基于双目立体视觉的无人车越野路面三维重构研究

Research on Three-dimensional Reconstruction of Unmanned Vehicle Off-road Pavement Based on Binocular Stereo Vision

利用双目立体视觉技术对自主地面车辆行驶环境进行三维感知,搭建了无人车辆的双目立体视觉硬件系统,采用中值滤波对采集到的图像进行预处理,进而通过半全局立体匹配算法实现三维环境信息获取,并最终通过三维重构技术构建环境的数字高程图。该研究成果对自主地面车辆的环境识别具有重要意义。

In this paper,binocular stereo vision technology is used to conduct three-dimensional perception of the driving environment of unmanned off-road vehicles,a binocular stereo vision hardware system for unmanned vehicles is constructed,the median filter is used to preprocess the collected images,and then the three-dimensional environment information acquisition is realized through the semi-global stereo matching algorithm,and finally the digital elevation map of the environment is constructed by three-dimensional reconstruction technology.The results of this study are of great significance for the environmental identification of unmanned off-road vehicles.