P0 - Convert CV Lane Detections to Lidar Topic - Githubissues

UTRA-ART / Caffeine

Caffeine on ROS for IGVC

18 stars 2 forks source link

P0 - Convert CV Lane Detections to Lidar Topic #133

Closed Raindancer0912 closed 10 months ago

Raindancer0912 commented 1 year ago

Description:

Transform the lane detection output (points in camera frame) to a scan message to use for costmap and SLAM

Goals:

[ ] Construct Pointcloud from CV output
[x] Transform from camera frame to laser frame
[x] Compute distances for each angle in scan message and publish
[ ] Possibly interpolate angles in between

Acceptance Criteria:

[x] Publish to a new scan topic in same frame as current lidar
[ ] Topic data should represent the position of the lanes
[ ] Feed the topic into costmap

Raindancer0912 commented 10 months ago

This task has been completed on branch IGVC-2023 instead. Implementation includes transforming the image segmentation into a lidar and PC topic and having SLAM package cartographer subscribe to it. Lane markings show up on map and costmap.