ibd's profile - activity

Hi all

I am using a .bag file with RTAB-map ROS. It uses some static transforms between different camera frames to do the mapping.

Edit: An example bag file that shows the buggy behavior can be downloaded here: https://www.dropbox.com/s/6ox0xq6ctg2...

The launch file to launch RTAB-Map can be found here: https://pastebin.com/ZhXb5r5U

When I feed "data.bag" into RTAB-map, it works fine. Now, I create a dummy filtered rosbag, using:

rosbag filter data.bag data_filtered.bag "topic != '/thistopicdoesntexist'"

As this topic doesn't exist, it should create an identical .bag file. However, when I try to feed "data_filtered.bag" into RTAB-map, I get many errors like:

[ WARN] [1494031985.023853640, 1493939596.985478852]: Could not get transform from camera_depth to camera_color after 0.100000 seconds (for stamp=1493939594.615704)! Error=". canTransform returned after 0.10093 timeout was 0.1.".

1. It seems that the filtered bag somehow is missing data from the static TF transforms. When I "rostopic echo -b data_filtered.bag", however, it shows that the static transforms are there.
2. It seems that the dummy filtering actually alters the .bag. Both MD5 hash and file size are changed between "data.bag" and "data_filtered.bag"! No compression is used. Why is that?

Maybe the problem is related to the time stamp? In any case, I am using use_sim_time=true, and I am also using --clock when playing back the .bag files.

Any help would be appreciated.

Cheers,

• ibd

I just did some tests, and it seems that the tango tablet clock will drift as much as 1 second in 30 minutes when not co

I just did some tests, and it seems that the tango tablet clock will drift as much as 1 second in 30 minutes!

Thank you for the workaround. At least these transforms are constant across all rosbags. I try to synchronize the time b

ahendrix, it seems related indeed. Is there a clean fix for the .bag files I already have? Maybe changing the timestamp

Updated the original post with a bag file and the launch file I use to launch RTAB-map.

Bug? Rosbag filter and /tf_static Hi all I am using a .bag file with RTAB-map ROS. It uses some static transforms betwe

I see, thank you for suggestions on how to solve it. Is there a reason that these restrictions are in place?

Hi all

I am playing back some rosbag file. In RVIZ, I see the error message "No transform from [area_description] to frame [start_of_service]". TF view_frames agrees and shows the following:

However, when I export the rosbag file, /tf topic only, to .csv format, I can clearly see that there are 3 kinds of transformations in the rosbag file: 1. map -> start_of_service 2. start_of_service -> device 3. area_description -> start_of_service Whereas the left side is the value in "field.transforms0.header.frame_id", and the right side is the value in "field.transforms0.child_frame_id" in the resulting .csv file.

Now, the resulting graph is acyclic. If we choose start_of_service as root, this is clearly a valid tree. The question remains: Is ROS able to invert the transformations so to form a valid TF tree?

Apologies if this has been asked before. I was unable to find an answer on google or here.

Cheers! - ibd

Bug? Rosbag filter and /tf_static Hi all I am using a .bag file with RTAB-map ROS. It uses some static transforms betwe

ROS TF claims tree disconnected, but clearly connected Hi all I am playing back some rosbag file. In RVIZ, I see the er

ROS TF claims tree disconnected, but clearly connected Hi all I am playing back some rosbag file. In RVIZ, I see the er

I did indeed install everything using apt-get. Posted my questions there.

If I "fix" the file in question, "dashboard.py", by changing line 8 to:

from PyQt4.QtGui import QMessageBox, QAction

Then python will segfault. Here's a backtrace from gdb:

    #0  0x00007fff9645e5a4 in ?? () from /usr/lib/x86_64-linux-gnu/libQtGui.so.4
#1  0x00007ffff7de74ea in call_init (l=<optimized out>, argc=argc@entry=4, 
    argv=argv@entry=0x7fffffffd858, env=env@entry=0xeb2920) at dl-init.c:72
#2  0x00007ffff7de75fb in call_init (env=0xeb2920, argv=0x7fffffffd858, 
    argc=4, l=<optimized out>) at dl-init.c:30
#3  _dl_init (main_map=main_map@entry=0x2deef60, argc=4, argv=0x7fffffffd858, 
    env=0xeb2920) at dl-init.c:120
#4  0x00007ffff7dec712 in dl_open_worker (a=a@entry=0x7fffffffb310)
    at dl-open.c:575
#5  0x00007ffff7de7394 in _dl_catch_error (
    objname=objname@entry=0x7fffffffb300, 
    errstring=errstring@entry=0x7fffffffb308, 
    mallocedp=mallocedp@entry=0x7fffffffb2ff, 
    operate=operate@entry=0x7ffff7dec300 <dl_open_worker>, 
    args=args@entry=0x7fffffffb310) at dl-error.c:187
#6  0x00007ffff7debbd9 in _dl_open (
    file=0x2b49180 "/usr/lib/python2.7/dist-packages/PyQt4/QtGui.so", 
    mode=-2147483646, caller_dlopen=0x5231c3 <_PyImport_GetDynLoadFunc+243>, 
    nsid=-2, argc=<optimized out>, argv=<optimized out>, env=0xeb2920)
    at dl-open.c:660
#7  0x00007ffff75edf09 in dlopen_doit (a=a@entry=0x7fffffffb540) at dlopen.c:66
#8  0x00007ffff7de7394 in _dl_catch_error (objname=0xa6c770, 
    errstring=0xa6c778, mallocedp=0xa6c768, 
    operate=0x7ffff75edeb0 <dlopen_doit>, args=0x7fffffffb540)
    at dl-error.c:187
#9  0x00007ffff75ee571 in _dlerror_run (
    operate=operate@entry=0x7ffff75edeb0 <dlopen_doit>, 
    args=args@entry=0x7fffffffb540) at dlerror.c:163
#10 0x00007ffff75edfa1 in __dlopen (file=<optimized out>, mode=<optimized out>)
    at dlopen.c:87
#11 0x00000000005231c3 in _PyImport_GetDynLoadFunc ()
#12 0x0000000000522d6f in _PyImport_LoadDynamicModule ()
#13 0x00000000004afbad in ?? ()
#14 0x00000000004af34d in PyImport_ImportModuleLevel ()
#15 0x00000000004b0f78 in ?? ()
#16 0x00000000004b0cb3 in PyObject_Call ()
#17 0x00000000004ce5d0 in PyEval_CallObjectWithKeywords ()
#18 0x00000000004c6ed6 in PyEval_EvalFrameEx ()
#19 0x00000000004c2765 in PyEval_EvalCodeEx ()
#20 0x00000000004c2509 in PyEval_EvalCode ()
#21 0x00000000004c061b in PyImport_ExecCodeModuleEx ()
#22 0x00000000004bd6ee in ?? ()
#23 0x00000000004afbad in ?? ()
#24 0x00000000004af34d in PyImport_ImportModuleLevel ()
#25 0x00000000004b0f78 in ?? ()
#26 0x00000000004cada2 in PyEval_EvalFrameEx ()
#27 0x00000000004c9d8f in PyEval_EvalFrameEx ()
#28 0x00000000004c9d8f in PyEval_EvalFrameEx ()
---Type <return> to continue, or q <return> to quit---
#29 0x00000000004c9d8f in PyEval_EvalFrameEx ()
#30 0x00000000004c9d8f in PyEval_EvalFrameEx ()
#31 0x00000000004c9d8f in PyEval_EvalFrameEx ()
#32 0x00000000004c9d8f in PyEval_EvalFrameEx ()
#33 0x00000000004c2765 in PyEval_EvalCodeEx ()
#34 0x00000000004ca099 in PyEval_EvalFrameEx ()
#35 0x00000000004c2765 in PyEval_EvalCodeEx ()
#36 0x00000000004ca099 in PyEval_EvalFrameEx ()
#37 0x00000000004c2765 in PyEval_EvalCodeEx ()
#38 0x00000000004ca099 in PyEval_EvalFrameEx ()
#39 0x00000000004c2765 in PyEval_EvalCodeEx ()
#40 0x00000000004ca099 in PyEval_EvalFrameEx ()
#41 0x00000000004c2765 in PyEval_EvalCodeEx ()
#42 0x00000000004ca099 in PyEval_EvalFrameEx ()
#43 0x00000000004c2765 in PyEval_EvalCodeEx ()
#44 0x00000000004ca099 in PyEval_EvalFrameEx ()
#45 0x00000000004c2765 in PyEval_EvalCodeEx ()
#46 0x00000000004c2509 in PyEval_EvalCode ()
#47 0x00000000004f1def in ?? ()
#48 0x00000000004ec652 in PyRun_FileExFlags ()
#49 0x00000000004eae31 in PyRun_SimpleFileExFlags ()
#50 0x000000000049e14a in Py_Main ()
#51 0x00007ffff7811830 in __libc_start_main (main=0x49dab0 <main>, argc=4, 
    argv=0x7fffffffd858, init=<optimized out>, fini=<optimized out>, 
    rtld_fini=<optimized out>, stack_end=0x7fffffffd848)
    at ../csu/libc-start.c:291
#52 0x000000000049d9d9 in _start ()

Any ideas? I concluded that it must be PyQt4.QtGui, and not PyQt5.QtGui, because the latter doesn't contain QMessageBox.