A Robust Line-Feature-Based Hausdorff Distance for Shape Matching

Abstract

The Hausdorff distance can be used to measure the similarity of two point sets. In matching the two point sets, one of the point sets is translated, rotated and scaled in order to obtain an optimal matching, which is a computationally intensive process. In this paper, a robust line-feature-based approach for model-based recognition is proposed, which can achieve a good performance level in matching, even in a noisy environment or with the existence of occlusion. The method is insensitive to noise and can find the rotation and scale of the image point set accurately and reliably. For this reason, instead of 4D matching, a 2D-2D matching algorithm can be used. This can greatly reduce the required memory and computation. Having rotated and scaled the image point set, the difference between the query point set and the model point set can be computed by considering translation only. The perfamvance and the sensitivity to noise of our algorithm are evaluated using simulated data. Experiments show that our 2D-2D algorithm can give a high performance level when determining the relative scale and orientation of two point sets.