| P. Werkhoveh, A. Toet, and J.J. Koenderink. Displacement estimates through adaptive affinities. PAMI, 12(7):658--663, July 1990. |
....methods can be classified into two categories: the local patch based methods and the model based methods. The local patch based methods simply abandon the idea of recovering a whole body description of the motion, and recover structure on a patch by patch basis in order to cope with nonrigidity [16, 20]. Unfortunately, using a local description limits the methods themselves to using only noise sensitive local measurements. Consequently, such patch bypatch recovery of structure is not likely to be either very meaningful or robust. On the contrary, the model based methods utilized predefined 3D ....
P. Werkhoveh, A. Toet, and J.J. Koenderink. Displacement estimates through adaptive affinities. PAMI, 12(7):658--663, July 1990.
.... of the motion field, but lacks good quantitative behavior, especially when the images involve rigid body motion with possible occlusion [2, 3] In these cases, parametric methods such as those reported in [4, 5, 6, 7, 8, 9, 10] and modified Horn and Schunck methods such as those reported in [2, 3, 11, 12, 13, 14, 15, 16, 10] show superior quantitative performance. When the images show an object undergoing deformable motion with no occlusion, however, SOF may still provide a high resolution, accurate estimate of the motion field. In these cases there are many parameters affecting the performance of SOF including ....
P. Werkhoven, A. Toet, and J.J. Koenderink. Displacement estimates through adaptive affinities. IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(7):658-- 663, July 1990.
....the optimal velocity estimate. These stochastic formulations of HSOF provide the basis for the motion estimation algorithms developed in this dissertation and are presented in more detail in Chapter 2. Other Optical Flow Methods Other optical flow algorithms have been developed from the BCE [41, 42, 43, 44, 45, 46, 33, 37]. Most are mainly aimed at relaxing the smoothness assumption in order to to handle image sequences where occluding objects cause significant discontinuities in the velocity field. Other algorithms have been developed using second order spatial derivatives of the brightness function [47, 48, 49, ....
P. Werkhoven, A. Toet, and J.J. Koenderink. Displacement estimates through adaptive affinities. IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(7):658--663, July 1990.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC