Generalized LOAM: LiDAR Odometry Estimation With Trainable Local Geometric Features
Generalized LOAM: LiDAR Odometry Estimation With Trainable Local Geometric Features
复制标题
DOI：
10.1109/lra.2022.3219022
复制
发表时间：
2022-10
影响因子：
5.2
通讯作者：
Kohei Honda;Kenji Koide;Masashi Yokozuka;Shuji Oishi;A. Banno
中科院分区：
文献类型：
--
作者：
Kohei Honda;Kenji Koide;Masashi Yokozuka;Shuji Oishi;A. Banno
这封信提出了一个激光雷达里程计估计框架，称为广义壤土。与传统的LiDAR里程计和测绘(LAAM)方法相比，我们提出的方法可以无缝地融合点周围的各种局部几何形状，从而提高位置估计精度。为了利用连续几何特征进行LiDAR里程计估计，我们将微小神经网络引入到广义迭代最近点(GICP)算法中。这些神经网络利用局部几何特征改进了数据关联度量和匹配代价函数。用KITTI基准进行的实验表明，与其他激光雷达里程计估计方法相比，我们提出的方法减少了相对轨迹误差。
This letter presents a LiDAR odometry estimation framework called Generalized LOAM. Our proposed method is generalized in that it can seamlessly fuse various local geometric shapes around points to improve the position estimation accuracy compared to the conventional LiDAR odometry and mapping (LOAM) method. To utilize continuous geometric features for LiDAR odometry estimation, we incorporate tiny neural networks into a generalized iterative closest point (GICP) algorithm. These neural networks improve the data association metric and the matching cost function using local geometric features. Experiments with the KITTI benchmark demonstrate that our proposed method reduces relative trajectory errors compared to the other LiDAR odometry estimation methods.