Attention: Answers.ros.org is deprecated as of August the 11th, 2023. Please visit robotics.stackexchange.com to ask a new question. This site will remain online in read-only mode during the transition and into the foreseeable future. Selected questions and answers have been migrated, and redirects have been put in place to direct users to the corresponding questions on Robotics Stack Exchange. Additional details are available here.

How the objects detected by vision are used in Autoware.AI

asked 2020-07-18 21:19:02 -0500

Ahmet Soyyigit
1 ●2 ●3 ●3

Hello,

I have been investigating the Autoware,AI perception source. As far as I understand, the objects detected by vision nodes (ssd, darknet) does not affect the the result of naive_motion_predict node. They are only used by range_vision_fusion and vision_beyond_track nodes, but output of these nodes does not affect any other node. I couldn't find where the bounding box information is used in the whole stack besides visualization. Is it the case that objects detected by vision have no use in the current status of Autoware.AI? Are they only used to visualize bounding boxes?

edit retag flag offensive close merge delete

add a comment

1 Answer

Sort by » oldest newest most voted

answered 2020-08-13 12:05:27 -0500

Josh Whitley

1766 ●4 ●72 ●24

The objects created by range_vision_fusion are published (by default) to the topic /detection/fusion_tools/objects. This topic is subscribed to by the lidar_imm_ukf_pda_track module which is used for object tracking. This module then publishes /detection/objects which is then used by naive_motion_predict.

edit flag offensive delete link

add a comment

Question Tools

1 follower

Stats

Asked: 2020-07-18 21:19:02 -0500

Seen: 621 times

Last updated: Aug 13 '20

How the objects detected by vision are used in Autoware.AI edit

1 Answer

Question Tools

Stats

Related questions

How the objects detected by vision are used in Autoware.AI