-
Notifications
You must be signed in to change notification settings - Fork 101
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Skimap with synthetic data #34
Comments
I think it is wrong to compute Current POse in that way. You need to multiply 4x4 matrices in order to obtain the correct position, because in your difference (position - init_pose) you are not taking into account the Roto-Traslation component. |
Thank you. It looks like I've nailed it with your help. |
@CaptainTrunky I find you have solved this skimap , I have tried to achieve it , but i failed , Can you help me ? i don't know how to change parameters in the launch file . |
Could you provide more details? What's exactly wrong when you're running skimap? |
this is my ros_rgbd.cc .
this is my skimap_live.launch
so i think my parameters is bad. |
First of all, thanks for the awesome tool.
I'm trying to build a map of a synthetic scene using SkiMap. To do so, I convert my own data into bag file and replay it. Resulting point clouds look decent to me:
However, whole map gets filled with noisy obstacles quite fast:
The black thing in front of the agent should be a few voxels thick, but it's drifting a lot.
Definitely, there is something fishy about agent's poses, which I have to convert from my own environment to /tf. I use simplified version of a transformation tree that includes just map, odom, base_link and optical_frame (see figures above). I fix agent's pose at the very first frame and during map reconstruction I store differences between initial state and a current one:
Could you provide any insights about what could I try?
The text was updated successfully, but these errors were encountered: