Towards Directing Convolutional Neural Networks Using Computational Geometry Algorithms: Application to Handwritten Arabic Character Recognition

Abstract

Suppose we want to classify a query item Q with a classification model that consists of a large set of predefined classes L and suppose we have a knowledge that indicates to us that the target class of Q belongs to a small subset from L. Naturally, this filtering will improve the accuracy of any classifier, even random guessing. Based on this principle, this paper proposes a new classification approach using convolutional neural networks (CNN) and computational geometry (CG) algorithms. The approach is applied and tested on the recognition of isolated handwritten Arabic characters (IHAC). The main idea of the proposed approach is to direct CNN using a filtering layer, which reduces the set of possible classes for a query item. The rules of the relative neighborhood graph (RNG) and Gabriel's graph (GG) are combined for this purpose. The choice of RNG-GG was based on its great capacity to correctly reduce the list of possible classes. This capacity is measured by a new indicator that we call "the appearance rate". In recent years and due to strong data growth, CNNs have performed classification tasks very well. On the contrary, CG algorithms yield limited results in huge datasets and suffer from high computational time, but they generally reach high appearance rates and do not require any training phase. Consequently, the proposed approach uses an optimal architecture to exploit the advantages of the two techniques and overcome the computational time issue. Experiments carried out on the IFHCDB database have shown that the suggested approach outperforms a normal CNN and yield satisfactory results.

CNN-based Methods for Offline Arabic Handwriting Recognition: A Review

“…In the study of Elkhayati and Elkettani [58], CNN was hybridized with two computational geometry algorithms, namely the Relative Neighborhood Graph and Gabriel's Graph. These algorithms function as a filtering layer during inference, effectively reducing potential classes for a query item.…”

Ancient Character Recognition: A Novel Image Dataset of Shui Manuscript Characters and Classification Model

“…The authors in [5]- [9] presented studies on the recognition of Tamil characters using CNN The difference lies in the improvement of the CNN structure. The authors in [10]- [12] presented studies on Arabic character recogni-tion, which mainly use CNNs to combine feature selection methods in machine learning, and obtain satisfactory results for Arabic character recognition. In [13]- [15], the authors combined deep learning methods with traditional methods of feature extraction or some preprocessing methods to improve the recognition accuracy of handwritten Persian characters.…”

The effect of technology on Arabic language learning in higher education

“…Information technology is closely related to the world of education which not only serves as a supporting tool but can also be the main tool for success in the world of education [30]. such as the Internet in the world of education where the internet can make it easier for educators and students to get information [31]. For this reason, educators can be good at utilizing this internet-based technology to apply more effective and efficient learning [32].…”

“…This education is developing with fairly rapid development. It is also inseparable from the education of the Arabic language where its development is increasing every year [31].…”