摘要
研究了机器视觉技术在工业分拣问题中的应用,介绍开发过程中使用的主要图像处理算法。利用Visual C++编程软件建立图像处理的算法库,实现对规则几何工件的识别和定位。分析视觉算法库中的摄像机模型,并且求解出摄像机内外参数、畸变参数、旋转和平移向量。然后对图像序列进行图像预处理、Hough圆检测、角点提取和轮廓识别等运算分析。提出计算多目标中心的算法和一种角点特征结合轮廓特征识别算法。仿真结果表明,该视觉算法可以有效解决规则几何工件的分拣问题,并且能准确地计算工件中心,周长,面积和边缘,以达到分拣的目的。
In this paper we study the application of machine vision technology in industrial sorting,and introduce the principal image processing algorithm used in development process.Visual C＋ ＋ programming software is employed to establish image processing algorithms library,which implements the recognition and positioning of the geometric workpieces.We analyse the camera model in OpenCV algorithms library,and calculate the internal and external parameters of the camera,distortion parameters,translation and rotation vectors,etc.For image sequence,the operations and analyses including image preprocess,Hough circle detection,corner extraction and contour recognition are conducted afterwards.Meanwhile,we put forward an algorithm of multi-objective centre calculation and an identification algorithm combines the corner feature with contour characteristics.Simulation results show that the vision algorithm can effectively solve the sorting problem of regular geometric workpieces,and can accurately calculate the workpiece centre,perimeter,area and edge to achieve the purpose of sorting.
出处
《计算机应用与软件》
CSCD
北大核心
2012年第11期87-91,共5页
Computer Applications and Software
基金
国家高技术研究发展计划项目(2012AA041405)
沈阳市工业科技攻关项目(F12-010-2-00)
关键词
机器视觉
工业机器人
分拣
目标识别
图像处理
Machine vision Industrial robot Sorting Target recognition Image processing
