1.引言
现代工业自动化生产中涉及到各种各样的检验、生产监视和零件识别应用，如汽车零配件批量加工的尺寸检查和自动装配的完整性检查、电子装配线的元件自动定位、IC上的字符识别等。通常这种带有高度重复性和智能性的工作是由肉眼来完成的，但在某些特殊情况下，如对微小尺寸的精确快速测量、形状匹配以及颜色辨识等，依靠肉眼根本无法连续稳定地进行，其它物理量传感器也难以胜任。人们开始考虑用CCD照相机抓取图像后送入计算机或专用的图像处理模块，通过数字化处理，根据像素分布和亮度、颜色等信息来进行尺寸、形状、颜色等的判别。这种方法是把计算机处理的快速性、可重复性与肉眼视觉的高度智能化和抽象能力相结合，由此产生了机器视觉测试技术的概念。
1. INTRODUCTION
Modern industrial automation involved in the production of a wide range of testing, production monitoring and parts identification applications such as automotive parts bulk of the size of check processing and automated assembly of the integrity of inspection, electronic components automatic assembly line positioning, IC on the character recognition And so on. This is usually a high degree of repeatability and intelligence work to be completed by the naked eye, but in certain special circumstances, such as the small size of the rapid precision measurements, shape and color matching identification,
8500字
