Principles of spatial processing
Try [http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/pdfs/Ch2.pdf John Ashburner's realignment chapter] in the [http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/ Human Brain Function book] - second edition.
See also the other chapters on "Computational anatomy" in [http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/ HBF2].
Mark Jenkinson and Steve Smith have an excellent introduction to the problems of optimizing in registration - see the [http://www.fmrib.ox.ac.uk/analysis/techrep/tr00mj2/tr00mj2/index.html HTML version] or [http://www.fmrib.ox.ac.uk/analysis/techrep/tr00mj2/tr00mj2.pdf pdf version].
You may get some benefit from the theory pages at the [http://bishopw.loni.ucla.edu/AIR5/index.html AIR website]. AIR is a realignment / coregistration / normalization program written by Roger Woods of UCLA. His pages go into details of the implementation of spatial transforms, some of which are relevant to SPM.
Note that, one problem with movement correction is that the estimation of movement parameters can be biased by large activation: see http://www-sop.inria.fr/epidaure/Collaborations/IRMf/INRIAlign.html, and L. Freire and J.-F. Mangin. Motion correction algorithms may create spurious brain activations in the absence of subject motion. Neuroimage 14(3), p. 709-722, september 2001.