Detail of Publication
Back to list
||Seiichi Uchida, Masakazu Iwamura, Shinichiro Omachi, Koichi Kise
||Data Embedding for Camera-Based Character Recognition
||Meeting on Image Recognition and Understanding 2005 (MIRU2005)
|Month & Year
||In this paper, we propose a data embedding method for camera-based
character recognition as easy and accurate as bar-code reading.
Specifically, a horizontal stripe pattern, called cross ratio pattern,
is embedded into a font image of a character category. Since the cross
ratio obtained from the pattern is invariant to projective distortion,
it always provides correct category candidates regardless of camera angle.
Experimental results showed that (i) the cross ratio can be extracted from
distorted character images with very high accuracy and thus (ii) the
cross ratio can be used for enhancing recognition performance
of conventional matching-based recognizers.