feat(viewer): add support of rendering pointcloud segmentation #261
+146
−21
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp>
Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp>
* chore(deps): restrict rerun-sdk to versions >=0.20.0,<0.28.0 (#248) * Bump version from 0.5.2 to 0.5.3 (#247) * Bump version from 0.5.2 to 0.5.3 * Pin rerun-sdk dependency to version 0.20.0 * refactor: update custom validator definition (#250) * refactor: update custom validator definition Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp> * Update t4_devkit/schema/tables/base.py Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> --------- Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> * fix: modity type of size to tuple[int, int] (#251) Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp> * fix: abstract away category indexing differences Semseg and non-semseg datasets behave differently (position-based vs. explicit indexing). Now computing index field in case position-based indexing is used. Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> * fix(rendering): don't fail when map is not present Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> * fix(rendering): fix iteration freezing when skipping a pointcloud without lidarseg file Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> * feat(rendering): auto-enable SEGMENTATION coloring when lidarseg is available Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> * docs: checkmark for semseg viz Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> * fix(rendering): fix wrongly named `color_mode` args Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> * refactor(compatibility): move to schema package Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> --------- Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp> Signed-off-by: Max SCHMELLER <max.schmeller@tier4.jp> Co-authored-by: Kotaro Uetake <60615504+ktro2828@users.noreply.github.com> Co-authored-by: Samrat Thapa <38401989+SamratThapa120@users.noreply.github.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp>
github-actions
bot
added
documentation
Contributor
☂️ Python Coverage
Overall Coverage
New Files
Modified Files
|
Signed-off-by: ktro2828 <kotaro.uetake@tier4.jp>
mojomex
requested changes
Jan 30, 2026
Contributor
mojomex
left a comment
mojomex
left a comment
There was a problem hiding this comment.
Looking good! Some small feedback:
| >>> viewer.render_lidarseg(seconds, lidar_channel, pointcloud) | ||
| ``` | ||
|
|
||
| <!----> |
Contributor
There was a problem hiding this comment.
This image is still commented out. I don't know if we have any internal data that's OK to share as an image. You can probably judge best which data to use.
| self._t4 = t4 | ||
| self._label2id: dict[str, int] = { | ||
| category.name: idx for idx, category in enumerate(self._t4.category) | ||
| category.name: category.index for category in self._t4.category |
Contributor
There was a problem hiding this comment.
This is part of #254 but I had it merged into #256 so it made its way into here 🙇
We can either
- close fix: abstract away category indexing differences #254
- merge fix: abstract away category indexing differences #254 first and update this PR (might be nicer for history but it's very minor)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What
This pull request adds support for rendering LiDAR segmentation point clouds, improves compatibility across different dataset revisions, and refines the handling of category indices. It also updates the documentation and user interface to reflect these new features and changes. The most important changes are grouped below:
LiDAR Segmentation Support:
SEGMENTATIONcolor mode toPointCloudColorModeand support for rendering LiDAR segmentation point clouds in the viewer via the newrender_lidarsegmethod inviewer.py. The rendering helper now automatically chooses segmentation rendering if segmentation data is available. [1] [2] [3] [4] [5] [6] [7] [8] [9] [10]Category Index Compatibility:
Introduced
fix_category_tableinschema/compatibility.pyto ensure category indices are correctly populated, handling cases where some or all indices are missing, and raising an error if the data is inconsistent. This function is now used during dataset initialization. [1] [2] [3]Updated all usages of category indices to use the
indexfield directly, instead of relying on list position, ensuring consistency across different dataset revisions. [1] [2]Documentation and UI Updates:
README.md,docs/index.md, anddocs/tutorials/render.mdto reflect new support for point cloud segmentation and to provide usage examples for rendering LiDAR segmentation. [1] [2] [3]Other Improvements:
Improved map rendering by making it conditional on the existence of the map file, preventing errors if the file is missing.
Updated type hints and imports to support the new segmentation point cloud functionality.