contains material from
Template Matching Techniques in Computer Vision: Theory and Practice
Roberto Brunelli © 2009 John Wiley & Sons, Ltd
Matching sets of points using techniques targeted at area matching is far from optimal, regarding both efficiency and effectiveness. This chapter shows how to compare sparse templates, composed by points with no textural properties, using the Hausdorff distance. Robustness to noise and template deformation as well as computational efficiency are analyzed. A probabilistic perspective on Hausdorff matching is briefly discussed. Invariant moments, a classical technique for shape matching is also considered.
keywords: Hausdorff distance, invariant moments, distance transform, metric pattern space, principal components analysis.