Multiresolution Motion Estimation Using an Affine ModelStefan Kruger, Andrew Calway, Multiresolution Motion Estimation Using an Affine Model. CSTR-96-002, Department of Computer Science, University of Bristol. June 1996. PDF, 843 Kbytes.
This report describes a novel method for estimating motion in image sequences. The approach is based on a multiresolution model in which the motion field is considered to consist of a collage of different size regions with different local motions associated with them. The local motions are defined in terms of an affine model. The estimation procedure consists of a local component in which the affine motion parameters are estimated using a generalised correlation technique and a global component in which the final motion field is determined using a coarse-to-fine matching scheme. Efficient implementation of the scheme is achieved by the use of a generalised wavelet transform, the Multiresolution Fourier Transform (MFT). Results of experiments performed on both artificial and natural image sequences illustrate the potential of the approach.