Shift based LBP

The function implements LBP (Local Binary Pattern analysis), in a shift based manner.

The LBP tests the relation between pixel and it's neighbors, encoding this relation into a binary word. This allows detection of texture and more complicated patterns & features.
The function is inspired by materials published by Matti Pietikainen in http://www.cse.oulu.fi/CMV/Research/LBP . This implementation however is not totally aligned with the methods proposed by Professor Pietikainen (see Issues & Comments). The current method proposes an implementation based on shifted images- which is beneficial in run time, especially in cases of large filter support. This is also a different way of thinking about the LBP algorithm. Two image shift functions are proposed- a circshift based one- which is faster, but less accurate, and filter2 based one- which takes more time and actually is exactly the same as efficientLBP proposed before.