This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

IWR6843AOPEVM: The relationship among point cloud tlv, target tlv, and target index tlv

Part Number: IWR6843AOPEVM

Hello,

In UART output data format

, there's description of target index tlv like this:

Qestion 1:

If I get it clear, it means we can only know which cluster does one point belong to in the next frame, am I right?

Qestion 2:

Does the cluster or target come from algorithm based on the point cloud tlv of the current frame (the same as the target tlv) or the former frame?

Thank you.

Best regards

Woody

  • Hi Woody,

    Qestion 1:

    If I get it clear, it means we can only know which cluster does one point belong to in the next frame, am I right?

    Points are clustered in the same frame in which they are identified. Each new frame will bring a set of points that will be clustered.

    Qestion 2:

    Does the cluster or target come from algorithm based on the point cloud tlv of the current frame (the same as the target tlv) or the former frame?

    The clusters/targets come from the same frame as the point cloud. Each frame produces a point cloud, which will create clusters and tracks for that same frame.

  • Hi Nate,

    Thank you.

    Now I'm clear about Qestion 2.

    For Qestion 1:

    I think it is this sentences 'NOTE: here the number of points are for frame n-1' and 'A targetID at index I is the target to which point I of the previous frame's point cloud was associated' that confused me. Does the target index tlv also indicate the relationship of point cloud and target in the same frame? What does it mean by 'frame n-1' and 'previous frame'?

    Plus: based on my experiment, the number of target indices in the current frame is always equal to 3 + the number of points in the previous frame. (Except for the first frame in which the number of target indices is equal to the number of points in the same frame.) Why is there 3 extra indices?

    If we've got the target from the point cloud in the same frame, it doesn't make sense not to include the current correspondence simultaneously. That's weird, isn't it?

    Thank you :-)

    Best Regards

    Woody

  • Hi Woody,

    It appears you are correct and I was incorrect before. The target index corresponds to the points from the previous frame. I do not know why the number of target indices in the current frame is 3 more than you're expecting. Perhaps you can generate a point cloud with very few points to check point by point why you're seeing more points than you expect? I think you could set the CFAR parameters so high that nothing gets picked up as a point.

    Best,

    Nate

  • Hi Nate,

    Thank you.

    Let's make it clear again:

    1. The target tlv is got from the current point cloud tlv.

    2. The target index tlv indicate the relationship between target tlv and point cloud tlv in the previous frame.

    Right?

    Best regards

    Woody

  • Hi Woody,

    Yes that's correct. This is because it takes more processing to associate each incoming point to a target, so in each TLV you get the points of the current frame, and then the target ID for the point's in the previous frame.

    Best,

    Nate

  • Hi Nate,

    Thanks a lot :-)

    Best regards

    Woody