Model-based Influence Diagrams for Machine Vision

TL;DR

This paper presents a method for automated control in machine vision using influence diagrams that represent hypotheses and processing decisions, integrating Bayesian inference for scene interpretation.

Contribution

It introduces a novel approach combining influence diagrams with hierarchical Bayesian inference for incremental scene analysis in machine vision.

Findings

Sequence of processing decisions matches the evaluation of the complete influence diagram.

Framework effectively represents and matches object instances and relationships in imagery.

Extends previous results to demonstrate equivalence in decision sequences.

Abstract

We show an approach to automated control of machine vision systems based on incremental creation and evaluation of a particular family of influence diagrams that represent hypotheses of imagery interpretation and possible subsequent processing decisions. In our approach, model-based machine vision techniques are integrated with hierarchical Bayesian inference to provide a framework for representing and matching instances of objects and relationships in imagery and for accruing probabilities to rank order conflicting scene interpretations. We extend a result of Tatman and Shachter to show that the sequence of processing decisions derived from evaluating the diagrams at each stage is the same as the sequence that would have been derived by evaluating the final influence diagram that contains all random variables created during the run of the vision system.