双目视觉的新颖实时局部2维地图构建方法

根据轮式机器人移动的特点，提出一种采用双目视觉的新颖实时局部2维栅格地图构建算法。首先，提出虚拟高度线投影成像原理，将场景均匀栅格化，在栅格中引入虚拟高度线，并将其投影到立体视觉系统的立体图对中产生投影线，将求解场景点高度值的问题转化为在投影线上和水平视差搜索范围内寻找具有最大相似测度的对应点问题；然后，提出一种新颖的局部2维地图构建方法，该方法以机器人所能越过的最大高度为阈值，对高于阈值部分的虚拟高度投影线上的场景点由其在图像对中的相似测度确定其是否为障碍物区域。实验结果表明，该方法满足机器人导航所要求的有效性和实时性，并可应用到构建3维地图。

Novel real-time local two dimension map building method based on binocular vision

According to the characteristics of wheeled mobile robots, a novel method for real-time local two dimension(2D) map building based on binocular vision is proposed. Firstly, the principle of virtual height lines imaging in binocular vision system is proposed. The scene is partitioned into uniform grids, and virtual height line segments(VHLS) which do not exist in the scene are introduced. Each VHLS was projected into the stereo images in binocular vision system and formed a projection line segment(PLS).Finding the height of the scene is converted to finding the correspondence that has the maximum similarity metric on the PLS and in the searched horizontal disparity range. Then, a novel local 2D map building method is proposed. Similarity metric of the scene on the part of the PLS whose height is greater than the maximum which robots can overleap, is to determine whether the scene grid is an obstacle region or not. The experiment results show that the proposed method is valid and real-time for mobile robots, and can be used for building three dimension(3D) map.