3D simultaneous localization and mapping based on multi-layered Normal Distribution Transform

Normal Distribution Transform (NDT) is an alternative scan matching approach to the well-known Iterative Closest Point (ICP) algorithm. NDT provides superior properties over ICP for large scale and outdoor SLAM since it represents the surface in a special compact form, and it is more robust in noisy environment. The main bottleneck of the NDT algorithm is that it does not offer a certain cell size and requires shifted overlapping grids to reduce discretization errors. To overcome this problem, we introduce Multi-Layered Normal Distribution Transform which differs from the conventional method in three ways. The main difference is that the cell size is automatically assigned by the introduced algorithm and no need for shifting grids, which reduces the computational cost effectively. Another difference is the implementation of different optimization techniques and minimized score function which provides faster convergence and long range measurement capability. Finally, the proposed method is combined with the Kalman Filter for navigation purpose in 3D space. The experimental results show that the ML-NDT can be used in SLAM problem successfully.