We developed a segmentation free word spotter, which means that each word do not have to be extracted from the text. Instead a sliding window (cyan box) is used to traverse the document. Key point matching was performed on four sets of different key points, using a descriptor based on the Fourier transform followed by a relaxed outlier removal method in two steps. The green box depicts the computed bounding box (the bounding box algorithm was not yet used) and the green dots are the key points that are found to correspond with the query word.
The word spotter is under constant development and a new version will be submitted for publication soon!
- A Segmentation-Free Handwritten Word Spotting Approach by Relaxed Feature Matching, A. Hast and A. Fornés, 2016 12th IAPR Workshop on Document Analysis Systems (DAS), Santorini, 2016, pp. 150-155. doi: 10.1109/DAS.2016.40
@INPROCEEDINGS{7490109, author={A. Hast and A. Fornés}, booktitle={2016 12th IAPR Workshop on Document Analysis Systems (DAS)}, title={A Segmentation-Free Handwritten Word Spotting Approach by Relaxed Feature Matching}, year={2016}, pages={150-155}, doi={10.1109/DAS.2016.40}, month={April}}
Anders Hast 