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: Modifiying oob_parser.py to read .bin and convert to .csv

Part Number: IWR6843ISK

Hi Team,

A customer would like to save the points clouds in ASCII(.csv) format. As per the visualizer user guide, they have changed self.saveTextFile equal to 1 in oob_parser.py. The customer runs the gui_main.py through Jupyter and they are able to produce a .csv file. However, the point cloud information (each point with corresponding range, velocity, and acceleration information) is not present in these files. 

This thread mentioned modifying the parser function in oob_parser.py to read its input from the saved .bin file (instead of UART) and save it as .csv. Can you explain how to proceed?

Thank you for your help.

Regards,

Marvin

  • Hi Marvin,

    Brief overview of the steps as I see them:

    1. Find the section of oob_parser.py which parses the specific demo which the customer is using (there are slightly different parsers for each).
    2. Find the area of the parsing function which contains the data you need (point cloud)
    3. Add functions to write this data to CSV (code examples for this are readily available online)

    Without additional information, I cannot get much more specific than this. Let me know if you have any further questions.

    Best Regards,
    Alec

  • Hi Alec,

    Thank you for your response.

    The customer is following the user guide which mentions:
    To save either of these files, in oob_parser.py, set self.saveBinary and/or self.saveTextFile equal to 1, save the file, then run start the gui as described.

    So, they have updated the oob_parser.py. and set the self.saveBinary to 0 and set the self.saveTextFile to 1. The script directly save the .csv files just that the data is not correct. (files from customer is provided below)

    The thread I mentioned earlier, points out that the script can read the .bin and convert it to .cvs. I was hoping you can provide guidance/example on how to do this as I am not familiar with this.

    2548.files.zip 

    Thanks again, and please let me know if you need more information. 

    Regards,

    Marvin

  • Marvin,

    Can you provide background on what demo they are running with this? 

    Best Regards,
    Alec

  • HI Alec,

    They were running the 3D people counting demo and were interested in getting the .csv fine for 3D people counting and HVAC labs.

    I hope this helps. Let me know if you need more information.

    Regards,

    Marvin

  • Hi Marvin,

    One more follow up question, what version of the toolbox is the customer using?

    Thank you,

    Angie 

  • Hi Angie,

    they are using Industrial tool box 4.11.0.,

    Regards,

    Marvin

  • Hi Marvin,

    When I run it, it seems to work. It appears to me that nothing got written in the file sent. What does the pHistText_0.csv file look like? Are they all empty?

    My file attached for comparison

    pHistText_0.csv

    Best,

    Nate

  • Hi Nathan,

    Can you tell me how you run this? can you share what you modify in the script to get this result?

    Thanks again.

    Regards,

    Marvin

  • Hi Marvin,

    I took the industrial toolbox 4.11 visualizer, modified oob_parser.py so that saveTextFile=1, then ran it in python with python gui_main.py I would also note that you have to wait 1000 frames to get the first text file, so perhaps that's where the problem is? The script saves the previous 1000 frames at each multiple of 1000, so for example if you want the results from frames 1300-1500, you'll need to run the script until it gets to frame 2000.

    Best, 

    Nate

  • Hi Nathan,

    Thanks for helping. 

    The customer is now running the long range people detection and would like to have the output parsed and saved in a .csv format.

    However, the file being saved does not contain any data. Can you confirm if this get point clouds even with long range people detection firmware and corresponding config file?

    I found this thread  and advice the customer to do the same but it does not work either.

    Thank you for helping furhter.

    Regards,

    Marvin

  • Hi Marvin,

    Please open a new thread for new questions. When you do so, can you please specify what happens when the customer tries to use the python visualizer for the long-range people detection, and why that isn't allowing them to save the data?

    Best,

    Nate