Modified Hausdorff distance for model-based 3-D object recognition from a single view

Abstract In this paper, we consider the problem of recognizing 3-D object from a single 2-D intensity image, obtained from unknown position and orientation. We propose the feature set based correspondence algorithm between 3-D model features and 2-D image features, in contrast to the conventional approaches which use a single local feature or fixed number of local features. We describe the model and the image as the set of the feature sets, which are then used as a basic unit to compare the similarity. As a measure of the similarity between features, the Hausdorff distance with explicit paring (HDEP) is proposed and extended to the partial HDEP, using the notion of the partial distance to cope with the problems which occur when there are backgrounds and some of image features are missing or severely deviated from the original. The proposed correspondence algorithm employing the partial HDEP is the generalized form which is able to vary the number of local relations to be considered flexibly from a single local relation to set of local relations. The probabilistic measure which reflects the view variation of the projected features is used to measure the similarity between the elements of the feature set. The 3-D object recognition system with hypothesis-verification scheme is implemented and tested on real images with backgrounds and occlusion.