大范围液位实时自动跟踪测量系统

针对传统液位测量方法的精度低、速度慢、测量范围受限制等问题，设计了一套基于视觉图像处理的高精度、快速的液位自动测量系统，它适用于测量范围大、液位变化速度快的实时液位测量。该系统以CCD摄像机采集的液位和钢尺图像为基础，通过钢尺刻度识别、钢尺刻线及液位识别三大识别模块完成液位高度测量，根据液位高度反馈控制电机带动CCD相机跟随液位实现实时自动跟踪测量。为了提高算法的识别速度，增强算法的鲁棒性，该系统采用了NCC(the normalized cross correlation)模板匹配数字识别方法，在刻线识别和液位识别过程中充分利用了图像的灰度特征和钢尺的结构特征完成识别。实验结果表明该测量系统的算法有着较高的鲁棒性以及较快的测量速度，可以完成精度在0.1mm以内的液位测量。

Automatic measuring system for high precision measurment on height of large scale water

In consideration of the traditional measurement method for the height of water with lower accuracy and slower measuring speed and small measuring scale, we design a set of height measuring system based on digital image process theory. This system process the image of ruler and water tube captured from CCD camera by three steps to complete the measuring task, that is: recognition of ruler number; identification of ruler scribed line; recognition of water level, then control the motor attached with the CCD camera to adjust the position of CCD camera to achieve real-time tracking measurement. During the recognition of ruler number, we use the NCC (the normalized cross correlation) template matching method with respect to the robustness of algorithm and stability of the measuring system. As to the identification of ruler scribed line and recognition of water level, we make full use of the gray features of the image and structure characteristics of the ruler. The experimental results show that this system has superior performance over the measurement of large scale water level with good robustness and it could complete a measurement with 0.1mm ac-curacy.