• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 30 subscribers
  • Views 561 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

IWR6843ISK: UART data Parser

Divyank Saxena
Prodigy 110 points
Part Number: IWR6843ISK

Hello There,

Lab Used: "Long range People Detection"

Chirp Config: "people_detection_and_tracking_100m_2D_advanced.cfg" 

 I am parsing sensor data using python by taking "People Counting Visualizer" as a reference.

 It is working properly except the range for people tracking is about 35 meters, whereas, I got approx. 80 meters of range in tm_visualizer. 

 Why there is range difference, is there any different algorithm used in tm_visualizer for people tracking??

  • 0
    TI__Genius 14415 points

    Hello,

    Can you be more specific about the 'range tracking is about 35 meters'? Do the tracks disappear at 35 meters, or do they stop moving after that distance is reached? The python based GUI was not really designed for really long range, so the draw distance might not extend so far.

    Have you tried running the python gui from the source file and debugging whether the track is still output on the UART line or if the GUI is receiving it? The GUI should not affect the data coming out of the device, it will only display it differently. All the tracking and processing is done on the device so the GUI will not affect performance.

    Regards,

    Jackson

  • 0 in reply to Jackson Thomas
    Prodigy 110 points

    Hello Jackson,

    I am using the oob_parser.py file to see the detected targets by simply printing the variable "targets" in the function "parseDetectedTracksSDK3x",  I am not using the gui_main.py for graphical interface, instead I am printing the values on terminal.

    After approx. 35 meters the variable targets print an empty array, means not target is being detected, while tm_visualizer gui perfectly detect and tracks people up to 80 meters.

    By reading the code of python visualizer and tm_visualizer, I found there is no major difference in parsing the TLV, but if any, please let me know.

    Have you tried running the python gui from the source file and debugging whether the track is still output on the UART line or if the GUI is receiving it? The GUI should not affect the data coming out of the device, it will only display it differently. All the tracking and processing is done on the device so the GUI will not affect performance.

    Yes I've tried it, the tracks are not displayed on terminal.

  • +1 in reply to Divyank Saxena
    TI__Genius 14415 points

    Hello,

    Understood. This is strange. I am assuming you have the same binary file loaded and are using the same CFG file and mounting setup?

    The gui does not change how the device is performing, it is only an output. So it seems there might be an error in the parsing once the range becomes too long in the python parser. I am not sure why that might happen. 2 suggestions.

    1. Pause the matlab TM gui when a person is being tracked at >35m and verify that the parsed data matches the gui and no values seem wrong.

    2. Look at a single frame of raw data in the python parser when a person should be tracked at >35m and verify if there is a target in the UART stream. This is what I expect you will find. From there it may make it easier to see why it is being parsed incorrectly.

    Regards,

    Jackson