A method for efficiently implementing 2D image rotations on distributed memory parallel machines

is proposed in this paper.The method has been developed in the context of developing the algorithms in terms of high level abstractions.With newly developed notations based on a 2-pass rotation technique

we show that the parallel image rotation can be expressed concisely and performed effectively on distributed memory parallel architectures.Error analysis has been done for the 2-pass operations

and compensation schemes suggested to reduce noises.