Deploying vision.PeopleDetector into Raspberry PI as standalone device.

I have read this documentation:

https://uk.mathworks.com/help/vision/ref/vision.peopledetector-system-object.html#d122e99012

Section "Inputs Arguments" mentions:

-Input image - I have one.

-ROI - I don't use that.

- Classification model - I specified one.

However, when I try to deploy an algorithim into RPI the Matlab Coder returns this error: "The library method 'vision.internal.buildable.HOGDescriptorBuildable.updateBuildInfo' failed. Caused by: Not enough input arguments.".

vision.People.Detector is supported for code generation.

How can I deploy vision.People.Detector into RPI?

What am I doing wrong?

My code:

function raspi() %#codegen

mypi = raspi();

cam = cameraboard(mypi,'Resolution','640x480');

peopledetector = vision.PeopleDetector;

peopledetector.ClassificationModel = ('UprightPeople_128x64');

peopledetector.ClassificationThreshold = 1;

peopledetector.MinSize = [];

peopledetector.MaxSize = [];

peopledetector.ScaleFactor = 1.05;

peopledetector.WindowStride = [8 8];

peopledetector.MergeDetections = 1;

peopledetector.UseROI = 0;

for f = 1:500

I = snapshot(cam);

[bboxes,scores] = peopledetector(I);

if bboxes ~= 0

% Insert boxes into image

I = insertObjectAnnotation(I,'rectangle', bboxes, scores);

end

% Display image

displayImage(mypi,I,'Title','Detector')

end