ABSTRACT
This chapter provides a description of different reference algorithms for motion estimation between two or more images, to analyze the displacement of rigid or non-rigid objects in image sequences.
The chapter describes two different approaches for this purpose: dense motion estimation and sparse estimation. Dense estimation is performed with the so-called optical flow (and its variants), with the well-known Horn-Schunk and Lucas-Kanade approaches. Sparse estimation is based on spatial correlation and phase correlation, that can be estimated in the Fourier space.
