SubME is the MATLAB implementation of a block matching motion estimation algorithm that achieves subpixel accuracy without interpolation.
In conventional block matching motion estimation algorithms, subpixel motion accuracy is achieved by searching the best matching block in an enlarged (interpolated) reference search area. This, however, is computationally expensive as the number of operations required is directly proportional to the interpolation factor.
SubME integrates the block matching algorithm and optical flow method to estimate the motion. Motion vectors are determined by a two-stage algorithm, with the first stage being a single layer block matching, and the second stage being a first order optical flow by solving a 2x2 linear system.
For additional information and citations, please refer to:
S. H. Chan, D. Vo, and T. Q. Nguyen, “Sub-pixel motion estimation without interpolation”, in Proceedings of IEEE Conference on Acoustics, Speech and Signal Processing (ICASSP '10). p. 722-725, 2010.