Page 107 - Phase Space Optics Fundamentals and Applications
P. 107

88   Chapter Three


                   P input                                          P out
                              L 1          L 2         L 3       L 4








                         Z            Z          Z           Z

               FIGURE 3.7 Scheme for the flexible setup performing the signal rotator.



               been recently proposed. 54  It has been shown that four is a minimal
               number of generalized lenses located in fixed positions needed to
               preform the signal rotator. The optical scheme of the signal rotator at
               angle   is displayed in Fig. 3.7. If the distances between all elements
               equal z, then the block matrices of the applied lenses are given by


                                          1  3 + cos    sin
                         G 1 ( ) = G 3 (− ) =
                                          2z   sin     5 − cos

                                 4 00
                         G 2 ( ) =
                                 z 01

                                 2 1 + cos    − sin
                         G 4 ( ) =                                  (3.81)
                                 z  − sin    1 − cos
               This scheme can be realized by implementation of analog generalized
               lenses, which consist of three cylindrical lenses for L 1 and L 3 and one
               cylindrical lens for L 2 and L 4 .
                 By using the optical setups performing the basic phase-space rota-
               tions, other phase-space rotators can be constructed as their cascades.
               Nevertheless there is no guarantee that the obtained setup is optimal.



          3.7 Applications of Phase-Space Rotators

               3.7.1 Generalized Convolution
               As mentioned before, the well-known phase-space rotator—the
               Fourier transform—plays a crucial role in signal and image process-
               ing.Itformsabaseforshift-invariantfilteringwhichisusedforpattern
               recognition, denoising, encryption, etc. Many good books are devoted
               tothissubject;see,e.g.,Refs.9and10.Hereweconsidertheapplication
   102   103   104   105   106   107   108   109   110   111   112