Efficient Off-Line Cursive Handwriting Word Recognition
Abstract
In this paper, we present an off-line cursive word handwriting recognition methodology. This is based on a novel combination of two different modes of word image normalization and robust hybrid feature extraction. Word image normalization is performed by using as a reference point the geometric center of the word image as well as by placing the baseline of the word in the center of a rectangular box. Additionally, image pre-processing is performed in order to correct word skew, word slant as well as to normalize the stroke thickness. At a next step, two types of features are combined in a hybrid fashion. The first one, divides the word image into a set of zones and calculates the density of the character pixels in each zone. In the second type of features, we calculate the area that is formed from the projections of the upper and lower profile of the word. The performance of the proposed methodology is demonstrated after testing with the reference IAM cursive handwriting database.
Loading...