velodyn | Dynamical systems methods for RNA velocity analysis | Genomics library

Looking at the link you shared, I see the OBB object has the following properties: center, extent and R. If you can access them then you can get position and orientation. Center is a point (x,y,z), extent are three lengths in x, y and z direction and R is a rotation matrix. Columns of R are three orthogonal unit-vectors pointing on rotated x, y and z directions.

I think you are interested in orientation, so R is the orientation matrix. You can convert it to quaternion using the matrix-to-quaternion method on this page: https://www.euclideanspace.com/maths/geometry/rotations/conversions/matrixToQuaternion/

Here's a code that produces exactly the same output as the Python version:

After a thorough examination of the network traffic both on macOS and on the emulated Ubuntu using Wireshark on both ends, I realised that there is no Network adapter present in the virtual machine configuration.

Following the answer from Desktop Parallel support, I did: Ubuntu on Mac -> configure -> hardware -> add device -> network 2 -> source: USB...

and when you check now in Wireshark, you will see the traffic coming from the USB adapter and the ROS driver, via an adequate port, will capture the packets.

I found a file which was causing the CMAKE to include 10 sm_arch in compatibility list. Here's the link. I will re-compile after editing the file for just 1 sm_arch and compare the size of binaries generated. – Anuj Patil Jan 22 at 18:10

So findCUDA was the culprit here. Editing the files to required sm_arch does the trick!

There is probably some problems with the routing. You can try debugging the routing issues, some googling will probably help there. Some commands which might help pinpoint the problem tracepath -n 10.0.1.201 and ip route list.

The easiest solution to your problem would be to just configure the "velodyne_interface" you created to the same subnet as your velodyne lidar. So in the "velodyne_interface" set your ip to 10.0.1.20 for example. Connect to the "velodyne_interface" connection, verify that you have the ip you set by typing ip a or ifconfig in some terminal and you should be able to access the Velodyne web interface from 10.0.1.201. From there you can configure the velodyne sensor networking to the settings you like.

As you can see, the points are available but they are not visible long enought because points are not showed simultaneously. The reason is that a single message does not contain all points of a complete (360°) beam. A beam gets splittet to several messages typically, of which only the latest gets shown by default. Checking the rviz point cloud documentation you will find a parameter called decay time:

The amount of time to keep a cloud/scan around before removing it. A value of 0 means to only display the most recent data.

Try increasing the value of this parameter in rviz, then you should be able to see more points simultaneously.

Maybe I'm a bit late, but for anybody having the same Problem:

For question 1.-3. see this. Also what you need to have in mind, is that the data is stored as uint8, but your points should be in float32, if I see it correctly. Therefore each value (x,y,z,intensity, etc.) or "field" is stored as multiple uint8 bytes. So you need 4 data entries to represent the x value of one point. The total length of one point in bytes is stored as "point_step", answering your fourth question.

1. The Field Offset is the number of bytes from the start of the point to the byte, in which this field begins to be stored. So every point has the first 4 bytes for x, then with an offset of 4 start the bytes for y etc.

2. and 3.: fields.datatype and fields.count: See this

3. point.step is number of bytes or data entries for one point

4. row_step: See your own link, so it is "number of points per row * point_step"

5. Probably your scanner publishes line after line? I'm actually not sure of this one.

6. No, the first 4 entries represent the x value, so 235, 171, 190, 53 equals: 11101011 10101011 10111110 00110101 and this represents a float32 value. The 171 has no direct information about the x, y or z value of the point.

For your first query regarding x and y dimension there are two explanation.

Reason 1.

• For image re-projection pin hole camera model is used which is in perspective coordinate or homogenous coordinate. Perspective projection uses the image origin as centre of projection and points are mapped to the plane z=1. A 3D point [x y z] is represented by [xw yw zw w] and the point it maps on the plane is represented by [xw yw zw]. Normalising with w gives.

  So (x,y) -> [x y 1]^T : Homogeneous Image Coordinates

  and (x,y,z) - > [x y z 1] ^T : Homogeneous Scene Coordinates

Reason 2.

• With respect to the paper you have attached, considering equation (4) and (5)

  It is clear that P is of dimension 3X4 and R is expanded to 4x4 dimension.Also x is of dimension 1x4. So as per matrix multiplication rule number of columns of first matrix must equal to the number of rows of second matrix. So for given P of 3x4 and R of 4x4, x has to be 1x4.

Now coming to your second question of LiDAR image fusion, It requires intrinsic and extrinsic parameters (relative rotation and translation) and camera matrix. This rotation and translation forms a 3x4 matrix called as transformation matrix. So the point fusion equations becomes

I was these days also working on the Kitti dataset, so I just added a new function to load a kiti velodyne data file directly into MRPT (see this PR).

However, after some thinking, I noticed that Kitti raw data does not match exactly with CObservationVelodyneScan, which is aimed at storing the raw ranges for each LiDAR beam, and only optionally, a pointcloud. The Kitti velodyne data are pointclouds, actually, hence I added a new PointCloud type with XYZ+Intensity mrpt::maps::CPointsMapXYZI and added a method loadFromKittiVelodyneFile() to it. Notice this is for the mrpt master git branch, "version 1.9.9".

Now, how to insert that into a gridmap? Your idea of using a velodyne CObservation to insert it into a gridmap is one of the pending issues on our queue but, anyway, as said above, the Kitti datasets are better loaded as pointclouds.

I would recommend you converting the pointcloud into a CObservation2DRangeScan, then inserting it into the grid. That would allow you to control what part of the 3D data you really want to be reflected in the grid (i.e. what heights, etc.)

Hope it helped!