A large annotated corpus is critical to the development of robust optical character recognizers (OCRs). However, creation of annotated corpora is a tedious task. It is laborious, especially when the annotation is at the character level. In this paper, we propose an efficient hierarchical approach for annotation of large collection of printed document images. We align document images with independently keyed-in text. The method is model-driven and is intended to annotate large collection of documents, scanned in three different resolutions, at character level. We employ an XML representation for storage of the annotation information. APIs are provided for access at content level for easy use in training and evaluation of OCRs and other document understanding tasks. 1.

Arabic cursive optical character recognition (OCR) based on the dynamic time warping (DTW) algorithm provides simultaneously very interesting segmentation and recognition rates. However, the computing complexity of the DTW algorithm restricts its widespread utilization and its consideration at a commercial scale. Accelerating the DTW execution time has attracted many researchers and several solutions have already been proposed. These solutions are commonly based on very specialized processors and hardware architectures and as such they remain very expensive and not amenable to a large scale utilization. In a previous work, we found that loosely coupled architectures can indeed provide viable infrastructures to implement a distributed Arabic OCR. Our objective here is to allow the recognition of huge quantities of Arabic documents such as those of certain national libraries. Undoubtedly, enough processing power and storage capabilities are needed. In this paper, we proposed and used a peer-to-peer (P2P) architecture using the scientific research Tunisian grid (SRTG). Conducted experiments testify that our proposed architecture provides very adequate speedups of the DTWbased

Abstract: In spite of the diversity of printed Arabic optical character recognition products and proposals, the problem seems to be not yet well solved. The complex morphology and calligraphy of the Arabic writing on one hand and the use of some light approaches on the other hand are behind the poorness of these products. However, some strong proposed approaches didn’t find the opportunity to be commercialised because of generally their corresponding complex computing. The dynamic time warping algorithm is considered as one among these strong approaches. In fact, several studies and experiments have shown and confirmed that the printed Arabic optical character recognition based on dynamic time warping algorithm provides a very interesting recognition rate especially for large and huge vocabularies. One of the attractive sides of the dynamic time warping algorithm is its ability to recognize properly connected or cursive characters (words or sub words) without prior segmentation. Furthermore, this algorithm performs the recognition process from within a reference library of isolated characters and owns a very good immunity against noises. Unfortunately, the big amount of its computing during the recognition process makes its execution time very slow and, hence, restricts its utilization. Many researchers attempted to speedup the execution time of this algorithm. Unfortunately, the corresponding proposed solutions require generally specific high cost architectures. Loosely coupled architectures such as grapes or grid computing can provide enough power without additional cost to distribute the complexity of some greedy applications. Consequently, we report in this paper the performance analysis of an analytical and an experimental study of a distributed Arabic optical character recognition based

This paper proposes a recognition based approach to handwritten numeral string segmentation. We consider two classes: numeral strings segmented correctly or not. The feature vectors containing recognition information for numeral strings segmented correctly are assumed to be of the distribution of Gaussian mixture model (GMM). Based on this modeling, the recognition based segmentation is solved under the Max-Min posterior Pseudo-probabilities (MMP) framework of learning Bayesian classifiers. In the training phase, we use the MMP method to learn a posterior pseudo-probability measure function from positive samples and negative samples of numeral strings segmented correctly. In the process of recognition based segmentation, we generate all possible candidate segmentations of an input string through contour and profile analysis, and then compute the posterior pseudo-probabilities of being the numeral string segmented correctly for all the candidate segmentations. The candidate segmentation with the maximum posterior pseudo-probability is taken as the final result. The effectiveness of our approach is demonstrated by the experiments of numeral string segmentation and recognition on the NIST SD19 database. 1.

In this paper, a novel technique is presented for creating unbiased training and testing sets based on performing random walks using a context free grammar. We describe a study conducted to generate a corpus of over 4500 hand-drawn mathematical expressions from 20 subjects. Finally, details are given for an automated system which generates high quality ground-truth data from the hand-drawn expressions. The techniques presented in this paper are illustrated through the creation of a ground-truthed corpus of mathematical expressions, but they are applicable to any sketching domain that can be described by a formal grammar.