Holger Rapp
0c3aa32ad2
Adds slides and announce next open house. ( #321 )
2017-06-09 15:10:20 +02:00
Holger Rapp
b427259214
Mark trimmed Submaps and remove from scan matching. ( #319 )
...
Related to #283 .
PAIR=wohe
2017-06-07 15:52:24 +02:00
Holger Rapp
c1d4d08a1d
Move AddProbabilityGridToResponse to the Submap object. ( #318 )
...
Working towards a thread-safe Submap object. Related to #283 .
PAIR=wohe
2017-06-07 15:13:56 +02:00
Wolfgang Hess
b3b9735a4b
Remove trimmed constraints. ( #317 )
...
Still missing marking submaps and nodes as trimmed.
PAIR=SirVer
2017-06-07 14:23:58 +02:00
Wolfgang Hess
a66147f131
Adds a PureLocalizationTrimmer. ( #316 )
...
Related to #315 .
2017-06-07 12:39:04 +02:00
Wolfgang Hess
c90b887bfb
Adds API to add trimmers. ( #314 )
...
This is related to #283 . First to be implemented for
pure localization and 2D SLAM.
PAIR=SirVer
2017-06-06 17:07:30 +02:00
Wolfgang Hess
56529e5968
Tiny cleanup of the MapBuilder. ( #313 )
...
GetTrajectoryConnectivity() was a convenience function.
Instead, sparse_pose_graph() should be used to get access to
GetConnectedTrajectories().
2017-06-06 15:20:10 +02:00
Wolfgang Hess
ba4f8ca9ee
Tiny cleanup. ( #312 )
2017-06-02 16:44:27 +02:00
Holger Rapp
0aa9730518
Announce Open House Hangout. ( #311 )
2017-06-02 14:00:25 +02:00
Wolfgang Hess
565e9c3eff
Simplify the handling of weights in the optimization problem. ( #309 )
...
This replaces the 3x3 (2D) or 6x6 (3D) matrix by the weights
for the translational and rotational component. The matrices
only contain these values now anyway, so this simplifies
the code and increases performance.
Also removes 2D from the UKF implementation which has not been
used for quite some time now.
2017-06-01 17:03:49 +02:00
Wolfgang Hess
92f154f561
Also use fixed covariances in 3D SLAM. ( #308 )
...
This follows #307 .
2017-06-01 15:04:07 +02:00
Wolfgang Hess
a2e52f81cf
Remove 2D scan matcher covariances. ( #307 )
...
This replaces the scaled covariances derived from the Ceres
scan matcher by directly configurable weights. Using covariances
did not provide the expected benefit, and replacing the scaling
matrix by two values will allow a faster evaluation of the cost
function in the future.
2017-05-31 11:56:32 +02:00
Wolfgang Hess
4bb02c7240
Add XRayPointsProcessor::bounding_box(). ( #306 )
...
Changes XRayPointsProcessor to use one bounding box for all floors.
Makes the bounding box accessible via a public member function.
2017-05-24 14:46:17 +02:00
Wolfgang Hess
3d0d1e46b5
Remove unused code. ( #304 )
2017-05-23 14:17:20 +02:00
Wolfgang Hess
515db3b4a9
Minor fixes. ( #303 )
...
Extract mapping_2d::CreateLocalTrajectoryBuilderOptions() into its own file.
Loop initialization fix in the trajectory_connectivity_test.
Thread safety fixes in sparse_pose_graph.
2017-05-22 14:55:05 +02:00
Wolfgang Hess
2d416589a7
Adds a NestedVectorsById<> class template. ( #300 )
2017-05-19 16:58:21 +02:00
Wolfgang Hess
0bba56428f
Remove flat indices. ( #299 )
...
This cleans up scan_index_to_node_id_, num_nodes_in_trajectory_,
scan_index. Related to #256 .
2017-05-19 14:01:59 +02:00
Damon Kohler
34b750977b
Do not ignore docker history command errors. ( #298 )
2017-05-19 08:29:09 +02:00
Wolfgang Hess
56fc2a9a92
Fix 3D SLAM loop closure issues. ( #296 )
...
3D submaps are now oriented approximately gravity aligned.
This is so that accumulating error in the local SLAM frame is
no longer a problem for finding loop closures. It also ensures
that the z search window size is approximately in the gravity
direction.
We now also pass an estimate of gravity orientation when doing
multi-trajectory matches. Otherwise trajectories starting with
different orientation relative to gravity could not be connected.
The gravity alignment is currently derived from the ImuTracker.
It might be possible to further improve on this by using the
latest gravity direction from the optimized poses.
2017-05-18 16:55:11 +02:00
Damon Kohler
22f673655f
First unstable development release ( #295 )
2017-05-18 13:32:08 +02:00
Wolfgang Hess
9033fad1ab
Always use the origin as the origin for grids. ( #294 )
...
Also removes convenience functions from mapping_3d::Submap.
PAIR=SirVer
2017-05-18 12:00:45 +02:00
Wolfgang Hess
33ce5dee37
Improve debug output and tiny bug fix. ( #293 )
...
The full submap search window in 3D SLAM was
not calculated correctly.
2017-05-17 15:21:39 +02:00
Wolfgang Hess
6c9e24d7c5
Keep 'trajectory_nodes_' per trajectory. ( #291 )
...
Fixes #256 .
PAIR=SirVer
2017-05-17 11:15:34 +02:00
Holger Rapp
5aa968748e
Sparse pose graph no longer uses Submaps* for trajectory identification. ( #290 )
...
Related to #256 .
PAIR=wohe
2017-05-16 14:24:49 +02:00
Holger Rapp
8612f5e6e5
ConstantData now has a trajectory_id. ( #289 )
...
Related to #256 .
PAIR=wohe
2017-05-16 11:42:18 +02:00
Wolfgang Hess
5effc4dac7
Remove flat indices from SparsePoseGraph::GetTrajectoryNodes(). ( #288 )
...
Related to #256 .
Also removes ComputeMapLimits with is moved to Cartographer ROS.
PAIR=SirVer
2017-05-16 10:51:08 +02:00
Wolfgang Hess
690d1893a7
Remove unused code. ( #287 )
2017-05-15 16:57:02 +02:00
Holger Rapp
8392a7e62c
Change SubmapState.scan_indices -> node_ids ( #286 )
...
Related to #256 .
PAIR=wohe
2017-05-15 16:31:59 +02:00
Jihoon Lee
f26e4366c9
Make SparsePoseGraph::constraints() public. ( #284 )
2017-05-15 16:06:20 +02:00
Wolfgang Hess
7d43eaa08c
Keep node data per trajectory in the optimization problem. ( #282 )
...
Related to #256 .
PAIR=SirVer
2017-05-15 14:09:45 +02:00
Holger Rapp
c084624958
Turn submap state into a per trajectory list. ( #281 )
...
Related to #256 .
2017-05-15 12:12:30 +02:00
Wolfgang Hess
fde9272533
Adds mapping:NodeId for Constraint. ( #279 )
...
Also extracted mapping::SubmapId into an id.h file.
Related to #256 .
PAIR=SirVer
2017-05-11 17:17:12 +02:00
Wolfgang Hess
32583e0c5d
Use trajectory IDs for global samplers. ( #278 )
...
Related to #256 .
PAIR=SirVer
2017-05-11 15:08:27 +02:00
Holger Rapp
1a9b18db69
Change num_submaps_in_trajectory to using trajectory IDs. ( #277 )
...
Related to #256 .
2017-05-11 14:32:55 +02:00
Holger Rapp
7b1915b906
Fix bug in ToProto introduced in #275 .
2017-05-11 14:02:04 +02:00
Holger Rapp
e6a6bab351
Make TrajectoryConnectivity use trajectory IDs. ( #275 )
...
Related to #256 .
2017-05-11 13:36:12 +02:00
Holger Rapp
95f0d013c6
Rename Constraint::i -> submap_id. ( #274 )
...
Related to #256 .
2017-05-11 11:43:47 +02:00
Holger Rapp
660eb341ff
Optimization problem only deals with trajectory IDs now. ( #273 )
...
Related to #256 .
PAIR=wohe
2017-05-10 17:12:38 +02:00
Wolfgang Hess
926b0320cb
Use per-trajectory SubmapData in OptimizationProblem. ( #272 )
2017-05-10 16:28:58 +02:00
Holger Rapp
94e8eec41d
Use (trajectory_id, submap_index) as submap IDs in places. ( #271 )
...
Replace flat submap index through (trajectory_id, submap_index) in Constraints.
Related to #256 .
PAIR=wohe
2017-05-10 12:26:58 +02:00
Holger Rapp
45de59b116
Move trajectory id mapping into the SparsePoseGraph. ( #270 )
2017-05-10 08:23:57 +02:00
Holger Rapp
524b613f2c
`submap_pose`s are now owned by the OptimizationProblem. ( #269 )
2017-05-09 17:24:48 +02:00
Holger Rapp
20e9cde53d
Fix race condition in extrapolation of submap poses. ( #268 )
...
PAIR=wohe
2017-05-09 15:21:30 +02:00
Wolfgang Hess
c664d7b966
Comply to CMake's CMP0037. ( #267 )
...
Test target names are generated from the directory and
source filename. Now "/"s are replaced by "."s which are
allowed by CMP0037.
2017-05-08 18:09:39 +02:00
Holger Rapp
772d02c00e
Move ConstantData into the SPG ( #263 )
2017-05-08 17:10:07 +02:00
Wolfgang Hess
9194c8679b
Clean up. ( #266 )
...
Move trajectory builder options.
Remove the barely used Interval template.
Inline the remaining CMakeLists.txt subdirectories.
2017-05-08 15:29:13 +02:00
Yutaka Takaoka
522b37979a
Enable multiple configuration for multi-trajectory ( #248 )
2017-05-08 11:58:53 +02:00
Holger Rapp
e2de27cded
Remove dead code. ( #262 )
2017-05-08 11:06:33 +02:00
Wolfgang Hess
79dc1f848f
Remove 'laser' references from parameter names. ( #259 )
...
Related to #250 .
2017-05-04 15:38:41 +02:00
Jochen Sprickerhof
12b3cc0d7b
Support new Debian version of gmock and add testing for debian systems ( #236 )
2017-05-04 12:47:12 +02:00