Abstract:Three-dimensional light detection and ranging (LiDAR) sensor-based outdoor environment truncated signed distance field (TSDF) mapping has been extensively studied in recent years. However, LiDAR points sparsity and varying density make it challenging to construct the smooth dense surface model. In this article, we alleviate such a problem by performing LiDAR data augmentation and formulating a novel TSDF mapping method to promote dense mapping quality. First, for each incoming LiDAR scan, range map projection and segmentation are performed, in which the point cloud segmentation plays the role of filtering unstable and irregular observations. Second, on the range map, 3-D points bilinear interpolation is performed in each effective patch with 2 pixels height and 2 pixels width. The generated points are regarded as the environmental information predictions based on a single LiDAR scan. Finally, we further augment the viewpoint along the normal of each point and input the augmented LiDAR data into a TSDF integrator. The proposed viewpoint augmentation can effectively realize projective TSDF values computation. Instead of using the widely applied dense mapping evaluation criteria, novel criteria are proposed, in which training scans are used for dense mapping, and testing scans are used for evaluating the surface model quality. The proposed mapping framework has been extensively evaluated on custom and public datasets. The experimental results demonstrate that our method is competitive with the state-of-the-art dense mapping methods. We have made our source code open to benefit the robotics community.

An Image Guided Algorithm for Range Map Super-resolution

Research On The Enhancement Of Laser Radar Range Image Recognition Using A Super-Resolution Algorithm

A MAP algorithm to super-resolution image reconstruction.

Range image super-resolution via guided image filter

Research on Super-resolution Reconstruction Method of Lidar 3D Range Profile

Three-dimensional Super-Resolution Range-Gated Imaging Based on Gaussian-range-intensity Model.

Hyperspectral Image Super Resolution Reconstruction with a Joint Spectral-Spatial Sub-Pixel Mapping Model.

Super-resolution of remotely sensed images based on map and discontinuity adaptive Markov Random field

A MAP Approach for Joint Motion Estimation, Segmentation, and Super Resolution

Multi-sensor super-resolution for hybrid range imaging with application to 3-D endoscopy and open surgery

A Coarse Elevation Map-based Registration Method for Super-resolution of Three-line Scanner Images

A Super-Resolution Reconstruction Algorithm for Surveillance Images

Adaptive algorithm of super-resolution image reconstruction

Novel Stereo Image Matching Technology Directed By Range Information

A high quality single-image super-resolution algorithm based on linear Bayesian MAP estimation with sparsity prior.

Single Depth Map Super-Resolution Via Joint Non-Local Self-Similarity Modeling and Local Multi-Directional Gradient-Guided Regularization

Range Map Interpolation Based 3D LiDAR Truncated Signed Distance Fields Mapping in Outdoor Environments

Super-resolution Single-Photon Imaging at 8.2 Kilometers

A High-Efficiency Super-Resolution Reconstruction Algorithm from Image/Video Sequences

A MAP Approach for Joint Image Registration, Blur Identification and Super Resolution

Single Depth Map Super-Resolution Via Joint Non-Local and Local Modeling