Using Corner Feature Correspondences to Rank Word Images by Similarity
Abstract
Libraries contain enormous amounts of handwritten historical documents which cannot be made available on-line because they do not have a searchable index. The wordspotting idea has previously been proposed as a solution to creating indexes for such documents and collections by matching word images. In this paper we present an algorithm which compares whole word-images based on their appearance. This algorithm recovers correspondences of points of interest in two images, and then uses these correspondences to construct a similarity measure. This similarity measure can then be used to rank word-images in order of their closeness to a querying image. We achieved an average precision of 62.57% on a set of 2372 images of reasonable quality and an average precision of 15.49% on a set of 3262 images from documents of poor quality that are even hard to read for humans.
Cite
Text
Rothfeder et al. "Using Corner Feature Correspondences to Rank Word Images by Similarity." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2003. doi:10.1109/CVPRW.2003.10021Markdown
[Rothfeder et al. "Using Corner Feature Correspondences to Rank Word Images by Similarity." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2003.](https://mlanthology.org/cvprw/2003/rothfeder2003cvprw-using/) doi:10.1109/CVPRW.2003.10021BibTeX
@inproceedings{rothfeder2003cvprw-using,
title = {{Using Corner Feature Correspondences to Rank Word Images by Similarity}},
author = {Rothfeder, Jamie L. and Feng, Shaolei and Rath, Toni M.},
booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
year = {2003},
pages = {30},
doi = {10.1109/CVPRW.2003.10021},
url = {https://mlanthology.org/cvprw/2003/rothfeder2003cvprw-using/}
}