ROS Index
Package Summary
| Tags | No category tags. |
| Version | 0.44.0 |
| License | Apache License 2.0 |
| Build type | AMENT_CMAKE |
| Use | RECOMMENDED |
Repository Summary
| Description | |
| Checkout URI | https://github.com/autowarefoundation/autoware_universe.git |
| VCS Type | git |
| VCS Version | main |
| Last Updated | 2025-04-25 |
| Dev Status | UNKNOWN |
| CI status | No Continuous Integration |
| Released | UNRELEASED |
| Tags | planner ros calibration self-driving-car autonomous-driving autonomous-vehicles ros2 3d-map autoware |
| Contributing |
Help Wanted (0)
Good First Issues (0) Pull Requests to Review (0) |
Package Description
Additional Links
Maintainers
- Kenzo Lobos-Tsunekawa
- Amadeusz Szymko
Authors
- Kenzo Lobos-Tsunekawa
autoware_cuda_pointcloud_preprocessor
Purpose
The pointcloud preprocessing implemented in autoware_pointcloud_preprocessor has been thoroughly tested in autoware. However, the latency it introduces does not scale well with modern LiDAR devices due to the high number of points they introduce.
To alleviate this issue, this package reimplements most of the pipeline presented in autoware_pointcloud_preprocessor leveraging the use of GPGPUs. In particular, this package makes use of CUDA to provide accelerated versions of the already established implementations, while also maintaining compatibility with normal ROS nodes/topics. <!– cSpell: ignore GPGPUs >
Inner-workings / Algorithms
A detailed description of each filter’s algorithm is available in the following links.
| Filter Name | Description | Detail |
|---|---|---|
| cuda_pointcloud_preprocessor | Implements the cropping, distortion correction, and outlier filtering (ring-based) of the autoware_pointcloud_preprocessor’s CPU versions. |
link |
| cuda_concatenate_and_time_sync_node | Implements pointcloud concatenation an synchronization following autoware_pointcloud_preprocessor’s CPU implementation. |
link |
(Optional) Future extensions / Unimplemented parts
The subsample filters implemented in autoware_pointcloud_preprocessor will have similar counterparts in this package.
Changelog for package autoware_cuda_pointcloud_preprocessor
0.44.0 (2025-04-18)
0.43.0 (2025-03-21)
- fix: update tool version
- Merge remote-tracking branch 'origin/main' into chore/bump-version-0.43
- chore(autoware_cuda_pointcloud_preprocessor): add maintainer (#10297)
- feat(autoware_cuda_pointcloud_preprocessor): a cuda-accelerated
pointcloud preprocessor
(#9454)
- feat: moved the cuda pointcloud preprocessor and organized from a personal repository
- chore: fixed incorrect links
- chore: fixed dead links pt2
- chore: fixed spelling errors
- chore: json schema fixes
- chore: removed comments and filled the fields
- fix: fixed the adapter for the case when the number of points in the pointcloud changes after the first iteration
- feat: used the cuda host allocators for aster host to device copies
* Update sensing/autoware_cuda_pointcloud_preprocessor/docs/cuda-pointcloud-preprocessor.md Co-authored-by: Max Schmeller <<6088931+mojomex@users.noreply.github.com>> * Update sensing/autoware_cuda_pointcloud_preprocessor/src/cuda_pointcloud_preprocessor/cuda_pointcloud_preprocessor.cu Co-authored-by: Manato Hirabayashi <<3022416+manato@users.noreply.github.com>> * Update sensing/autoware_cuda_pointcloud_preprocessor/src/cuda_pointcloud_preprocessor/cuda_pointcloud_preprocessor.cu Co-authored-by: Manato Hirabayashi <<3022416+manato@users.noreply.github.com>>
- style(pre-commit): autofix
* Update sensing/autoware_cuda_pointcloud_preprocessor/docs/cuda-pointcloud-preprocessor.md Co-authored-by: Max Schmeller <<6088931+mojomex@users.noreply.github.com>> * Update sensing/autoware_cuda_pointcloud_preprocessor/README.md Co-authored-by: Max Schmeller <<6088931+mojomex@users.noreply.github.com>> * Update sensing/autoware_cuda_pointcloud_preprocessor/README.md Co-authored-by: Max Schmeller <<6088931+mojomex@users.noreply.github.com>> * Update sensing/autoware_cuda_pointcloud_preprocessor/src/cuda_pointcloud_preprocessor/cuda_pointcloud_preprocessor.cu Co-authored-by: Max Schmeller <<6088931+mojomex@users.noreply.github.com>>
- style(pre-commit): autofix
* Update sensing/autoware_cuda_pointcloud_preprocessor/src/cuda_pointcloud_preprocessor/cuda_pointcloud_preprocessor.cu Co-authored-by: Manato Hirabayashi <<3022416+manato@users.noreply.github.com>>
- style(pre-commit): autofix
* Update sensing/autoware_cuda_pointcloud_preprocessor/src/cuda_pointcloud_preprocessor/cuda_pointcloud_preprocessor.cu Co-authored-by: Manato Hirabayashi <<3022416+manato@users.noreply.github.com>>
- style(pre-commit): autofix
- chore: fixed code compilation to reflect Hirabayashi-san's memory pool proposal
- feat: generalized the number of crop boxes. For two at least, the new approach is actually faster
- chore: updated config, schema, and handled the null case in a specialized way
- feat: moving the pointcloud organization into gpu
- feat: reimplemented the organized pointcloud adapter in cuda. the only bottleneck is the H->D copy
- chore: removed redundant ternay operator
- chore: added a temporary memory check. the check will be unified in a later PR
- chore: refactored the structure to avoid large files
- chore: updated the copyright year
- fix: fixed a bug in the undistortion kernel setup. validated it comparing it with the baseline
- chore: removed unused packages
- chore: removed mentions of the removed adapter
- chore: fixed missing autoware prefix
- fix: missing assignment in else branch
- chore: added cuda/nvcc debug flags on debug builds
- chore: refactored parameters for the undistortion settings
- chore: removed unused headers
- chore: changed default crop box to no filtering at all
- feat: added missing restrict keyword
- chore: spells
- chore: removed default destructor
- chore: ocd activated (spelling)
- chore: fixed the schema
- chore: improved readibility
- chore: added dummy crop box
- chore: added new repositories to ansible
- chore: CI/CD
- chore: more CI/CD
- chore: mode CI/CD. some linters are conflicting
- style(pre-commit): autofix
- chore: ignoring uncrustify
- chore: ignoring more uncrustify
- chore: missed one more uncrustify exception
* chore: added meta dep ---------Co-authored-by: Max Schmeller <<6088931+mojomex@users.noreply.github.com>> Co-authored-by: Manato Hirabayashi <<3022416+manato@users.noreply.github.com>> Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]\@users.noreply.github.com> Co-authored-by: Amadeusz Szymko <<amadeusz.szymko.2@tier4.jp>>
- Contributors: Amadeusz Szymko, Hayato Mizushima, Kenzo Lobos Tsunekawa
Wiki Tutorials
Package Dependencies
System Dependencies
Dependant Packages
Launch files
- launch/cuda_pointcloud_preprocessor.launch.xml
-
- input/pointcloud [default: /sensing/lidar/top/pointcloud_raw_ex]
- input/imu [default: /sensing/imu/imu_data]
- input/twist [default: /sensing/vehicle_velocity_converter/twist_with_covariance]
- output/pointcloud [default: /sensing/lidar/top/test]
- cuda_pointcloud_preprocessor_param_file [default: $(find-pkg-share autoware_cuda_pointcloud_preprocessor)/config/cuda_pointcloud_preprocessor.param.yaml]
- launch/cuda_concatenate_and_time_sync_node.launch.xml
-
- input/twist [default: /sensing/vehicle_velocity_converter/twist_with_covariance]
- output [default: /sensing/lidar/concatenated/pointcloud]
- param_file [default: $(find-pkg-share autoware_pointcloud_preprocessor)/config/concatenate_and_time_sync_node.param.yaml]