Intelligent system for Arabic character recognition

Abstract

In this work, a new system for Arabic letter recognition is designed and implemented. New approaches for segmentation, processing, classification and hence recognition of characters and scripts are shown. The research concentrates on two important subjects: First, segmentation on the basis of word histogram and baseline estimation -a convenient algorithm is worked out for this aim. Second, the process of feature extraction to find the most useful points is implemented upon the given algorithm. Feature coding is executed as a string of eight digits through two counterclockwise passes. The code is filtered up provided with eight basic pairs. The filtered code goes through processing to form an array of 9*9 elements, in addition to an array of 2*2 elements determined to resemble the four parts of the extracted character image. The 85 obtained elements are the input to a Backpropagation Neural Network used for classification purposes. A 98.7% rate of recognition is achieved for Arabic character classification. Results have proved high recognition of Arabic letters for varieties of fonts and sizes. They have also assured that computing time is negligible with very small errors.

A Survey on Arabic Handwritten Character Recognition

“…In [3], found a popular defined size, skeletonization and thinning. The technique of thinning will convert the binary image of character into one pixel thick image used by [4]. Thresholding in this technique includes converting gray input images into document images of binary type for getting good features from the image [5], skew detection and correction [6], noise removal; the basic noise objective removal which is to eliminate all bit-patterns which are unwanted those did not have any value in the result.…”

“…In [23], used the longest spike which represented the baseline. In [4], used start-point also end-point of a character features and branch. Wavelet transforms features used by [32].…”

“…In [8], authors presented sequenced networks to recognition of the characters with 120 features for each character's image. In [4], the neural network algorithm of back-propagation was also used with histogram features. In [22], they have designed radial-basis neural network classifier, investigated and compared among results of four different artificial neural network models.…”

Recognizing Cursive Typewritten Text Using Segmentation‐Free System

“…Feature extraction tackles the obstacle of cursiveness of Arabic in twofold: the global approach and the analytical approach. While global approach treats the word as a whole, extracts features from the unsegmented word, and then compares those features to a model [ 7 , 8 ], analytical approach decomposes the word into smaller units called glyphs [ 9 ]. Glyphs may or may not correspond to characters, although previous research has confirmed the difficulties in attempting to segment Arabic words into individual characters [ 10 ].…”