P. Parent and S. W. Zucker. Trace inference, curvature consistency, and curve detection. Technical Report CIM-86-3, Computer Vision and Robotics Laboratory, McGill Research Center for Intelligent Machines, McGill University, Quebec, Canada, June 1985.  Home/Search   Document Not in Database   Summary   Related Articles

....is a good starting point for the perception of objects in general, as was first strongly proclaimed by Gibson [28, p.8#1] 2.2. Cooperation between curve and surface sensing. There has been and still is a lot of research in the perception of two dimensional curves in intensity images [54, 51] [46, 42, 4][34] this research is interesting in its own right, because as for the reading example above, there are a number of visual tasks that can be restricted to a two dimensional world. But methods that are successful in some dimensionality are not necessarily easily extended to higher dimensions, 1 ....

....neighbours is not insured. The initial estimates are therefore further refined by an iterative process which enforces a locally constant curvature constraint over the trace points. Sander attempts to do this in a framework similar to the relaxation labelling process used to extract image curves in [46] and [34] However, the complexity of the representation precludes the use of a discrete set of labels, as in relaxation labelling. Instead, Sander applies the same local minimization principle using continuous values and variational methods, thus the name variational relaxation. The iterative ....

P. Parent and S. W. Zucker. Trace inference, curvature consistency, and curve detection. Technical Report CIM-86-3, Computer Vision and Robotics Laboratory, McGill Research Center for Intelligent Machines, McGill University, Quebec, Canada, June 1985.

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