Nao Robot and "FaceDetected" variable

Question

I am trying to do some visual debugging of the Nao's face detection, just draw a rectangle around the detected(s) face(s). Instead of acquire the image and use an opencv algorithm to do the face detection, I am acquiring the image (for visual debugging) and using the parameters returned by Nao's face detection system.

This system is provided by Oki and all the information is available into the variable "FaceDetected". I have not problem to access to this variable. For my porpoise it is supposed I have to use "ShapeInfo" information.

My questions is: How can I manipulate them to get the face position into the Cartesian image space?

Note 1: I suppose the units are radians, but I have not official confirmation

Note 2: After some experiments, I have discovered that alpha and beta are relatives to the image centre and sizeX and sizeY increases/decreases in relation with the face size (as is expected). So, I suppose the information is given as an Orthographic projection

Note 3: Some examples of alpha, beta (range [-0.5,0.5]), sizeX, sizeY (range [0,1]) for a face located into the area top-right of the camera:

-0.232393 -0.0899552 0.358549 0.37366 
-0.232393 -0.0899552 0.358549 0.37366 

Answer 1

After a lot of tests I have finally discovered the meaning of the FaceDetected variable. I've written a explanation here

I hope it could be useful to others