摘要
以双目立体视觉测量为背景,以显著性标志物中的关键点为目标,提出了基于颜色阈值分割的关键点的实时检测和定位方法;关键点的检测和定位主要分为3个过程,一是图像的预处理部分,为后续的检测和定位提供基础;二是对预处理后的图像进行关键点检测,检测的方法首先分别通过颜色阈值分割、轮廓提取、多边形逼近以及设置矩形轮廓提取出关键点所在的显著性标志物,其次根据该显著性标志物的特点,采用hough变换提取线段,并通过最小二乘法进行直线拟合,求出关键点的精确的像素坐标;三是利用立体视觉三角测量原理,对求取的关键点进行精确的位姿计算;该方法实时性好、精度高,为后续的机器人视觉避障提供了一定的理论依据。
With the background of binocular stereo vision measurement and aiming at key point of obvious marker,a method based on color segmentation for key point detection and localization is presented.The detection and localization of key point includes three parts.The first part is image preprocessing for the following detection and localization.The second is key point detection,the detection method is through color segmentation,contour extraction,polygonal approximation,setting rectangular boundary and find the obvious marker.Then according to the feature of the obvious marker,we adopt Hough line segment detection,then fits the edge points by Lease Square Error（LSE） method,The coordinates of the key points are calculated by the equations of lines.Finally,the distance and orientation from camera to the key point is acquired accurately based on triangulation of binocular vision.The experiment result shows that the method greatly improves accuracy and real-time performance.It provides foundation for obstacle-avoidance of mobile robot.
出处
《计算机测量与控制》
CSCD
北大核心
2011年第7期1565-1568,共4页
Computer Measurement &Control
基金
国家863计划项目(2007AA04Z226)
北京市教育委员会科技发展计划面上项目(KM200810005016)
北京市教委科技创新平台项目(0020005466018)
关键词
双目视觉
关键点
阈值分割
定位
binocular vision
key point
threshold segmentation
localization
