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: explanation of fall detection algorithm

Part Number: IWR6843ISK

Someone explain how to calculate fall detection in IWR6843ISK and explain algorithm for calculating average height and absolute height for fall detection experiment

  • Hello,

    Fall detection is currently using a simple algorithm to determine a fall event. If the height of the tracked person changes by more than a certain threshold in the previous 10 frames, it is considered a fall. This could be more robust if you use the point cloud to measure target height instead of the tracker centroid. 

    The source code for the visualizer is at the toolbox path below. You can see the fall detection logic in the updateGraph function.

    "C:\ti\mmwave_industrial_toolbox_4_9_0\labs\people_counting\visualizer\fallDetection_main.py"

    Regards,

    Jackson

  • I want know how this formula (self.targetSize[5,tid,fNum]= (1/self.numFrameAvg*self.targetSize[0,tid,fNum])+((self.numFrameAvg-1)/self.numFrameAvg)*self.targetSize[5,tid,(fNum-1)%100] #avg height over 10 frames) for average height is determined and please explain it.

  • Hello,

    This function is just as the comment states, it is a simple average height measurement of the tracker height over the previous 10 frames. This uses a sliding window to weight the average.

    This code is provided as an example, you are free to modify anything you need. You may need to step through and debug the code to understand all of the numbers.

    Regards,

    Jackson