针对传统激光扫描仪在地图创建中的缺点,提出一种基于Kinect深度数据模拟激光扫描的SLAM算法,对室内未知环境进行二维地图创建.通过装载Kinect传感器的移动机器人采集室内环境的彩色图和深度图,将深度图转换为三维点云,通过模拟激光扫描方法获得对应的激光扫描数据;利用SIFT算法与GTM算法相结合的方法进行特征点提取与匹配,实现室内未知环境的二维地图创建.实验对比结果表明,该方法对于室内二维地图创建有较好的效果,具有一定的理论和实际应用价值.%Aiming at the disadvantages of traditional laser scanner in map building,a SLAM algorithm for 2D map creation was proposed in unknown environment,which based on the Kinect depth data simulation of laser scanning.The indoor environmental color image and depth image were sampled by loading Kinect sensor of the mobile robot,depth image was converted to 3D point cloud and the corresponding laser scanned data were obtained using simulating laser scanning method.Through the combination of SIFT algorithm and GTM algorithm,the feature points were extracted and matched.Experimental comparison results show that this method has good effects on indoor 2D map building,and has certain theoretical and practical application value.
展开▼
