We introduce a novel method for odometry estimation using convolutional neural networks from 3D LiDAR scans. The original sparse data are encoded into 2D matrices for the training of proposed networks and for the prediction. Our networks show significantly better precision in the estimation of translational motion parameters comparing with state of the art method LOAM, while achieving real-time performance. Together with IMU support, high quality odometry estimation and LiDAR data registration is realized. Moreover, we propose alternative CNNs trained for the prediction of rotational motion parameters while achieving results also comparable with state of the art. The proposed method can replace wheel encoders in odometry estimation or supplement missing GPS data, when the GNSS signal absents (e.g. during the indoor mapping). Our solution brings real-time performance and precision which are useful to provide online preview of the mapping results and verification of the map completeness in real time.

We present a novel way of odometry estimation from Velodyne LiDAR point cloud scans. The aim of our work is to overcome the most painful issues of Velodyne data - the sparsity and the quantity of data points - in an efficient way, enabling more precise registration. Alignment of the point clouds which yields the final odometry is based on random sampling of the clouds using Collar Line Segments (CLS). The closest line segment pairs are identified in two sets of line segments obtained from two consequent Velodyne scans. From each pair of correspondences, a transformation aligning the matched line segments into a 3D plane is estimated. By this, significant planes (ground, walls, ...) are preserved among aligned point clouds. Evaluation using the KITTI dataset shows that our method outperforms publicly available and commonly used state-of-the-art method GICP for point cloud registration in both accuracy and speed, especially in cases where the scene lacks significant landmarks or in typical urban elements. For such environments, the registration error of our method is reduced by 75% compared to the original GICP error.

This paper presents a novel method for ground segmentation in Velodyne point clouds. We propose an encoding of sparse 3D data from the Velodyne sensor suitable for training a convolutional neural network (CNN). This general purpose approach is used for segmentation of the sparse point cloud into ground and non-ground points. The LiDAR data are represented as a multi-channel 2D signal where the horizontal axis corresponds to the rotation angle and the vertical axis represents channels - laser beams. Multiple topologies of relatively shallow CNNs (i.e. 3-5 convolutional layers) are trained and evaluated, using a manually annotated dataset we prepared. The results show significant improvement of performance over the state-of-the-art method by Zhang et al. in terms of speed and also minor improvements in terms of accuracy.

This paper deals with a scene pre-processing task - depth image segmentation. Efficiency and accuracy of several methods for depth map segmentation are explored. To meet real-time capable constraints, state-of-the-art techniques needed to be modified. Along with these modifications, new segmentation approaches are presented which aim at optimizing performance characteristics. They benefit from an assumption of human-made indoor environments by focusing on detection of planar regions. All methods were evaluated on datasets with manually annotated real environments. A comparison with alternative solutions is also presented.