ting1,2, CHU Zefan1,2

(1. Nanjing Automation Institute of Water Conservancy and Hydrology, Ministry of Water Resources, Nanjing 210012, China;

2. Hydrology and Water Resources Engineering Research Center for Monitoring, Ministry of Water Resources, Nanjing 210012, China;

3. Jiangsu NAIWCH Co.,Ltd., Nanjing 210012, China)

　　Abstract: In order to effectively improve the automatization and intellectualization of on-line monitoring technology of water level data, aiming at the problems of low frequency, low intelligence and poor robustness of traditional hydrological monitoring equipment microcontrollers (C51，MSP430 and STM32, etc.), and inaccuracy, poor real-time performance and high cost of water level monitoring data, this paper realizes on-line automatic monitoring of water level data using embedded Linux system, image processing technology and wireless transmission technology. Image recognition and processing is carried out in the embedded platform, by using embedded Linux system( bootloader, Linux kernel and rootfs file system) to control network camera to periodically acquire high-definition image. Then, the recognized water level information and acquired image are transmitted to the control center by wireless module, so as to realize on-line video monitoring and image query of water level. Compared with traditional Micro controller monitoring of water level, this method has the advantages of quicker processing speed, higher intelligence, higher data precision, lower cost and stronger real-time performance, which can further promote the construction and development of hydrological instruments modernization.

　　Keywords: water level monitoring, image recognition, embedded Linux, wireless transmission