A key component for advanced driver assistance systems (ADAS) applications and autonomous robots is enabling awareness of where the vehicle or robot is, with respect to its surroundings and using this information to estimate the best path to its destination. The simultaneous localization and mapping (SLAM) process uses algorithms to estimate the pose of a vehicle and the map of the environment at the same time.
Lidar Toolbox™ provides a point cloud registration workflow that uses the fast point
feature histogram (FPFH) algorithm to stitch together point cloud sequences. You can use
this feature for progressive map building. Such a map can facilitate path planning for
vehicle navigation or can be used for SLAM. For an example of how to use the
extractFPFHFeatures function in a 3-D SLAM workflow for aerial data, see
Aerial Lidar SLAM Using FPFH Descriptors.
Lidar Toolbox also provides features for scan matching and simulating range-bearing sensor readings. These features are used in 2-D SLAM and obstacle detection workflows
|Register two point clouds using ICP algorithm|
|Register two point clouds using CPD algorithm|
|Register two point clouds using NDT algorithm|
|Extract eigenvalue-based features from point cloud segments|
|Extract fast point feature histogram (FPFH) descriptors from point cloud|
|Find matching features between point clouds|
|Map of segments and features for localization and loop closure detection|
|Display point clouds with matched feature points|
Understand point cloud registration and mapping workflow.
This example shows how to estimate a rigid transformation between two point clouds.
This example shows how to match corresponding features between point clouds using the
pcmatchfeatures function and visualize them using the