masking for computer vision

Hi everybody, im working on a proyect that uses the camera in the roof to track the peoples position in a space.

I would like to track only a specific area of the videos and not the whole video. The camera in the roof shows the whole space , and i would like to track just a reduced space.

The camera in the roof allows me to track a space of 6 * 5 meters but i just want track a reduced space of 2*3 and not the whole space. Is there is a way of masking a specific part of the video , i would like to select a specific area of the video and just track that. Is that possible

any advices?

thanks

Adam

check for the setROI methods in ofxOpenCv images