基于双目视觉的小场景三维轮廓提取

3D contour extraction of small scene based on binocular vision

为实现机器人移动过程中自主避障及轨迹修正,研究了小场景内基于立体视觉的环境信息的三维重构技术。用MATLAB双目相机标定工具箱对相机进行标定,SGM算法计算出相机左右视图的视差图,三角测距原理求取场景的三维点云信息。为适当简化环境参数,提出了基于二维轮廓图的三维轮廓提取方法,并借助模板匹配算法验证三维重建的准确性。在MATLAB环境下进行试验,结果表明:使用的方法比较准确地实现了小场景的三维轮廓的提取,说明提出的方法可以为机器人移动及其路径规划提供相应参考。

In order to realize autonomous obstacle avoidance and trajectory correction in the process of robot movement,the three-dimensional reconstruction technology of environmental information based on stereoscopic vision in small scene are studied.MATLAB binocular camera calibration toolbox is used to calibrate the camera.The SGM algorithm is used to calculate the parallax graph of the left and right view of the camera,and the three-dimensional point cloud information of the scene is obtained by using the principle of triangular distance measurement.In order to simplify the environment parameters,a three-dimensional contour extraction method based on the two-dimensional contour map is proposed.The accuracy of 3D reconstruction is verified by template matching algorithm.In the MATLAB environment,the results show that the method is used to extract more accurately the 3D contour in small scenes,which shows that the proposed method can provide reference for robot movement and its path planning.