v0.22.1

First, a big thanks for our new contributors (see below) and everyone reporting/discussing issues!

Dependencies included in the Windows binaries:

• Core: OpenCV 4.7.0 with xfeatures2d and nonfree modules (GPU options enabled on CUDA 11.7 release), PCL 1.15.0 with VTK 9.3, Qt 6.8.3.
• Drivers:
  • OpenNI and OpenNI2 (Kinect for XBOX 360, Xtion Live Pro, Orbbec Astra) *Kinect for XBOX 360 reported not working on Windows 11 (but works on Windows 10),
  • Freenect2 (UsbDk v1.00-22 should be installed, cuda pipeline available in cuda11.7 release),
  • Kinect for Windows SDK 2.0 (Kinect for XBOX ONE),
  • Kinect For Azure,
  • RealSense2 2.56.3 (T265, D415, D435, D435i, L515, D400 Series visual presets can be used),
  • Tara (side by side opencv video capture),
  • ZED SDK 5.0.3 (enabled on cuda11.7 release),
  • DepthAI 2.24.0 (OAK-D, OAK-D Lite)
• Optimizers: TORO, g2o 2024-12-14, GTSAM 4.2.0, Ceres 2.2.0
• ICP: libpointmatcher 1.3.1, CCCoreLib 1.0
• Reconstruction: OctoMap
• DNN: SuperPoint support with libtorch (enabled on cuda11.7 release)

Dependencies included in the MacOS binaries:

• Core: OpenCV 4.11.0 with xfeatures2d and nonfree modules, PCL 1.15.0 with VTK 9.4, Qt 6.9.0.
• Drivers:
  • RealSense2 2.56.3 (T265, D415, D435, D435i, L515, D400 Series visual presets can be used) *may crash on some systems or require launching with sudo (e.g., not currently working on M1 Sonoma/Sequoia),
  • Tara (side by side opencv video capture),
  • DepthAI 2.30.0 (OAK-D, OAK-D Lite)
• Optimizers: TORO, g2o, GTSAM 4.1.0
• ICP: libpointmatcher 1.4.4
• Reconstruction: OctoMap 1.10.0
• Homebrew binaries now available: brew install rtabmap (add --HEAD for latest master version). Note that it has only basic dependencies to visualize, edit or post-process databases (e..g, coming from RTAB-Map for iOS app).

What's Changed

• Remove dependency on qt_gui_cpp. by @clalancette in #1072
• not limitKps inside Feature2D::generateKeypoints when using pyDetector by @hellovuong in #1223
• OAK Internal Sync & DepthAI HF-Net Support by @borongyuan in #1193
• iSAM2 integration by @matlabbe in #1249
• GlobalDescriptor (PyDescriptor / netvlad) by @matlabbe in #1163
• Create scheduled-stats.yml by @doumdi in #1260
• copy global descriptor by @borongyuan in #1263
• LiDAR capture support in standalone library by @matlabbe in #1264
• adjust likelihood using z-score by @borongyuan in #1272
• Add ANMS (SSC method) by @borongyuan in #1276
• Add SeerSense Support by @borongyuan in #1294
• Add flags for OpenNI by @Serafadam in #1299
• Charuco calibration + depthai calibration support by @matlabbe in #1302
• Switch the octomap dependency to the system version. by @clalancette in #1314
• Remove iOS bitcode support by @kemchenj in #1326
• Fixing ChArUco board support on opencv >= 4.7.0 by @matlabbe in #1331
• Adding OpenCV's GPU GFTT/ OpticalFlow and CudaSift support by @matlabbe in #1330
• Supporting GTSAM bearing/range factors for landmarks by @matlabbe in #1346
• Clicking Nodes\Links brings them up in viewer by @Nosille in #1344
• Node highlighting addition to previous pull request by @Nosille in #1348
• Update iOS app's install_deps.sh to remove PCL's SSE dependency by @tomaja in #1351
• iOS various updates by @matlabbe in #1340
• Sort keypoints before SSC by @borongyuan in #1364
• Adding FOURCC usb camera support by @matlabbe in #1366
• Draw "1 - Loop Thr" along with posterior by @borongyuan in #1373
• Concise layout by @Nosille in #1381
• Applying suggestion of #1388 by @matlabbe in #1390
• Added floor filtering options for features extraction by @matlabbe in #1408
• Support Ceres 2.1+ Manifolds in backwards-compatible manner by @ahojnnes in #1405
• Export tool memory optimization by @matlabbe in #1391
• add support for RVL depth image compression by @borongyuan in #1409
• fix build with ceres<2.1 by @borongyuan in #1437
• fixed ORB_SLAM preproc breaking gui conditionals by @laurence-diack-pk in #1445
• RAM: add entilement com.apple.developer.kernel.increased-memory-limit… by @zeitlerv in #1446
• Adding options to bound OdometryF2M keyframes by @matlabbe in #1448
• DBReader: added fine-grained camera selection by @matlabbe in #1457
• Further improvements to DepthAI integration by @borongyuan in #1458
• Adding support for compressed input data and right color image. by @matlabbe in #1463
• Improved depth estimation of mono features by @matlabbe in #1460
• iOS feb 2025 various updates by @matlabbe in #1451
• Enable on-device feature detectors for color camera by @borongyuan in #1470
• Fix rgb camera size setting and capture delay by @borongyuan in #1476
• Fixed cv::remap() crach when the neural network did not detect any features by @borongyuan in #1488
• New parameter: RGBD/LocalizationSecondTryWithoutProximityLinks by @matlabbe in #1499
• Fix ctor by @phodina in #1496
• Fix 6DoF graph angular check by @matlabbe in #1511
• Handle edge case when SuperPoint detects only 1 keypoint by @torjusti in #1509
• New feature: Depth confidence by @matlabbe in #1520
• Fixed a typo with SuperPoint by @pzvelev in #1535
• Remove typo in ceres optimizer branch by @HO4X in #1544
• Added libceres-dev dep in CI by @matlabbe in #1547
• ImageView: Depth color map in base frame by @matlabbe in #1553

New Contributors

• @clalancette made their first contribution in #1072
• @hellovuong made their first contribution in #1223
• @doumdi made their first contribution in #1260
• @Serafadam made their first contribution in #1299
• @kemchenj made their first contribution in #1326
• @Nosille made their first contribution in #1344
• @tomaja made their first contribution in #1351
• @ahojnnes made their first contribution in #1405
• @laurence-diack-pk made their first contribution in #1445
• @zeitlerv made their first contribution in #1446
• @phodina made their first contribution in #1496
• @torjusti made their first contribution in #1509
• @pzvelev made their first contribution in #1535
• @HO4X made their first contribution in #1544

Full Changelog: 0.21.4...0.22.1

Other

• Looking for ICP configuration file for L515 and Kinect Azure, see release 0.20.8.
• Windows High CPU Usage: this can be related to OpenMP, try setting environment variable OMP_WAIT_POLICY to passive.
• SuperPoint weights can be downloaded in assets below for convenience (tested with CUDA release).
• For ZED's neural depth option, download and extract the files from RTABMap-0.22.1-win64-cuda117-zed-neural-extra.7z in the binary folder of RTAB-Map. Note that the first time neural option is used, RTAB-Map may appear frozen for some time, just wait till the model is initialized (could take a be a couple of minutes).
• See Installation page for additional installation instructions depending on your platform and the camera you are using (e.g., drivers).