On the Recognition of Arabic Characters Using Hough Transform Technique

Abstract

A method for the recognition of Arabic characters entered from an image scanner is presented. After the process of smoothing, midline skew correction, Arabic text is segmented into characters. Then features are extracted using Hough transform. Next, characters are classified using Dynamic programming matching technique and the extracted features. Characters classification is done in two steps: in the first one, the character’s main body is classified using DP matching technique, and features extracted in the Hough transform space. In the second one, simple topological features extracted from the geometry of the secondary parts are used by the topological classifier to completely recognize the characters. The topological features used to classify each type of the secondary part are the width, the height, and the number of the secondary part. Knowing both the main body of the character and the type of the secondary parts (if any), the character is completely identified. A recognition rate of about 95% was obtained.