Probabilistic Model of Multiple Dynamic Curve Matching for Semitransparent Scene of Nonrigid Objects (1995)
| Venue: | in Vision Geometry IV |
| Citations: | 2 - 2 self |
BibTeX
@INPROCEEDINGS{Mallouche95probabilisticmodel,
author = {Hail Mallouche and Hail Mallouche and Hail Mallouche and Jacques A. De Guise and Jacques A. De Guise and Yves Goussard and Yves Goussard},
title = {Probabilistic Model of Multiple Dynamic Curve Matching for Semitransparent Scene of Nonrigid Objects},
booktitle = {in Vision Geometry IV},
year = {1995},
pages = {148--157}
}
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
High-level and knowledge-based modeling for scene understanding, organization, and recognition are of great interest in psychophysical and machine vision research. On a lower level, dynamic contours have become an attractive subject in computer vision because of their connectivity and closure properties. Here to fore, proposals were restricted to using one deformable curve, disjoint curves, or elastic models. In this paper, we present a probabilistic model of multiple dynamic free curve matching with 2-D images. The model is conceived for images from semitransparent scenes such as additive projection imaging -- planar radiographical transmission or planar emission imaging. The model is made up of three parts: (i) a data driven term, (ii) a model driven term, and (iii) smoothing constraints. The model represents a hierarchical structure of three processes: lines, regions and relational graph. Thus, an object is modeled as a set of linked subobjects according to a 3-D relational graph th...
