Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat(autoware_detected_object_validation): set validate distance in the obstacle pointcloud based validator #9663

Conversation

technolojin
Copy link
Contributor

@technolojin technolojin commented Dec 16, 2024

Description

A set of detected objects is validated by the obstacle pointcloud, which is the result of a rule-based detection ground segmentation.
In far ranges, however, it is difficult to separate non-ground obstacle since reflection of ground in far range is very week or absent due to the shallow angle.

A parameter validate_max_distance_m is added to not validate object far than this parameter.

Related links

Launcher PR autowarefoundation/autoware_launch#1277

How was this PR tested?

TIER IV INTERNAL

Experiment condition

  • ground segmentation setting : change parameters to make obstacle pointcloud less in far distance (for test purpose)

    • global_slope_max_angle_deg: 30.0
    • local_slope_max_angle_deg: 30.0
    • use_recheck_ground_cluster: false
  • obstacle pointcloud based validator set a short validation distance (for test purpose)

    • validate_max_distance_m: 40.0

Before

Blue bounding box: detected objects
Yellow bounding box: validated objects
White points: obstacle pointcloud

Screenshot from 2024-12-16 18-43-04

Since the obstacle pointclouds are fail to segment obstacles in far ranges, far objects are filtered out.

After

Blue bounding box: detected objects
Yellow bounding box: validated objects

Screenshot from 2024-12-16 18-59-16

Even the obstacle pointclouds are absent, objects far than 40m could pass the validator.

Notes for reviewers

The parameter validate_max_distance_m need to be set where the ground segmentation is less accurate. If this parameter is set too close, there is a risk that the detected is object is not validated.

Interface changes

None.

Effects on system behavior

None.

@technolojin technolojin self-assigned this Dec 16, 2024
@github-actions github-actions bot added component:perception Advanced sensor data processing and environment understanding. (auto-assigned) tag:require-cuda-build-and-test labels Dec 16, 2024
Copy link

github-actions bot commented Dec 16, 2024

Thank you for contributing to the Autoware project!

🚧 If your pull request is in progress, switch it to draft mode.

Please ensure:

@technolojin technolojin added the run:build-and-test-differential Mark to enable build-and-test-differential workflow. (used-by-ci) label Dec 16, 2024
@technolojin technolojin marked this pull request as ready for review December 16, 2024 10:13
Copy link

codecov bot commented Dec 16, 2024

Codecov Report

Attention: Patch coverage is 0% with 8 lines in your changes missing coverage. Please review.

Project coverage is 29.81%. Comparing base (219b8ca) to head (f63d6a3).
Report is 11 commits behind head on main.

Files with missing lines Patch % Lines
...tacle_pointcloud/obstacle_pointcloud_validator.cpp 0.00% 8 Missing ⚠️
Additional details and impacted files
@@           Coverage Diff           @@
##             main    #9663   +/-   ##
=======================================
  Coverage   29.80%   29.81%           
=======================================
  Files        1443     1443           
  Lines      108592   108581   -11     
  Branches    42664    42645   -19     
=======================================
  Hits        32369    32369           
+ Misses      73044    73033   -11     
  Partials     3179     3179           
Flag Coverage Δ *Carryforward flag
differential 5.29% <0.00%> (?)
total 29.81% <ø> (+<0.01%) ⬆️ Carriedforward from a85c67b

*This pull request uses carry forward flags. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

@Shin-kyoto Shin-kyoto added the run:deploy-docs Mark for deploy-docs action generation. (used-by-ci) label Dec 16, 2024
Copy link
Contributor

@Shin-kyoto Shin-kyoto left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

  • I did NOT check this PR by running autoware.

Copy link

Documentation URL: https://autowarefoundation.github.io/autoware.universe/pr-9663/
Modified URLs:

Copy link
Contributor

@YoshiRi YoshiRi left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@technolojin technolojin merged commit 178b6c0 into autowarefoundation:main Dec 17, 2024
48 of 49 checks passed
@technolojin technolojin deleted the feat/obst-point-validator/set-validate-range branch December 17, 2024 05:31
kyoichi-sugahara pushed a commit to kyoichi-sugahara/autoware.universe that referenced this pull request Dec 25, 2024
…he obstacle pointcloud based validator (autowarefoundation#9663)

* chore: add validate_max_distance_m parameter for obstacle_pointcloud_based_validator

Signed-off-by: Taekjin LEE <[email protected]>

* chore: optimize object distance validation in obstacle_pointcloud_validator

Signed-off-by: Taekjin LEE <[email protected]>

* chore: add validate_max_distance_m parameter for obstacle_pointcloud_based_validator

Signed-off-by: Taekjin LEE <[email protected]>

---------

Signed-off-by: Taekjin LEE <[email protected]>
saka1-s pushed a commit to saka1-s/autoware.universe that referenced this pull request Jan 9, 2025
…he obstacle pointcloud based validator (autowarefoundation#9663)

* chore: add validate_max_distance_m parameter for obstacle_pointcloud_based_validator

Signed-off-by: Taekjin LEE <[email protected]>

* chore: optimize object distance validation in obstacle_pointcloud_validator

Signed-off-by: Taekjin LEE <[email protected]>

* chore: add validate_max_distance_m parameter for obstacle_pointcloud_based_validator

Signed-off-by: Taekjin LEE <[email protected]>

---------

Signed-off-by: Taekjin LEE <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
component:perception Advanced sensor data processing and environment understanding. (auto-assigned) run:build-and-test-differential Mark to enable build-and-test-differential workflow. (used-by-ci) run:deploy-docs Mark for deploy-docs action generation. (used-by-ci) tag:require-cuda-build-and-test
Projects
Archived in project
Development

Successfully merging this pull request may close these issues.

3 participants