A skeleton based descriptor for detecting text in real scene images.

Mehdi Felhi, Nicolas Bonnier, Salvatore Tabbone

Browse the full ICPR paper archive.