Using bottom-up image processing algorithms to predict human eye fixations and extract the relevant embedded information in images has been widely applied in the design of active machine vision systems. Scene text is an important feature to be extracted, especially in vision-based mobile robot navigation as many potential landmarks such as nameplates and information signs contain text. This paper proposes an edge-based text region extraction algorithm, which is robust with respect to font sizes, styles, color/intensity, orientations, and effects of illumination, reflections, shadows, perspective distortion, and the complexity of image backgrounds. Performance of the proposed algorithm is compared against a number of widely used text localization algorithms and the results show that this method can quickly and effectively localize and extract text regions from real scenes and can be used in mobile robot navigation under an indoor environment to detect text based landmarks.
An Edge-based Text Region Extraction Algorithm for Indoor Mobile Robot Navigation
2007-07-20
oai:zenodo.org:1332556
International Journal of Information, Control and Computer Sciences 0.0(7)
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
DDC: | 629 |
Indoor mobile robot navigation using doorplate landmarks
Kraftfahrwesen | 2006
|A New Method of Extraction of Mobile Robot Navigation Line Based on Edge Detection
Trans Tech Publications | 2010
|