• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 30 subscribers
  • Views 592 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: About 3D People Counting for Falling Detection GUI

James Lee
Intellectual 485 points
Part Number: IWR6843ISK

Hello TI,

I want to comprehend the 3D people counting about Falling Detection GUI in industrial_toolbox_4_5_1(The folder is \ti\industrial_toolbox_4_5_1\labs\people_counting\visualizer\fallDetection_main.py).

The following figure is code in fallDetection_main.py.

I have two questions and there are two red rectangles that I made:

1.In first red rectangle,what is the meaning about the age? I don't understand the physical meaning.

2.In second rectangle,I want to know the condition of deciding falling detection situation.

Why "self.targetSize[6,tid,fNum]= self.targetSize[5,tid,fNum]-self.targetSize[5,tid,(fNum-10)%100]"  is made as the falling detection?


I'll appreciate your response!

Best regards,

James

  • 0
    TI__Genius 14415 points

    Hi James,

    Age (or targetSize[4,:,:]) is the number of frames a particular target has been active. We only start reporting height data after 2 seconds (age>40 frames).

    targetSize[5,:,:] is the rolling average height over the past 10 frames. So the logic is that if the average height  changes by more than the fallThreshold compared to 10 frames previous, the fall detection alert is set. A simplified equation for illustration would be targetSize[6,,] = height(fnum) - height(fnum-10)

    Hopefully that makes sense. 

    Regards,

    Jackson

  • 0 in reply to Jackson Thomas
    Intellectual 485 points

    Hello Jackson,

    It's nice to receive your response.I know the logic that you have explained.

    As we discussed previously,you simplified the equation as "targetSize[6,,] = height(fnum) - height(fnum-10)",but the original expression in the fallDetection_main.py is "targetSize[6,,] = height(fnum) - height((fnum-10)%100)".

    It confuses me why (fnum-10)%100? Why the frame-10 need to take modulo 100 ?

    Thank you for your reply.

    Best regards,

    James

  • 0 in reply to James Lee
    TI__Genius 14415 points

    Hi James,

    If you look at the definition of the variable fnum, it is not actually frame number directly, it is also modulo 100.

    "fNum = self.frameNum%100"

    fNum loops every 100 frames to keep the memory requirement down. Since we only use a 10 frame rolling average, this is plenty.

    Regards,

    Jackson

  • 0 in reply to Jackson Thomas
    Intellectual 485 points

    Hello Jackson,

    It is clear for your response.

    I know the definition of the variable fnum and I confirm my idea after your reply,it's same as I thought.

    But I still have other questions,when we test the falling detection,why "Long Range People Detection" is recommended?

    And when I tried to use the bin file of Long Range People Detection,the sensor and GUI couldn't work.

    Best regards,

    James