IWR6843: static cloud points

Hello Tom,

Please see the info on StaticRACfarCfg in the 3DPeoplecountingDemo_ConfigurationDetails.pdf included with the toolbox download.

C:\ti\mmwave_industrial_toolbox_4_5_1\labs\people_counting\68xx_3D_people_counting\docs

Regards,

Jackson

Hi Jackson,

Thanks for your reply. The doc you mentioned says the RAcfar method is hard coded and couldn’t be configured.

However code in people counting overhead seems to support different cfar methods. We simply changed all parameters of static range cfar to to be same with dynamic range cfar except the last one. But no big difference was observed.

We also tried different set of parameters and failed. Did we miss some thing else?

Cheers,

Tom

Hi Jackson,

We double checked the code and found the cfar method parameter is ignored in the cli and cannot be changed. The pdf doc in pc3d overhead project also recommends a parameter set:

dynamicRACfarCfg -1 4 4 2 4 8 16 4 4 4.00 4.50 0.50 1 1
staticRACfarCfg  -1 4 4 2 4 8 16 4 6 8.00 13.0 0.30 0 0

// ods
dynamicRACfarCfg -1 10 1 1 1 8 8  6 4 4.00 6.00  0.50 1 1
staticRACfarCfg  -1 4  4 2 2 8 16 4 6 6.00 13.00 0.50 0 0

// aop
dynamicRACfarCfg -1 10 1 1 1 8 8  6 4 4.00 6.00  0.50 1 1
staticRACfarCfg  -1 4  4 2 2 8 16 4 6 6.00 13.00 0.50 0 0

Hi Tom,

Both of these CFG files have been tested, but are intended for different antenna configurations.

Can you confirm which version of the SDK and toolbox you are using? And which demo you are trying to run? These values do not seems to be hardcoded in the latest 3D people counting demo.

Also, please see the documentation for staticRangeAngleCfg parameters. This will also control the static point cloud.

Regards,

Jackson

Hi Jackson,

Sorry for not listing the details in previous posts,  which are:

1. SDK 3.5.0

2. ITB 4.5.1

3. people counting overhead

4. AOP es2

We also found the capon method could be different and selected by the number  of parameters passed via cli. It seems a zoom-in based method is activated  according to the CFG file. Do you have any doc about this zoom-in based method?

Cheers

Tom

Hello Tom,

Can you please explain a little bit more about what you want to use the static point cloud for? This will help what parameters are best to change.  These points are not utilized in the tracker so the people counting configurations are not tuned and optimized for static points.

You can also try increasing the maximum allowed static points, depending on your scene this may be limiting the static reporting.

"C:\ti\mmwave_industrial_toolbox_4_5_1\labs\common\src\dpu\capon3d_overhead\radarProcess.h" (for overhead people counting)

#define  MAX_DYNAMIC_CFAR_PNTS       150
#define  MAX_STATIC_CFAR_PNTS        150

Regards,

Jackson

Hi Jackson,

We would like use static points to improve the tracking performance.

As mentioned as in your previous post, only dynamic points are used in the gtrack module. Our idea is to use static points to separate two close targets if they stay quasi-static in the dynamic scene.

Cheers,

Jason

Hi Jaskon,

Yes, it did output more points when lowering thresholds. We'v made trial modifications to this lab project and started testing that.

Thanks for your help again.

Cheers,

Tom