高精度移动小车载具位姿检测平台仿真设计  

Orientation detection platform simulation design for the high precision moving vehicle carrier

作　　者：占红武[1] 丁信斌 胥芳[1] ZHAN Hongwu;DINGXinbin;XU Fang(College of Mechanical Engineering,Zhejiang University of Technology,Hangzhou 310014,China)

机构地区：[1]浙江工业大学机械工程学院,浙江杭州310014

出　　处：《浙江工业大学学报》2018年第5期564-569,共6页Journal of Zhejiang University of Technology

基　　金：国家自然科学基金资助项目(U1509212);浙江省科技计划项目(2017C31028;2011R50011-01)

摘　　要：高精度移动小车载具的定位机制和控制算法中,需要先对小车位姿进行高精度、实时的检测。为此设计了机器视觉位姿检测实验平台,选用高分辨率高帧率相机对小车进行实时拍照,通过面积法对相机光轴垂直度进行调节保证相机安装精度,选用世界坐标已知的X靶面作为标定板,采用改进的Harris算子提取图像亚像素级的特征角点。利用Matlab对相机进行标定和安装检测,得到世界坐标和图像坐标之间的映射关系和相机倾斜方向。实验结果表明:该实验平台可以达到预期的功能和精度需求,为后续小车位姿检测奠定了基础。In the location mechanism and control algorithm of the high precision moving vehicle carrier,the position and orientation should be detected firstly.The experimental platform for machine vision pose detection was designed and the carrier was constantly photographed by high resolution and high frame rate camera.In order to ensure camera mounting accuracy,the X target surface with world coordinate was used as calibration board,and the area method was used to adjust verticality of camera axis.Improved Harris operator was adopted to extract sub-pixel feature points.Mapping relation between world coordinates and image coordinates was acquired through using Matlab to calibrate the camera and test installation.The result shows that the experimental platform achieves the expected function and precision,which lays the foundation of orientation detection.

关 键 词：摄像机标定 数字图像处理 位姿检测 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置] TP391.7[自动化与计算机技术—控制科学与工程]