autoware_probabilistic_occupancy_grid_map

Purpose

This package outputs the probability of having an obstacle as occupancy grid map.

References/External links

• Pointcloud based occupancy grid map
• Laserscan based occupancy grid map
• Grid map fusion

Settings

Occupancy grid map is generated on map_frame, and grid orientation is fixed.

You may need to choose scan_origin_frame and gridmap_origin_frame which means sensor origin and gridmap origin respectively. Especially, set your main LiDAR sensor frame (e.g. velodyne_top in sample_vehicle) as a scan_origin_frame would result in better performance.

Parameters

• binary bayes filter updater

  {{ json_to_markdown(“perception/autoware_probabilistic_occupancy_grid_map/schema/binary_bayes_filter_updater.schema.json”) }}

• grid map

  {{ json_to_markdown(“perception/autoware_probabilistic_occupancy_grid_map/schema/grid_map.schema.json”) }}

• laserscan based occupancy grid map

  {{ json_to_markdown(“perception/autoware_probabilistic_occupancy_grid_map/schema/laserscan_based_occupancy_grid_map.schema.json”) }}

• multi lidar pointcloud based occupancy grid map

  {{ json_to_markdown(“perception/autoware_probabilistic_occupancy_grid_map/schema/multi_lidar_pointcloud_based_occupancy_grid_map.schema.json”) }}

• pointcloud based occupancy grid map

  {{ json_to_markdown(“perception/autoware_probabilistic_occupancy_grid_map/schema/pointcloud_based_occupancy_grid_map.schema.json”) }}

• synchronized grid map fusion

  {{ json_to_markdown(“perception/autoware_probabilistic_occupancy_grid_map/schema/synchronized_grid_map_fusion_node.schema.json”) }}

Downsample input pointcloud(Optional)

If you set downsample_input_pointcloud to true, the input pointcloud will be downsampled and following topics are also used. This feature is currently only for the pointcloud based occupancy grid map.

• pointcloud_based_occupancy_grid_map method

# downsampled raw and obstacle pointcloud
/perception/occupancy_grid_map/obstacle/downsample/pointcloud
/perception/occupancy_grid_map/raw/downsample/pointcloud

• multi_lidar_pointcloud_based_point_cloud

# downsampled raw and obstacle pointcloud
/perception/occupancy_grid_map/obstacle/downsample/pointcloud
/perception/occupancy_grid_map/<sensor_name>/raw/downsample/pointcloud

Test

This package provides unit tests using gtest. You can run the test by the following command.

colcon test --packages-select autoware_probabilistic_occupancy_grid_map --event-handlers console_direct+

Test contains the following.

• Unit test for cost value conversion function
• Unit test for utility functions
• Unit test for occupancy grid map fusion functions
• Input/Output test for pointcloud based occupancy grid map

Changelog for package autoware_probabilistic_occupancy_grid_map

0.53.0 (2026-09-29)

• Merge remote-tracking branch 'origin/main' into prepare-0.53.0-changelog

• feat(probabilistic_occupancy_grid_map): apply [agnocast_wrapper::Node]{.title-ref} to [pointcloud_based_occupancy_grid_map]{.title-ref} (#13387)

  • feat(probabilistic_occupancy_grid_map): apply [agnocast_wrapper::Node]{.title-ref} to [pointcloud_based_occupancy_grid_map]{.title-ref}

  * feat(probabilistic_occupancy_grid_map): run the occupancy grid map node as its own process under Agnocast ---------

• Contributors: Koichi Imai, Ryohsuke Mitsudome

0.52.0 (2026-06-30)

0.51.0 (2026-05-01)

• Merge remote-tracking branch 'origin/main' into tmp/bot/bump_version_base

• chore(perception): move perception node configuration file to each package (#12440) move perception node configuration file to each package

• refactor(autoware_ground_segmentation/launch, autoware_probabilistic_occupancy_grid_map/launch): use launch substitution instead of get_package_share_directory (#12392)

  • refactor: use launch substitution instead of get_package_share_directory

  * use existing file for default ---------

• refactor(autoware_universe): use autoware_ament_auto_package in perception utility packages (#12281) Co-authored-by: github-actions <<github-actions@github.com>>

• feat(autoware_probabilistic_occupancy_grid_map): cuda 12.0 build compatibility (#12187)

  • feat(autoware_probabilistic_occupancy_grid_map): CUDA 12.0+ build compatibility

  * feat: restore Turing arch ---------

• fix(autoware_probabilistic_occupancy_grid_map): remove dead operation (#12159) fix: remove dead operation

• Contributors: Amadeusz Szymko, Taekjin LEE, Taeseung Sohn, Vishal Chauhan, github-actions

0.50.0 (2026-02-14)

• Merge remote-tracking branch 'origin/main' into humble
• feat(autoware_probabilistic_occupancy_grid_map): update nvcc flags (#12052)
• feat!: remove ROS 2 Galactic codes (#11905)
• fix(probabilistic_occupancy_grid_map): add missing [#include <cstdint>]{.title-ref} (#11919)
• Contributors: Amadeusz Szymko, Mete Fatih Cırıt, Ryohsuke Mitsudome

0.49.0 (2025-12-30)

• Merge remote-tracking branch 'origin/main' into prepare-0.49.0-changelog

• docs: fix broken links (#11815)

• fix(autoware_probabilistic_occupancy_grid_map): prevent local_map leak in occupancy grid map updateOrigin (#11765)

  • fix(occupancy_grid_map): avoid local_map leak in updateOrigin
  • style(pre-commit): autofix

  * Fix local map resize type and format copyMapRegion ---------Co-authored-by: pre-commit-ci-lite[bot] <117423508+pre-commit-ci-lite[bot]@users.noreply.github.com>

• Contributors: Hongliang Liu, Mete Fatih Cırıt, Ryohsuke Mitsudome

0.48.0 (2025-11-18)

File truncated at 100 lines see the full file