智能小车室内定位系统的设计与实现

室内定位技术越来越多地受到人们的关注。结合视频跟踪算法和单目视觉技术设计了一种室内小车定位系统,首先采用Camshift物体跟踪算法求出每帧图像中目标小车的中心位置像素坐标;然后再结合相机成像原理,建立室内场景的定位模型,推导出室内实际位置坐标和像素坐标之间的几何关系,并且利用Matlab中张正友标定工具箱求出相机的内部参数;最后结合推导出的几何关系公式计算出室内小车的实际位置坐标。实验表明,设计的定位系统能准确地实现室内小车的连续定位跟踪。

Design and Implementation of Intelligent Car Indoor Positioning System

This paper designs an indoor automotive positioning system based on video tracking algorithm and monocular vision technology.First,the coordinates of the central position of the targeted car in each frame are obtained by the Camshift object tracking algorithm.Then,combining with the principle of camera imaging,the positioning model of indoor scene is established,and the geometric relationship between the actual indoor position coordinates and the pixel coordinates is deduced in this paper.The internal parameters of the camera are obtained by using Zhang Zhengyou's Calibration Toolbox in matlab.Finally,the actual position coordinates of the indoor car are calculated with the derived geometric relationship formula.