laser_filter.yaml limits in linorobot2 are incorrect for ld19 lidar #59
Replies: 12 comments 3 replies
-
|
The limit should be -3.14 and +3.14 for 360 degrees scan. I updated the wiki. https://github.com/hippo5329/linorobot2_hardware/wiki#laser-filter |
Beta Was this translation helpful? Give feedback.
-
|
Thank you for your quick reply. I don't think it's correct for the ld19. (I don't have experience with laser sensors - first time trying - so please excuse any mistakes I make.) Here's the beginning of the data on /scan/unfiltered:
Notice the angle_min is 0.0, so that is the minimum angle that can be filtered. So setting lower_angle to -3.14 does nothing. Look at line 72 on in angular_bounds_filter.h in the laser_filters package - it compares lower_angle to start_angle and throws away anything less than lower_angle, and includes anything higher, so negative lower_angle values have no effect. Here's /scan/unfiltered - you see 360 degrees: Here's /scan with lower set to -3.14 and upper set to 3.14. You only see 180 degrees, because the lower_value has no effect - everything from 0 degrees on is included but nothing above 180 degrees. Here's /scan with lower set to 0 and upper set to 3.14. It is the same as when lower was -3.14 - only shows 180 degrees. The negative lower_angle has no effect. Do you think it's possible the problem with ignoring negative lower_angle might be deeper - maybe the angular range coming out of the laser driver is wrong? I read that the ld19 is considered incompatible with ROS. It's clearly not completely incompatible, but maybe whatever the driver is doing isn't quite right? |
Beta Was this translation helpful? Give feedback.
-
|
I think the angle bound should be depended on angle_min and angle_max. So 0 and 6.28 should be used for ld19. |
Beta Was this translation helpful? Give feedback.
-
|
I think it will be easier to revert to square filter. I will update the filter later. |
Beta Was this translation helpful? Give feedback.
-
|
Hi Thomas, thanks for your reply. The real question in my mind is, is the laser driver working incorrectly on ld19? Where does that angle_min come from? I wonder if there could be other side effects of angle_min being 0. Presumably ld19 is the only device that reports angle_min as 0, or other sensors would have the same problem with angle bounds. Do you have an ld19? I think the angle filter is not bad - but it doesn't seem to be able to handle two obstructions. |
Beta Was this translation helpful? Give feedback.
-
|
I'm experiencing the same issue with the LD19. Just our of curiosity, is it possible to disable the filtering altogether? |
Beta Was this translation helpful? Give feedback.
-
|
ldrobot ld19 scan angle: 0 to 360 degrees. The scan message def said only, "with zero angle being forward along the x axis". All of them are correct. But ldlidar is different from others. https://docs.ros2.org/foxy/api/sensor_msgs/msg/LaserScan.html |
Beta Was this translation helpful? Give feedback.
-
|
@PaulBouchier you may chain/cascade filters as you want. |
Beta Was this translation helpful? Give feedback.
-
|
I change the laser filter to box filter with default all pass. I am traveling and do not have robot/lidar to test. Please verify and let me know. Thanks. |
Beta Was this translation helpful? Give feedback.
-
|
https://github.com/hippo5329/linorobot2_hardware/wiki#laser-filter |
Beta Was this translation helpful? Give feedback.
-
|
I will try to update ld19 driver to change the scan range to -180 to 180 degrees. |
Beta Was this translation helpful? Give feedback.
-
|
Hi Thomas - thank you for always being willing to try and make things better. I honestly don't think it's necessary to change the ld19 driver scan range to -180 to 180 degrees, unless it breaks something else in ROS that I don't know about. The problem with scan angle range can easily be solved with documentation. There are dangers in modifying someone else's package. I leave it to your judgement as to whether modifying LDLidar's package is necessary. (I see you already modified it to support the UDP client). |
Beta Was this translation helpful? Give feedback.
-
|
Hello, I'm using Jazzy for the first time along with your Linorobot2 changes. The lidar scan is now a full 360 degrees (thank you!), however when I run the bringup file, I get a "base_ laser" frame does not exist error message. Any suggestions would be most appreciated. See here: tb@tb-750XDA:~$ ros2 launch linorobot2_bringup bringup.launch.py micro_ros_transport:=udp4 micro_ros_port:=8888 lidar_transport:=udp_server lidar_server_port:=8889 [INFO] [launch]: All log files can be found below /home/tb/.ros/log/2025-02-25-11-55-26-243213-tb-750XDA-2351 |
Beta Was this translation helpful? Give feedback.
-
|
I reverted again to angle bound with default all pass. Please pull and check again. |
Beta Was this translation helpful? Give feedback.
-
|
Hello Again, Here are the results after you reverted to angle bound: tb@tb-750XDA:~$ ros2 launch linorobot2_bringup bringup.launch.py micro_ros_transport:=udp4 micro_ros_port:=8888 lidar_transport:=udp_server lidar_server_port:=8889 [INFO] [launch]: All log files can be found below /home/tb/.ros/log/2025-02-25-19-05-18-078138-tb-750XDA-6551 I shared the above with ChatGPT. It answered with the following: It looks like your ROS 2 launch process started successfully, but there are a few warnings and errors to note: Next Steps: |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
linorobot2_bringup/config/laser_filter.yaml appears based on experimentation to require the lower_angle and upper_angle to be in the range 0 - 2*pi for an ld19 lidar. I was only seeing 90 degrees of scan data (about 105 data items) on /scan when running
ros2 launch linorobot2_bringup bringup.launch.py micro_ros_transport:=udp4 micro_ros_port:=8888 lidar_transport:=udp_server lidar_server_port:=8889Yet on the /scan/unfiltered topic I saw the full 360-degree scan that matched my room.
lower_angle was -1.57 and upper_angle was 1.57 as in the current repo. After some experimentation I found that if I set lower_angle to 0.0 and upper_angle to 6.1 I got the full scan on /scan when viewed with rviz.
I have your latest code pulled in linorobot2:
bouchier@xps15:linorobot2$ git remote -v origin https://github.com/hippo5329/linorobot2 (fetch) origin https://github.com/hippo5329/linorobot2 (push) bouchier@xps15:linorobot2$ pwd /home/bouchier/linorobot2_ws/src/linorobot2 bouchier@xps15:linorobot2$ git branch \* rollingThe link in the Laser Filters section of the wiki page points to wiki.ros.org, which is ROS1 documentation, and actually specifies upper and lower in degrees. The current ros2 laser-filters package source seems to be at
https://github.com/ros-perception/laser_filters/blob/ros2/examples/angular_filter_example.yamlbut there's no documentation there, and the examples angular_filter_example.yaml has -1.57 and +1.57 as lower and upper bounds.What is your finding for the correct values to use for upper and lower angular bounds?
TIA!
Beta Was this translation helpful? Give feedback.
All reactions