Natural scene text localization using edge color signature

Abstract

Localizing text regions in images taken from natural scenes is one of the challenging problems due<br />to variations in font, size, color and orientation of text. In this paper, we introduce a new concept so<br />called Edge Color Signature for localizing text regions in an image. This method is able to localize<br />both Farsi and English texts. In the proposed method rst a pyramid using different scales of the<br />input image is created. Then for each level of the pyramid an edge map is extracted. Afterward,<br />several geometric features are employed to lter out the non-text edges from the extracted edges.<br />At this stage we describe an edge using colors of its neighboring pixels. We use the mean-Shift<br />algorithm to obtain the color modes surrounding each edge pixel. Subsequently, the connected<br />edge pixels with similar color signatures are clustered using Single-Linkage clustering algorithm to<br />construct meaningful groups. Finally, each of the clusters is labeled as text or non-text using an<br />MLP based cascade classi er. The proposed method has been evaluated on well-known ICDAR 2013<br />and our Farsi dataset, the result is very promising.