Title :
An Adaptive-Angle-Method for Uyghur Location
Author :
Deng Rengreng ; Halidan, A.
Author_Institution :
Dept. of Electr. Eng. Coll., Xinjiang Univ., Urumqi, China
Abstract :
According to the limitation of too many text location methods only be used for horizontal texts by now, the paper adopts an adaptive-angle-method for text image location. Firstly, we do the first-location for the original image, and by means of projection with a threshold and projection features on pixels, starting from the Uyghur baseline characteristics, and combined with the proposed adaptive method to choose the angle for the smallest text area, so we can make the second-location. The method can adapt to horizontal and inclined text. A large number of tests show that the adaptive method have great practicality and high accuracy.
Keywords :
document image processing; text analysis; Uyghur baseline characteristics; Uyghur location; adaptive-angle-method; horizontal texts; inclined text; original image first-location; original image second-location; projection features; text image location method; threshold features; Accuracy; Classification algorithms; Educational institutions; Feature extraction; Image edge detection; Licenses; Noise;
Conference_Titel :
Engineering and Technology (S-CET), 2012 Spring Congress on
Conference_Location :
Xian
Print_ISBN :
978-1-4577-1965-3
DOI :
10.1109/SCET.2012.6341964