• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 615 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.

IWR1843BOOST: The elevation value is wrong by 10 degree.

Big Fields
Intellectual 475 points
Part Number: IWR1843BOOST
Other Parts Discussed in Thread: IWR1843, , IWR6843

Hi,

I am doing some experiments using traffic_monitoring lab in mmwave_industrial_toolbox_4_4_0.

What I did:

 1. Run traffic_monitoring lab.

 2. Start walking in front of the radar from 0m to 40m.

 FW: 18xx_68xx_traffic_monitoring\prebuilt_binaries\traffic_monitoring_18xx_demo.bin

config: 18xx_68xx_traffic_monitoring\chirp_configs\18xx_traffic_monitoring_70m_MIMO_3D.cfg

Please see the point cloud data attached below. This shows all the point cloud data over the time from this person starts walking until he stops.

As you can see, the elevation value is slightly wrong. I expected it to be 0 degree but actually is is around 10 degree.

Why is the elevation value is wrong by 10 degree?

Best regards,

Fields

  • 0
    TI__Mastermind 30405 points

    Hi Fields,

    1. If you are looking to detect and track people at long range, you should use the Long range people detection and tracking demo. Both Traffic Monitoring and Long Range People Tracking demos use the same processing chain, but the configuration of the Traffic Monitoring Demo is specifically designed and optimized for objects with larger RCS, namely vehicles (cars, buses etc) and it is not suitable for detecting and tracking people. The Configuration of Long range people detection and tracking demo is optimized for detecting and tracking Human targets at up to 100m.

    2. However, you will need to migrate the Long Range People Detection Demo to IWR1843 since it is provided for the IWR6843 device in Industrial Toolbox. But this migration should be very easy if you use the Traffic Monitoring Demo as reference which provides CCS projects for both IWR6843 and IWR1843. You can compare the differences between the two and make similar changes to the Long Range People Detection Demo to migrate it to IWR1843BOOST. For more information, you can also refer to the following migration guide: Migrating to xWR68xx and xWR18xx Millimeter Wave Sensors

    3. Regarding the error in angle measurement, please note that both Long Range People Detection and Traffic Monitoring Demos use 3TX beamforming (and beam-steering) to increase detection range. This reduces the number of virtual receivers (4x1=4 in 3TX SIMO as compared to 4x3=12 in 3TX TDM MIMO) and hence the angular resolution, which in turn reduces the angular accuracy as well. Note that the configurations for these demos are targeted at macro motion detection and tracking and not achieving the best range and angle accuracy. You can check/confirm the angular accuracy of your IWR1843BOOST EVM by using the SDK Out of Box demo.

    Hope the above helps.

    Regards
    -Nitin

  • 0 in reply to Nitin Sakhuja
    Intellectual 475 points

    Hi Nitin,

    Thank you for your answer, but this still doesn't resolve my question.

    Please see the pictures below.

    On the left side is the point cloud data and on the right side is a graph that shows the number of points that fall in each 0.5 elevation degree range.

    As you can see, the average of elevation angle is obviously wrong. It must be around 0 degree but it is actually around 9~10 degree.

    If this is due to the angle resolution, the average is still around 0 degree, isn't it?

    Where does this "10 degree" come from?

    Best regards,

    Fields

  • 0 in reply to Big Fields
    TI__Mastermind 30405 points

    Hi Fields,

    Could you run the Out of Box demo first to confirm the accuracy of the setup as suggested in my last response?

    Regards

    -Nitin