【作者】 江晓；

【导师】 彭侠夫； 陈嘉川；

【作者基本信息】 厦门大学 ， 控制工程， 2008， 硕士

【摘要】 图像测量是近年来在测量领域新兴的一种高性能测量技术,在机器视觉、图像匹配、图像检测和模式识别中都有重要应用。它以图像处理技术为基础,将光学、电子技术、计算机技术、几何测量和信号与信息处理技术等多种现代技术融为一体,构成综合体的测量系统。图像测量就是把测量对象的图像作为检测和传递信息的载体加以利用的精确测量技术,可有效地进行定位和识别。本文以图像处理技术为基础,深入研究了图像测量系统的总体方案及其实现技术。分别从数字图像测量系统的建立、图像预处理、直线的Hough变换、最小二乘法拟合、亚像素边缘定位等进行了理论分析和探讨,并在此基础上,构造了图像测量系统。用实验系统完成了对基本图像的线、圆、椭圆、不规则元件和边界分割等进行测量,对存在的问题和产生的误差进行了理论分析。实验结果表明,本文中提出的方法理论上正确,方法切实可行。采用通常非专用摄像机构造的系统,图像测量系统可以达到0.05-0.1mm精度,可以在一般精度要求系统中应用。更多还原

【Abstract】 Image measuring is a novel measure technique of high quality in the fields of measuring recently, and it has played an important role in machine video, image matching, and image detecting and pattern recognition. On the basis of image processing, the modern techniques of optics, electronics, computer, geometric, signal and information processing is melt into it as a whole to construct a measuring system. Image measuring is that the measured object image is considered as media to detect and transmit the measured object information, which takes advantages of accurately measuring technique, and the position and recognition of objects is efficiently preceded.On the basis of image processing, a total device and it’s implementation applications of image measuring system is deeply researched in this paper. Digital image measuring system setup, image preprocessing, line Hough transform, least-square fitting, sub-pixel edge location and so on are studied and analyzed, then an image measuring system is constructed on the basis of these work. The basic image elements of line, circle, ellipse, non-normal device and border segmentation are measured by our experimental system, and the existed problems and error derived from experiments are analyzed in theory. The experimental results show that the measuring method is correct in theory and can be implemented. When a normal camera is used to construct the image measuring system, a measure resolution of 0.05-0.10mm can be got by this system, and it can be used in normal applications.更多还原