Title :
An Efficient Slope and Slant Correction Technique for Off-Line Handwritten Text Word
Author :
Gupta, Jija Das ; Chanda, Bhabatosh
Author_Institution :
Appl. Opt. &Photonics, Univ. of Calcutta, Kolkata, India
Abstract :
Slope and Slant correction of off-line hand written text words is one of the important preprocessing step because it reduces the variations in person based writing styles and, thus, improves word recognition accuracy. This paper presents a novel slope and slant correction technique. We have developed a database and tested our algorithm on it. We compare our result with well-known slope and slant correction techniques. Comparative study reveals that our method is faster and more accurate.
Keywords :
handwritten character recognition; text detection; off-line handwritten text word; slant correction technique; slope correction technique; word recognition accuracy; Accuracy; Databases; Entropy; Estimation; Handwriting recognition; Text recognition; Writing; connected-component; hough transform; slant; slope; stroke-width;
Conference_Titel :
Emerging Applications of Information Technology (EAIT), 2014 Fourth International Conference of
DOI :
10.1109/EAIT.2014.19
