110 - Implement traffic light detection #148
Merged
Conversation
This file contains 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
…b.com/UNA-AuxMe/paf23 into 110-implement-traffic-light-detection
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
Simulation results
|
JuliusMiller
approved these changes
Jan 12, 2024
|
Fix failed checks! |
|
Linter error fixed (came from main-branch). |
JuliusMiller
pushed a commit
that referenced
this pull request
Jan 14, 2024
* Add publisher * Publish segmented traffic lights * Implement TrafficLightNode * Add TrafficLightState msg. WIP * Add publisher * Publish segmented traffic lights * Implement TrafficLightNode * Add TrafficLightState msg. WIP * Added side view classification * Finish traffic light node * Add manual control launch file * Make linter happy * Add documentation * Add missing traffic light detection model * Fix color issues in rviz * Limit simulator's max. RAM usage to prevent system crash * fix: Linter fixes for other team's code
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Outdated suggestions cannot be applied.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.
Description
Extended the VisionNode to publish the cut out traffic lights and implemented a new traffic light detection node which uses them to recognize the traffic light state via a small neural network.
The node can publish the states UNKNOWN, GREEN, YELLOW, RED which are int8 enums as a ROS message.
The neural network has been extended to recognize side-facing traffic lights as a first measure of filtering those out. Additionally, a bounding box was laid over the source image, as the correct traffic light is towards the center of the image.
It was necessary to switch from 'yolov8x-seg' to yolo 'rtdetr-x' as the traffic lights were detected too late (had to be almost in the intersection to see them properly). Lack of segmentation wasn't an issue, cutting out by the bounding box of the recognized object was perfectly fine in this case.
Fixes #110
Type of change
Please delete options that are not relevant.
Does this PR introduce a breaking change?
No
Most important changes
Checklist: