Structural sensitivity for large-scale line-pattern recognition

Abstract

This paper provides a detailed sensitivity analysis for the problem of recognising line patterns from large structural libraries. The analysis focuses on the characterization of two different recognition strategies. The first is histogram-based while the second uses feature-sets. In the former case comparison is based on the Bhattacharyya distance between histograms, while in the latter case the feature-sets are compared using a probabilistic variant of the Hausdorff distance. We study the two algorithms under line-dropout, line fragmentation, line addition and line end-point position errors. The analysis reveals that while the histogram-based method is most sensitive to the addition of line segments and end-point position errors, the set-based method is most sensitive to line dropout.