On the Verification of Hypothesized Matches in Model-Based Recognition
MASSACHUSETTS INST OF TECH CAMBRIDGE ARTIFICIAL INTELLIGENCE LAB
Pagination or Media Count:
In model-based recognition a number of ad hoc techniques are used to decide whether or not a match of data to a model is correct. Generally an empirically determined threshold is placed on the fraction of model features that must be matched. In this paper we present a more rigorous approach in which the conditions under which to accept a matched are derived based on fundamental grounds. We obtain an expression that relates the probability of a matched occuring at random to the reaction of a model features that are accounted for by the match. This expression is a function of the number of model features, the number of image features, and a bound on the degree on the degree of sensor noise. One implication of our analysis is that a proper threshold for matching must vary with the number of model and data features. Thus, it is important to be able to set the threshold as a function of a particular matching problem, rather than setting a single threshold as a function of a particular matching problem, rather than setting a single threshold based on experimentation. We analyze some existing recognition systems and find that our method yields threshold similiar to the ones were determined empirically for these systems, providing evidence of the validity of the technique.