Tool/software:
- there is a difference between points and track position.
-
even though the dots are coming, the track start is late.
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.
Tool/software:
Hi,
We can review the configuration and will provide you feedback.
That being said, the performance of the tracker is related to the detection performance.
If the sensor is used in an environment where the detection is not optimal the performance of the tracker will not be optimal
thank you
Hello, Jin Lee,
Yes, this is a borad we designed. The test object is an adult riding a bicycle. You can see from the points in the video that we have no problem with detection. This is our interface but we communicate in exactly the same tlv and points structure. There is a problem here because the track position is not exactly on the points for some reason, I shared information to solve this, only tracker specific. If you want additional information, I can share it again.
we have changes in some commands compared to standard ti commands, but in general it is the same.
{"sensorStop \n\r"},
{"flushCfg \n\r"},
{"dfeDataOutputMode 1 \n\r"},
{"channelCfg 15 7 0 \n\r"},
{"adcCfg 2 1 \n\r"},
{"adcbufCfg -1 0 1 1 1 \n\r"},
{"bpmCfg -1 0 0 1 \n\r"},
{"profileCfg 0 77 6 6 49 0 0 24 1 128 3000 0 0 48\n\r"},
{"chirpCfg 0 0 0 0 0 0 0 3 \n\r"},
{"chirpCfg 1 1 0 0 0 0 0 6 \n\r"},
{"frameCfg 0 1 172 0 200 1 0\n\r"},
{"lowPower 0 0 \n\r"},
{"guiMonitor -1 1 0 0 0 0 0 \n\r"},
{"cfarCfg -1 0 2 8 4 3 0 15 0 \n\r"},
{"cfarCfg -1 1 0 4 2 3 1 10 0 \n\r"},
{"multiObjBeamForming -1 0 0.5 \n\r"},
{"clutterRemoval -1 0 \n\r"},
{"calibDcRangeSig -1 0 -5 8 256 \n\r"},
{"aoaFovCfg -1 -15 0 0 15 -2 1 -0.2 0.2 0 0.3 -0.3 0.3 -90 0\n\r"},
{"cfarFovCfg -1 0 0.25 15 \n\r"},
{"cfarFovCfg -1 1 -8 8 \n\r"},
{"CQRxSatMonitor 0 3 15 125 0 \n\r"},
{"CQSigImgMonitor 0 115 6 \n\r"},
{"analogMonitor 0 0 \n\r"},
{"ransacCfg 70 0.2 \n\r"},
{"lvdsStreamCfg -1 0 0 0 \n\r"},
{"dbscanCfg 4 3 20 20 2 256 0 \n\r"},
{"compRangeBiasAndRxChanPhase 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 \n\r"},
{"measureRangeBiasAndRxChanPhase 0 1. 0.2 \n\r"},
{"calibData 0 0 0 \n\r"},
{"extendedMaxVelocity -1 0\n\r"},
{"sensorPosition 1.0 0 0 \n\r"},
{"staticBoundaryBox -15 0 0.25 4 0.25 3 \n\r"},
{"boundaryBox -15 0 0.25 5 0.25 3 \n\r"},
{"gatingParam 10 4 4 4 10 \n\r"},
{"stateParam 3 3 3 3 3 10 \n\r"},
{"allocationParam 50 50 0 5 4 2 \n\r"},
{"maxAcceleration 10 10 0.1 \n\r"},
{"trackingCfg 1 2 500 20 100 346 200 \n\r"},
{"presenceBoundaryBox -15 0 0.25 5 0.25 3 \n\r"},
{"sensorStart \n\r"}
Thank you.
Hi Emre,
Thanks for the details. Given that you are testing with a person on a bicycle, I might start by looking at the following parameters:
For a detailed description of the gtrack algorithm and parameter tuning, please refer to the Tuning Guide: /cfs-file/__key/communityserver-discussions-components-files/1023/3D_5F00_people_5F00_counting_5F00_detection_5F00_layer_5F00_tuning_5F00_guide.pdf
Thank you,
Jin