Short Description
This plugin computes a displacement map from a sequence of at least two consecutive images.Documentation
This plugin analyses a pair of images and retrieves the displacement that transforms the first image into the second.
This is implemented as a 2D variational optical flow method. The problem formulation follows the energy-based Horn-Schunck method (B. Horn and N. Schunck, Determining Optical Flow, Artificial Intelligence, 17, 185-203, 1981). The discretization and the solving method differ from the original article in the following ways: the discretization is done with finite elements; the resulting linear system is solved with a Conjugate Gradient algorithm.
Note: there is a major limitation in the current implementation. There is no guarantee that it will detect properly movements larger than 1 pixel. This will be fixed in a future version, by doing a multiscale analysis.
2 reviews on “Optical Flow – Horn-Schunck”