I've pulled the code from depthai-experiments/gen2-people-tracker/api, and repurposed it. I changed the script.py file and removed the ability for it to run on a pre-recorded video, and only on streaming video. I am now trying to run it in a new python file, where I will have two separate pipelines running at once. I have the pipeline being instantiated, then setting the device with a "with" statement. It then is returning the RGB frames alongside the tracklets with the .getOutputQueue() method.

inside of a while loop, I have the RBG output queue (rgbQ), trying to pull the RGB frames usingrbgQ.tryGet() This runs for a few seconds, until I then get the following error

[system] [critical] Fatal error. Please report to developers. Log: 'Fatal error on MSS CPU: trap: 00, address: 00000000' '0'

This hasn't happened in any other scripts I've ran. What does this mean, and how can I get around it?

Hi @AlexMcGraw
MSS CPU trap usually indicates a pipeline problem. Try changing the tracking type (tracking issues are usually the culprit).

Thanks ,
Jaka

    jakaskerl

    I have the following object tracker set up

    # Create and configure the object tracker

    objectTracker = pipeline.create(dai.node.ObjectTracker)

    objectTracker.setDetectionLabelsToTrack([1]) # Track people

    # possible tracking types: ZERO_TERM_COLOR_HISTOGRAM, ZERO_TERM_IMAGELESS, SHORT_TERM_IMAGELESS, SHORT_TERM_KCF

    objectTracker.setTrackerType(dai.TrackerType.ZERO_TERM_COLOR_HISTOGRAM)

    Are you referring to the objectTracker.setTrackerType() line? This works in a different script that I have, but I'm having to create a new file, and it just all of a sudden doesn't want to work, even though the pipeline is identical.

    jakaskerl

    I have this running in a multi-threaded set up, and it errors out on .join() when looping thru the threads.

    Hi @AlexMcGraw
    Hmm, try a newer version of depthai if not already on the latest. If that doesn't solve the issue, please create a MRE.

    Thanks,
    Jaka