search

RobotLocomotion / LabelFusion

LabelFusion: A Pipeline for Generating Ground Truth Labels for Real RGBD Data of Cluttered Scenes
http://labelfusion.csail.mit.edu
Other
387 stars 98 forks source link

How to run the pipeline at 1280x720 resolution? #91

Closed Tchaikovic closed 4 years ago

Tchaikovic commented 4 years ago

Hi there, I would like to run the entire pipeline so that the final images have 1280x720 resolution. Everything works fine at VGA resolution.

When I acquire a HD sequence though, run_prep goes into seg fault.

Do you know what is required to solve this issue?

peteflorence commented 4 years ago

Yes, ElasticFusion is compiled to run at VGA resolution: https://github.com/mp3guy/ElasticFusion/blob/master/GUI/src/MainController.cpp#L37

Won't be able to help with more follow-up, but wanted to point that out. You can either try recompiling ElasticFusion with your new resolution or flags to dynamically support different resolutions.

On Thu, Jun 11, 2020 at 9:10 AM Tchaikovic notifications@github.com wrote:

Hi there, I would like to run the entire pipeline so that the final images have 1280x720 resolution. Everything works fine at VGA resolution.

When I acquire a HD sequence though, run_prep goes into seg fault.

Do you know what is required to solve this issue?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/RobotLocomotion/LabelFusion/issues/91, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABLBBKGDMJUZWKWSRAJMOPLRWD6VVANCNFSM4N3RVGVA .

tiexuedanxin commented 4 years ago

Hi there, I would like to run the entire pipeline so that the final images have 1280x720 resolution. Everything works fine at VGA resolution.

When I acquire a HD sequence though, run_prep goes into seg fault.

Do you know what is required to solve this issue?

hello, do you solve the problem? could you tell me which camera you use and Can you camera collect data in a real time.

Tchaikovic commented 4 years ago

Hi @tiexuedanxin . Yes, I managed to solve the problem. I'm using both Primesense and Realsense cameras. Follow the instructions here to collect the data https://github.com/ianre657/LabelFusion.

yangxingjian commented 2 years ago

Hello @Tchaikovic, do you happen to remember how you solved the problem?

Tchaikovic commented 2 years ago

Hi @YangXJ95 , I don't remember the details as it was quite some time ago. I can tell you that the resolution is hardcoded across 2-3 files in the repository. I just changed those values consistently and re-compiled. Hope this helps.

yangxingjian commented 2 years ago

Thank you @Tchaikovic, this definitely helps! Also as it's hardcoded in the repository, then I am wondering do I need to change anything in the ElasticFusion?