I have a lot of upper body/face pictures and i'm trying to create a page similar to this: http://www.mediatechsummit.com/ehome/index.php?eventid=45432&tabid=76964&
The problem is my pictures vary in size and type, some are more face while some are upper body. I want to write an algorithm that will scale/crop pictures to a specific ratio (1/1.3) using my implemented face detection - EMGU CV library.
it needs to be "smart", taking the square from face detected and decide how to keep it in the center while not loosing important parts from the picture (ears, forehead).