JustinBlack02's profile - activity

A PointCloud2 is simply a raw memory buffer, to interpret it, you need to use a PointCloud2Iterator which can be found h

I think your remap tag is backwards, try <remap from="/robot_2/scan" to="/scan"/>

Yes it is possible, however the tricky part is selecting the correct process when you are attaching your debugger. As a

Yes it is possible, however the tricky part is selecting the correct process when you are attaching your debugger. As a

Is this ROS1 or ROS2? Have you tried putting the remapping inside the node tag?

which is the proper syntax to do that ? void left_ecn_cb(const std_msgs::msg::UInt16::ConstPtr& ltick_msg) {

Yes it is correct.

@gvdhoorn, interesting, I'm not getting the same results even after pulling the latest docker image.

You should post the actual warning itself. i.e. a copy of your console

For your robot to accurately control itself and follow the path it will invariably need to do localization or have a dec

Please mark this answer as accepted or add your own answer for future users.

Your entire problem is self inflicted. You localize in 3D and navigate in 2D. Why don't you make your dimensions match?

This will be a lot of work so make sure it is worth it. Import your point cloud into Meshlab or CloudCompare and gene

You should use pbstream_map_publisher and/or occupancy_grid_node to create the occupancy grid. If you are running pure l

In theory it should be possible, if you assume your world is basically 2D then you could project on the xy plane your ra

Define what you mean by "map". For a lot of people, having a point cloud representation of the environment would be a ma

I don't think there is a specific way to do that using the existing sync policies. To handle your case you would have to

It's impossible to clearly answer your question, you will have to figure out how precise of a solution you want on your

I don't think there is a specific way to do that using the existing sync policies. To handle your case you would have to

I hadn't thought about that. The option @gvdhoorn suggest is probably the best one.

This can be caused by a variety of reasons which only you can really debug. I would suggest trying the following: rospa

While true, on the driver side you don't know when the application will start listening on the messages so you should ju

Pixel disparity is unrelated to rtabmap. If you want to follow that formula then you will need to retrieve the intrinsic

I don't see the problem here. If you want to avoid obstacles, then by definition you are moving. If you are moving then

For those wondering the same thing, sometimes IMUs will not give the magnetic field but rather the magnetic vector. If y

I can't find it anymore but I seem to recall reading somewhere that when you record a rosbag, the timestamp encoded to the message is the time at which the message was received by the rosbag record process instead of the time at which it was published.

If this is true, is there a way to force a rosbag to instead encode messages with the timestamp which were published with them?

I don't think anyone other than you can answer this question. You would need to first define what your performance metric is (speed, accuracy, map quality, ...) before trying to find how to compare the two.

I have a camera which gives me images which are already rectified. I'm writing the ROS driver for it and I'm wondering what I should do with the camera_info message. Should I :

• Calibrate the camera and publish the intrinsic parameters in the camera_info message
• Calibrate the camera and publish the projection matrix but set the distortion parameters to 0
• Not calibration and set all the camera_info parameters to 0

For the record, I do not have access to the prerectified image, only the final result.