Abstract

Three efficient algorithms for sub-pixel image registration, based on nonlinear optimization and the discrete Fourier transform, are developed to compute a translation invariant error metric. Their accuracy and computational performance is investigated and compared.

© 2007 OSA

PDF Article