仲志远.一种基于图像识别的水位测量算法[J].国外电子测量技术,2017,36(6):96-99
一种基于图像识别的水位测量算法
Method of Water Level Data Capturing based on Video Image Recognition
  
DOI:
中文关键词:  视频  图像识别  水位量测  K-means聚类分析
英文关键词:video image pattern recognition water level measurement K-means cluster analysis
基金项目:
作者单位
仲志远 南京理工大学 机械工程学院 南京 210098 
AuthorInstitution
Zhong Zhiyuan College of Machanical Engineering, Nanjing University of Science and Technology, Nanjing 210098,China 
摘要点击次数: 499
全文下载次数: 1919
中文摘要:
      对传统水尺测量需人工读取,而压力式,超声波式,光电式等自动水位传感器存在维护困难,易损坏,需其他设施配套辅助等问题,本文提出一种利用摄像机捕获含有水尺图像的视频,并从视频图像中实时截取水位图像,对图像进行灰度转换,中值滤波,边缘检测等处理后,计算出水位结果的方法。水位数据主要由软件识别算法获得,具有测量精度高、维护方便、设备简单、易于推广等特点,实用化后可应用于城市涵洞、水库水位甚至闸位的测量。
英文摘要:
      Due to the traditional gauge measurement need human to read, and the pressure, ultrasonic, automatic water level sensor exists problems like difficult to maintain, easy to damage, need other auxiliary facilities, this paper proposes a method that capture video images containing draft using the camera, and intercept real-time water level image from the video, after gray scale conversion, median filtering, edge detection and other treatment, calculate the results of water level. The water level data is obtained by the software identification algorithm, which has the characteristics of high measurement precision, easy maintenance, simple equipment, easy to popularize and so on. This method can be applied to the measurement of the water level of reservoir, city culvert and even the gate position.
查看全文  查看/发表评论  下载PDF阅读器