• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 1362 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

IWR1642BOOST: Detection of a wall

Bob Bacon
Genius 10815 points
Part Number: IWR1642BOOST

Hello
This is releted to E2E post

The detection demo seems to detect objects (Humans, cars etc..) but does not detect a wall. In the below graph there's a solid wall running parallel to the detector 2m to the left of the sensor centre line. I am expecting a solid (or at least a row of dots) to represent the wall. It does not detect this wall (I’m sure it’s actually detecting it, but not classifying it as a wall) but it does detect individual point objects. What am I missing ?

We need this knowledge of a wall - i.e. the output from the primary RANGE FFT in the DSS application pipeline, prior to the secondary Doppler FFT. I cannot find a hook to access this data - where can I find such a hook in the API? Is this the correct approach ?

For information, I have copied the demo program for the 1642 from: ~/ti/mmwave_sdk_01_00_00_05/packages/ti/demo/xwr16xx/mmw. My plan is to modify this demo to include our algorithm and CAN interface, compile and download to the eval board and demonstrate this as a concept to management.

Regards

Bob Bacon

[unlocked 1/19/18]>LE

  • 0
    TI__Mastermind 24955 points

    Bob,

    We may need some more information about the wall you are trying to detect. What of material is the wall? Is it a traditional drywall style wall? Is it a solid metal wall or is it glass?

    I ask this because we have conducted some experiments and found that the angle of the mmWave senor relative to drywall does affect the ability to detect a wall and objects that may be behind the wall. I will post the link that summarizes the results of our experiments. We found that the best case scenario occurs when the mmWave senor is directly perpendicular to the wall. So the angle at which our device is pointed towards a particular way is important.

    Please let us know if you have any questions. I want to make sure we have eliminated this potential issue before proceeding down a particular debug path.

    Thanks and cheers,

    Kyle

  • 0 in reply to Kyle Cousino
    TI__Mastermind 24955 points
    Bob,

    As a followup, here is the link to the set of experiments that were conducted with different types of wall materials.

    dev.ti.com/.../

    Kyle
  • 0 in reply to Kyle Cousino
    Prodigy 70 points

    Hi Kyle,

    Bob posted my message on my behalf. In my initial tests, the wall is solid brick with plaster covering. The angle of incidence is around 45 Degrees to the wall normal.

    Also, (not on the picture) There's a single target at the far end wall, so it's certainly picking the wall up at 90 degrees incidence, but again, I would expect a series of targets describing the outline of the wall, not a single target.

    Thanks, Richard.

  • 0 in reply to Richard Gears
    Prodigy 70 points
    p.s. the link to the set of experiments on glass, metal etc.. was very useful.
  • 0
    TI__Guru 71526 points
    unlocked 1/19/18[LE]