• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 366 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.

IWR6843AOP: Object Identification and Visualization

Jenn-Bing Ong
Prodigy 75 points
Part Number: IWR6843AOP
Other Parts Discussed in Thread: IWRL6432

Hi TI,

Thanks for the great work !

Regarding the picture shown below, I have 2 questions:
(1) Inside the plot, the subjects and moving clutter are shown as ellipses with different colors and shape / size, how to obtain the parameters of the ellipses from radar output and what are the physical meaning of these parameters ?
(2) How to identify the moving clutter from the human subjects ? 

Another question is how to differentiate humans from other moving objects such as robots and vehicles as mentioned in this video: https://www.youtube.com/watch?v=5B07EwvphUw, what are the useful parameters that can be derived from TI mmwave radar output for object identification ?


Reference: https://www.youtube.com/watch?v=VS_bE61gC3s 

  • +1
    TI__Mastermind 20190 points

    Hello.

    You can look at the tracking command parameters in the configuration file to modify the shape and size of the tracks.  Commands such as the gatingParams and allocationParams control when tracks are assigned based on how big the detected object is and the quality of the points.  You can refer to the tracking guide linked here to understand the various commands and their parameters.

    For (2) and your additional question, you can refer to the Human vs. Non-Human classification demo on the toolbox.  You can also refer to section 4 of this document for implementation details.

    Sincerely,

    Santosh

  • 0 in reply to Santosh Krishnan
    Prodigy 75 points

    Hi Santosh,

    Thanks for the information.

    Regarding the Human vs. Non-Human classification demo in the Radar toolbox, the embedded software runs on the IWRL6432 chip, can it be run on the IWR6843AOP ? If yes, would the change in mmWave chip affects the performance of the classification model ?

    Regarding the plot below, how to draw the ellipses ? Is it using the point cloud data or other information derived from the mmWave radar data ?

  • 0 in reply to Jenn-Bing Ong
    TI__Mastermind 20190 points
    Jenn-Bing Ong said:
    Regarding the Human vs. Non-Human classification demo in the Radar toolbox, the embedded software runs on the IWRL6432 chip, can it be run on the IWR6843AOP ? If yes, would the change in mmWave chip affects the performance of the classification model ?

    Currently, human vs non-human classification runs only the IWRL6432 device.

    Jenn-Bing Ong said:
    Regarding the plot below, how to draw the ellipses ? Is it using the point cloud data or other information derived from the mmWave radar data ?

    Tracks are drawn based on the parameters configured by the tracker layer configuration commands.  It uses a combination of information, which can be found in the link I mentioned in my previous response.  It uses the point cloud data along with information such as the quality of points(SNR) and gating parameters.

    Sincerely,

    Santosh