IWR6843ISK-ODS: [ITB4.2.0 - 3D people counting] What is the calibration just after 1D fft for?

Hi Fields,

We do something similar in the OOB demo as well. When the device is run, there is always bleed over from the Tx directly to the Rx. This causes the first few range bins to always have massive amounts of energy, which can result in false detection if it isn't dealt with. We usually call this dc signature. This is removed with static clutter removal - however, in the 3D People Counting lab, after running the dynamic Capon based chain, we run a second more traditional chain on the static data only. The dc signature would appear in this secondary chain - however, and easy technique to remove this is to capture data from the first few frames at startup then remove the average value of the energy in the first few bins every frame after in post processing.

While this isn't very accessible in the 3D People Counting demo, you can tinker with it in the OOB demo using the "calibDcRangeSig" command. This can even be modified live in the online visualizer. There is documentation on usage of the cli command in the SDK user's guide.

Let me know if you have more questions.

Regards,

Justin