Empty output with own data

[ghost]

Hi, firstly, thanks for sharing your work!

After running your algorithm with my recorded data, it seems not to give any output when I play the bag file.
I just used horizontal 2D laser data, without IMU, I guess that's fine. Could it be because I don't provide intensity values?

When I run the launch file, a fatal error arises. I am not sure if this is the problem because when I use Cartographer (sample) data it also appears this error, but the algorithm works well (new topics are correctly published and rviz visualization looks good).

I can't find any bug in my input data. Do you know why this problem could be happening?

[libing64]

@polbolso You should modify the path to your own rosbag in launch file.

[ghost]

Thanks! I have included the path to my bag file and the error opening the file disappeared. But now there is another error that prevents me from showing anything in rviz.

Is it something related to my input data? As I see something about the topic 'horizontal_laser_2d' mentioned in the error.

[libing64]

@polbolso Please print the result of
"rostopic list", "rostopic info horizontal_laser_2d" and "rostopic info your_own_laser_topic"

[ghost]

I saved my 2D LiDAR raw data into a bag file, under the topic 'horizontal_laser_2d', so I guess in this case this is the same as 'my_own_laser_topic".

After playing my input bag file, this is what the recorded bag file shows (I exited after 50 s), only the initial pose is obtained as output:

[libing64]

@polbolso Can you share your rosbag? And I would try to debug with your data.

[ghost]

20210330_132458.zip

Thanks!

[ghost]

Hi @libing64, have you already tried to debug it with my data?

[libing64]

@polbolso Yeah, it crashed with your data and I am trying to figure out why.

[libing64]

@polbolso The header of laser scan has invalid data, for example, angle_min and angle_max.

header: 
  seq: 20
  stamp: 
    secs: 46833
    nsecs: 118433999
  frame_id: "horizontal_laser_link"
angle_min: 5.454115390777588
angle_max: 5.8087310791015625
angle_increment: 0.0015219557099044323
time_increment: 9.699571137389285e-07
scan_time: 0.00022600000374950469
range_min: 0.4160001873970032
range_max: 45.38550567626953

[ghost]

@libing64 thanks for looking into the data.
But why is it invalid data? Do you mean because the angle_min and angle_max are not between [-180° : 180°]?
These values are between [0° : 360°] (expressed in radians), as my laser scanner scans 360° starting from the origin.
If that is the problem, I could try to generate a new bag file converting the angle data to fit into [-180° : 180°] and try to run it again.
Does this make sense?

[ghost]

Hi @libing64 , I tried generating a new ros bag file (attached) with the angle_min and angle_max within [-3.14 : 3.14 rad], but I still have the same effect when I run it.
Why did you mean that the header of my laser scan had invalid data? Can you please explain me this?

210330_132458.zip

[libing64]

@polbolso What is your radar fov? 5.4~5.8 rad? 24 degree?

[ghost]

@libing64 My lidar's fov is 360°.
The reason why there are some strange values like 5.4~5.8 rad is that as there are some gaps (angles without points), when I converted the raw data into ros bag, I splitted the data only considering the angles with scanned points. In this way, all the separated parts are continuous and the time increment would make sense.

[libing64]

@polbolso Don't splite the lidar data into different parts, and you could set the invalid points to invalid value such as 0 or -1.

[ghost]

@libing64 I did that, I have now a continuous file, just split per each scan, so in angle intervals of [0 : 6.28 rad].
But again, when running my new bag file, I still can't get any output.

Link to the new bag file: https://drive.google.com/file/d/12sQXlGgxZ0AlBEEgWL94JIFgnH3oAHAn/view?usp=sharing