Decision Trees for Geometric Models
- 1 June 1998
- journal article
- Published by World Scientific Pub Co Pte Ltd in International Journal of Computational Geometry & Applications
- Vol. 8 (3) , 343-363
- https://doi.org/10.1142/s0218195998000175
Abstract
A fundamental problem in model-based computer vision is that of identifying which of a given set of geometric models is present in an image. Considering a "probe" to be an oracle that tells us whether or not a model is present at a given point, we study the problem of computing efficient strategies ("decision trees") for probing an image, with the goal to minimize the number of probes necessary (in the worst case) to determine which single model is present. We show that a ⌈l g k⌉ height binary decision tree always exists for k polygonal models (in fixed position), provided (1) they are non-degenerate (do not share boundaries) and (2) they share a common point of intersection. Further, we give an efficient algorithm for constructing such decision tress when the models are given as a set of polygons in the plane. We show that constructing a minimum height tree is NP-complete if either of the two assumptions is omitted. We provide an efficient greedy heuristic strategy and show that, in the general case, it yields a decision tree whose height is at most ⌈l g k⌉ times that of an optimal tree. Finally, we discuss some restricted cases whose special structure allows for improved results.Keywords
This publication has 22 references indexed in Scilit:
- An efficient algorithm for identifying objects using robot probesThe Visual Computer, 1994
- Probing polygons minimally is hardComputational Geometry, 1993
- Model-based probing strategies for convex polygonsComputational Geometry, 1992
- Interactive reconstruction via geometric probingProceedings of the IEEE, 1992
- An optimal algorithm for intersecting line segments in the planeJournal of the ACM, 1992
- Simplified linear-time jordan sorting and polygon clippingInformation Processing Letters, 1990
- How to find a battleshipNetworks, 1989
- Decision tree design from a communication theory standpointIEEE Transactions on Information Theory, 1988
- Determining the shape of a convex n-sided polygon by using 2n + k tactile probesInformation Processing Letters, 1986
- Model-based recognition in robot visionACM Computing Surveys, 1986