In order to recover an accurate representation of a scene containing multiple moving objects, one must use estimation methods that can recover both model parameters and segmentation at the same time. We introduce a new layered model of scene segmentation based on explicitly representing the support of a homogeneous region. Our model employs parallel robust estimation techniques, and uses a minimal-covering optimization to estimate the number of objects in the scene. Using a simple direct motion model of translating objects, we successfully segment real image sequences containing multiple motions.
Trevor J. Darrell,
Alexander P. Pentland,
"Multilayer robust estimation for motion segmentation", Proc. SPIE 1613, Mobile Robots VI, (14 February 1992); doi: 10.1117/12.135181; https://doi.org/10.1117/12.135181