一种相机标定辅助的单目视觉室内定位方法  

An Indoor Location Method of Monocular Vision Assisted by Camera Calibration

作　　者：王勇[1,2] 陈国良 李晓园[3] 田鹏 周雨[2] 

机构地区：[1]中国矿业大学机电工程学院,江苏徐州221000 [2]徐州市产品质量监督检验中心,江苏徐州221000 [3]中国矿业大学环境与测绘学院,江苏徐州221000 [4]徐州市勘察测绘研究院,江苏徐州221000

出　　处：《测绘通报》2018年第2期35-40,共6页Bulletin of Surveying and Mapping

基　　金：国家重点研发计划(2016YFB0502105);国家自然科学基金(41371423);江苏省自然科学基金(BK20161181);江苏省质监局科技项目计划(KJ168351)

摘　　要：目前随着室内定位技术的发展,人们对室内定位精度提出了更高的要求,高精度、低成本的室内定位技术成为研究的热点。本文结合相机标定提出了一种单目视觉室内定位方法,将室内环境按照是否共面进行区域分割,并对分割后的各区域进行相机标定,对摄像机获取的视频进行运动目标检测获取目标像素坐标,利用像素坐标和相机与空间的位置关系并结合平面约束方程进行目标位置解算。试验表明:该算法定位精度在X方向小于0.15 m,Y方向小于0.35 m,平均单点定位时间小于5 s,获得了良好的室内定位效果。At present,with the development of indoor positioning technology,people put a higher demand for the accuracy of indoor positioning. High precision and low cost indoor positioning technology has become a hot research. In this paper,a kind of monocular vision indoor positioning technology is proposed,which combines the indoor environment according to whether or not the area is divided and the camera is calibrated.The moving target of the video acquired by the camera is used to obtain the target pixel coordinates.And it uses the moving target to detect the acquired pixel coordinates and the positional relationship between the camera and the space and combine the plane constraint equation for the target position.The experimental results show that the accuracy of the algorithm is less than0.15 m in the X direction and less than 0.35 m in the Y direction,and the average single point positioning time is less than 5 seconds,and a good indoor positioning effect is obtained.

关 键 词：室内定位 相机标定 平面分割 目标检测 坐标解算 

分 类 号：P228[天文地球—大地测量学与测量工程]