Character recognition in external scenes by means of vanishing point grouping

We present a method to extract characters from images of outdoor scenes by means of a suitable grouping procedure. The candidate characters are recognized by simple low-level segmentation techniques, and a further elaboration allows to group them in consistent chains by means of their relation to the vanishing points of the scene. The method has been trained and tested using two different sets of images, recorded without any particular attention to the light conditions and the tilt angle of the camera, and very good results are obtained.

[1]  Jake K. Aggarwal,et al.  Determining vanishing points from perspective images , 1984, Comput. Vis. Graph. Image Process..

[2]  Beatrice Brillault-O'Mahony,et al.  New method for vanishing point detection , 1991, CVGIP Image Underst..

[3]  C. Coelho,et al.  Extraction of vanishing points from images of indoor and outdoor scenes , 1993, Image Vis. Comput..

[4]  Josef Kittler,et al.  Vanishing point detection , 1993, Image Vis. Comput..

[5]  Shigeru Akamatsu,et al.  Recognizing Characters in Scene Images , 1994, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Alessandro Mecocci,et al.  Outdoor scenes interpretation suitable for blind people navigation , 1995 .

[7]  Paolo Gamba,et al.  Vanishing point detection by a voting scheme , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[8]  Paolo Gamba,et al.  Scene interpretation by fusion of segment and region information , 1997, Image Vis. Comput..