• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 33 subscribers
  • Views 615 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

Linux/AWR1443BOOST: Noise near the sensor: even there is nothing before the sensor, there is always some specks, how to remove...

Linux/AWR1443BOOST: Noise near the sensor: even there is nothing before the sensor, there is always some specks, how to remove that?

Tao Wang94
Intellectual 840 points
Part Number: AWR1443BOOST

Tool/software: Linux

What is a clutter and how to remove it? I tried to play with the configuration parameters, it seems I cannot find a way to remove the "noise": there is always something "seen" something about 1 meter in front of the radar, while there is nothing before the radar sensor, this is especially "obvious" if the sensor is under sunlight. If the sensor is put under shade or indoors, the situation is better. The horizontal view area is 180 degree and the vertical area is 60 degree. The "cfarCfg" is set to "0 2 8 4 3 0 768". 
  • 0
    TI__Guru*** 138329 points
    Hi,

    Can you please let us know what application/system you are trying to develop?

    What type of objects are you trying to detect? Are they static or moving?

    Thank you
    Cesar
  • 0 in reply to Cesar
    Intellectual 840 points
    Hello Cesar,
    Thanks for taking care my question. We are doing obstacle avoidance on UAV. Till now all the test is still on the ground of performance analysis. The basic test I am doing now is to make sure the radar can detect objects without fault alarms. This "one meter" stuff we got is something that is always there, (the basic configuration I use is the one for robotics), esp. in the sunlight, i.e., if I move the sensor to the shade or indoor, the "one meter" stuff will be much small but still there. The environment I tested I had attached to the previous topic linked to this topic, if you cannot find, I will attach again. No matter what object, static or moving, to detect, my problem now is the when there is nothing before the sensor, I always have this "one meter" staff.
    Best Regards,
    Tao Wang
  • 0 in reply to Tao Wang94
    Intellectual 840 points
    On our observation, when there is something in the view area of the radar, this "one meter" stuff mostly will not show. Is it possible that when there is nothing in the view area of the radar, the firmware will deliberately put something in its view area?
  • 0 in reply to Tao Wang94
    Intellectual 840 points
    I find the parameter "clutterRemoval", anyway, each time I change it from "0" to "1", the sensor will crash. Is this parameter conflict with other parameters? Here is the configure file when I enable clutter removal:

    sensorStop
    flushCfg
    dfeDataOutputMode 1
    channelCfg 15 7 0
    adcCfg 2 1
    adcbufCfg 0 1 0 1
    profileCfg 0 77 7 7 57.14 0 0 70 1 240 4884 0 0 34
    chirpCfg 0 0 0 0 0 0 0 1
    chirpCfg 1 1 0 0 0 0 0 4
    chirpCfg 2 2 0 0 0 0 0 2
    frameCfg 0 2 16 0 33.333 1 0
    guiMonitor 1 0 0 0 0 0
    %cfarCfg 0 2 8 4 3 0 768
    cfarCfg 0 2 8 4 3 0 768
    peakGrouping 1 1 0 1 229
    multiObjBeamForming 1 0.5
    clutterRemoval 1
    calibDcRangeSig 0 -5 8 256
    compRangeBiasAndRxChanPhase 0.0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0
    measureRangeBiasAndRxChanPhase 0 1.5 0.2
    CQRxSatMonitor 0 3 5 123 0
    CQSigImgMonitor 0 119 4
    analogMonitor 1 1
    sensorStart


    I am working with ROS driver under Ubuntu 16.04.
  • 0 in reply to Cesar
    Intellectual 840 points
    Hello Cesar,

    Any answer for my questions?
  • 0 in reply to Tao Wang94
    TI__Mastermind 21295 points
    Hello Tao Wang,

    The ROS Point Cloud Visualizer lab relies on the mmWave SDK mmw demo running on the mmWave sensor. Many of the parameters in the device configuration (.cfg) file are inter-dependent and can cause the demo to fail if they are not adjusted in concert. Therefore, the config file should be generated using the mmWave Demo Visualizer tool (dev.ti.com/mmWaveDemoVisualizer ) as takes all of the demo constraints into account. All of the configs provided in the ROS Point Cloud Visualizer lab were generated using the mmWave Demo Visualizer tool and a snapshot of the actual GUI settings used to generate each config file is located in the ti_mmwave_rospkg/cfg directory where the cfg files are located.

    Instructions for creating the config file and using it with the ROS Point Cloud Visualizer lab are given in the first 5 minutes of the following training video.
    training.ti.com/mmwave-sensors-robotics-technical-deep-dive

    As far as seeing phantom objects, the default Range Detection Threshold in the ROS Point Cloud Visualizer lab is set to an extremely small value to maximize the number of points seen in the visualization. This can cause very weak Radar reflections to be reported which may not be desired. The Range Detection Threshold can be increased in the mmWave Demo Visualizer to eliminate many of the weak reflections.

    If you are using a 3D configuration, then you can also look at the elevation of the reported objects. Often false objects can be reported that are outside the desired field-of-view of the antenna if there are strong reflectors outside the desired field-of-view. These can be eliminated by constraining the allowed field-of-view for objects reported by the ROS mmWave driver. This can be controlled using the "max_allowed_elevation_angle_deg" parameter in the ROS launch file (ti_mmwave_rospkg/launch/rviz_1443_3d.launch if using the 1443 3D ROS configuration).

    Please mark the thread as answered if your question is resolved or reply if more support is desired.

    Best Regards,
    John