基于数学形态学光笔测量系统的特征点图像识别算法  

Feature Point Image Recognition Algorithm of the Light Pen Measurement System based on Mathematical Morphology

作　　者：徐玉玲[1] 李中伟[1] 钟凯[1] 王从军[1] 

机构地区：[1]华中科技大学材料成形与模具技术国家重点实验室,湖北武汉430074

出　　处：《新技术新工艺》2014年第6期65-71,共7页New Technology & New Process

基　　金：中央高校基本科研业务专项资金资助项目(HUST:2012QN036);材料成形与模具技术国家重点实验室自主创新基金资助项目(02-25-110093);国家科技支撑计划资助项目(2012BAF08B03);欧盟框架七项目"Casting of Large Ti Structures"资助项目(265697)

摘　　要：介绍了一种便携式光笔三坐标视觉测量系统,在分析测量系统成像特点的基础上,提出了一种基于数学形态学算法的图像数字处理方法,即通过在CCD摄像机镜头上加滤光片来隔离背景对图像产生亮斑的干扰。数学形态算法以光笔上被测特征点在CCD摄像机像面上的光斑图像轮廓所占的像素个数为基准,剔除过大和过小的图像轮廓,进而采用基于本算法的特征点图像识别方法来实现目标图像的获取,从而实现椭圆形光斑图像的识别。本文给出了该方法的具体实现过程,试验结果表明,该方法能够对背景复杂和灰度分布不均匀的图像实行有效的图像分割,实现目标图像和背景图像的分离,加快图像分析和处理速度,提高系统测量的效率和稳定性。The paper introduced a portable light pen vision coordinate measurement system,In the analysis of the imaging characteristics of the measurement system,Proposed a digital image processing method based on mathematical morphology algorithm.The usual method is to install a filter on the CCD camera to isolate the bright spots interference of the image from background light.This algorithm adopts the pixel number of the contour of the spot image of the measured feature points on the light pen on the CCD camera image plane as a benchmark,exclude too big and too small outline of the image contour;Further mathematical morphology algorithm based on feature point image recognition algorithm was adopted to achieve target image acquisition in order to achieve an oval spot image recognition.The implementation of the method and the experimental results showed that the method is able to do effective image segmentation for the complex background image of uniform intensity distribution to achieve the separation of the target image and the background image,accelerate the image analysis and processing speed,and improve the efficiency and stability of the measurement system.

关 键 词：便携式光笔测量系统 视觉坐标测量 图像处理 数学形态算法 回光反射标志点 

分 类 号：TN209[电子电信—物理电子学]