IWR6843LEVM: How to support clutter removal functionality?

Hi

Thanks for your query. Please allow us a couple of days to respond

Regards

Hi Josh,

  I have a little confused about your reply "it is always enabled in code and no option is provided for it on the CLI for the same reasons".

From 3D_people_tracking_demo_implementation_guide.pdf, it mentions "The static clutter removal is performed by function RADARDEMO_aoaEst2DCaponBF_clutterRemoval(). 

From the source code, this function will do nothing since it's not defined RADARDEMO_AOARADARCUDE_RNGCHIRPANT in 3D people tracking.

Can you please help to check it again?

Thanks.

Hi Josh,

  I can see the doppler value of point cloud is zero in my experiment. Can you please help to explain this result?

Thanks.

Hi, 

Can you provide more details about your experiment? Also please let me know if you have modified the example code or configuration.

Best Regards,

Josh

Hi Josh,

  The experiment is only put mmWave sensor in the desktop and record the point clouds information.

Thanks.

Hi Josh,

  Do you find the reason?

Thanks.

Hello, 

Sorry for the delayed response here.

Are you using the prebuilt binary files and which configuration are you using? 

If you could provide more information about the data captured in your screenshot it would be helpful. For example, how are you capturing and parsing the data from the device? It appears that there are >27000 rows in the data that you are showing, am I correct to assume that all of these are detected points in the point cloud and only ~15 points have a doppler value of 0? Also, since the points seem to have 0 for azimuth and range I am wondering if these are even valid detections or if there is some issue with the parsing/corrupt data/something else. I'll keep looking into this to see if this issue of invalid points is a known issue and get back to you with my findings in the next couple of days.

Best Regards,

Josh

Hi Josh,

  Let's clarify the environment setup in my experiment first. 

Device: IWR6843-ODS

Example: 3D People tracking

Radar firmware: prebuilt binary file 3D_people_track_6843_demo.bin in radar_toolbox_2_00_00_06\source\ti\examples\People_Tracking\3D_People_Tracking\prebuilt_binaries ) 

Radar config:  ODS_6m_default.cfg in radar_toolbox_2_00_00_06\source\ti\examples\People_Tracking\3D_People_Tracking\chirp_configs

Steps:

1. Modify parseTLVs.py in radar_toolbox_2_00_00_06\tools\visualizers\Applications_Visualizer\common to save point clouds information.

   The change is only add save point clouds inforamtion to file in parseCompressedSphericalPointCloudTLV()

2. Run gui_main.py in radar_toolbox_2_00_00_06\tools\visualizers\Applications_Visualizer\Industrial_Visualizer

3. Start mmWave sensor

4. Put the mmWave sensor in desktop and monitor point clouds information.

I do the experiment again in this morning and get the same result. There are 7 points with zero doppler value. 

Can you please help to check it?

I attached the full point clouds information for your reference.

20240925-095316.csv

There are some questions about this detection layer flow chart.

1. Are the flow A and flow B use in this 3D People tracking?

    If yes, is it possible to remove flow B?

-----

2. Is the flow A enable clutter removal by default?

    Is it possible to change the setting?

-----

3. Is the flow B disable clutter removal by default?

    Is it possible to change the setting?

Thanks.

Hi Josh,

  Do you have any findings?

Thanks

Hello, 

Sorry for the delayed response. And thanks for the additional information about your experiment. I'm still trying to check if this is a known issue. 

Regarding the detection layer flowchart questions, the processing in A removes the static clutter, this is not configurable and is required for the Capon algorithm. 

The processing in B is the static scene processing mentioned in section 2.1.7 of the same document, this is optional processing which can be enabled via CLI configuration but it is disabled by most of the default configuration files. The command to enable this is processing as well as the applicable trade offs can be found in section 3.2.4. 

Best Regards,

Josh