Showing papers on "Intelligent word recognition published in 2014"
Posted Content•
TL;DR: This work presents a framework for the recognition of natural scene text that does not require any human-labelled data, and performs word recognition on the whole image holistically, departing from the character based recognition systems of the past.
Abstract: In this work we present a framework for the recognition of natural scene text. Our framework does not require any human-labelled data, and performs word recognition on the whole image holistically, departing from the character based recognition systems of the past. The deep neural network models at the centre of this framework are trained solely on data produced by a synthetic text generation engine -- synthetic data that is highly realistic and sufficient to replace real data, giving us infinite amounts of training data. This excess of data exposes new possibilities for word recognition models, and here we consider three models, each one "reading" words in a different way: via 90k-way dictionary encoding, character sequence encoding, and bag-of-N-grams encoding. In the scenarios of language based and completely unconstrained text recognition we greatly improve upon state-of-the-art performance on standard datasets, using our fast, simple machinery and requiring zero data-acquisition costs.
875 citations
TL;DR: An approach in which both word images and text strings are embedded in a common vectorial subspace, allowing one to cast recognition and retrieval tasks as a nearest neighbor problem and is very fast to compute and, especially, to compare.
Abstract: This paper addresses the problems of word spotting and word recognition on images. In word spotting, the goal is to find all instances of a query word in a dataset of images. In recognition, the goal is to recognize the content of the word image, usually aided by a dictionary or lexicon. We describe an approach in which both word images and text strings are embedded in a common vectorial subspace. This is achieved by a combination of label embedding and attributes learning, and a common subspace regression. In this subspace, images and strings that represent the same word are close together, allowing one to cast recognition and retrieval tasks as a nearest neighbor problem. Contrary to most other existing methods, our representation has a fixed length, is low dimensional, and is very fast to compute and, especially, to compare. We test our approach on four public datasets of both handwritten documents and natural images showing results comparable or better than the state-of-the-art on spotting and recognition tasks.
522 citations
01 Mar 2014
TL;DR: The Urdu, Pushto, and Sindhi languages are discussed, with the emphasis being on the Nasta'liq and Naskh scripts, with an emphasis on the preprocessing, segmentation, feature extraction, classification, and recognition in OCR.
Abstract: We survey the optical character recognition (OCR) literature with reference to the Urdu-like cursive scripts. In particular, the Urdu, Pushto, and Sindhi languages are discussed, with the emphasis being on the Nasta'liq and Naskh scripts. Before detaining the OCR works, the peculiarities of the Urdu-like scripts are outlined, which are followed by the presentation of the available text image databases. For the sake of clarity, the various attempts are grouped into three parts, namely: (a) printed, (b) handwritten, and (c) online character recognition. Within each part, the works are analyzed par rapport a typical OCR pipeline with an emphasis on the preprocessing, segmentation, feature extraction, classification, and recognition. HighlightsA literature review of the Nasta'liq and Naskh cursive script OCR.The peculiarities and challenges are described a priori.Printed, handwritten and online OCR efforts are being explored.Analyses based on the stages of a typical OCR pipeline.
121 citations
15 Dec 2014
TL;DR: A system based on recurrent neural networks and weighted finite state transducers was used both for printed and handwritten recognition, in French, English and Arabic, for multi-lingual text recognition.
Abstract: This paper describes the system submitted by A2iA to the second Maurdor evaluation for multi-lingual text recogni- tion. A system based on recurrent neural networks and weighted finite state transducers was used both for printed and handwritten recognition, in French, English and Arabic. To cope with the difficulty of the documents, multiple text line segmentations were considered. An automatic procedure was used to prepare annotated text lines needed for the training of the neural network. Language models were used to decode sequences of characters or words for French and English and also sequences of part-of- arabic words (PAWs) in case of Arabic. This system scored first at the second Maurdor evaluation for both printed and handwritten text recognition in French, English and Arabic.
58 citations
27 Mar 2014
TL;DR: A recently proposed machine learning approach called deep learning for Bangla hand written character recognition, with a focus on automatic learning of good representations, employs the deep belief network (DBN) that takes the raw character images as input and learning proceeds in two steps - an unsupervised feature learning followed by a supervised fine tuning of the network parameters.
Abstract: Recognition of Bangla handwritten characters is a difficult but important task for various emerging applications. For better recognition performance, good feature representation of the character images is a primary requirement. In this study, we investigate a recently proposed machine learning approach called deep learning [1] for Bangla hand written character recognition, with a focus on automatic learning of good representations. This approach differs from the traditional methods of preprocessing the characters for constructing the handcrafted features such as loops and strokes. Among different deep learning structures, we employ the deep belief network (DBN) that takes the raw character images as input and learning proceeds in two steps - an unsupervised feature learning followed by a supervised fine tuning of the network parameters. Unlike traditional neural networks, the DBN is a probabilistic generative model, i.e., we can generate samples from the model and it can fit both the semi-supervised and supervised learning settings. We demonstrate the advantages of unsupervised feature learning through the experimental studies carried on the Bangla basic characters and numerals dataset collected from the Indian Statistical Institute.
57 citations
01 Oct 2014
TL;DR: This paper proposes a shared-hidden-layer deep convolutional neural network (SHL-CNN) for image character recognition, and discusses several issues including architecture of the network, training of thenetwork, and the effectiveness of the learned SHL-CNN.
Abstract: This paper proposes a shared-hidden-layer deep convolutional neural network (SHL-CNN) for image character recognition. In SHL-CNN, the hidden layers are made common across characters from different languages, performing a universal feature extraction process that aims at learning common character traits existed in different languages such as strokes, while the final softmax layer is made language dependent, trained based on characters from the destination language only. This paper is the first attempt to introduce the SHL-CNN framework to image character recognition. Under the SHL-CNN framework, we discuss several issues including architecture of the network, training of the network, from which a suitable SHL-CNN model for image character recognition is empirically learned. The effectiveness of the learned SHL-CNN is verified on both English and Chinese image character recognition tasks, showing that the SHL-CNN can reduce recognition errors by 16–30% relatively compared with models trained by characters of only one language using conventional CNN, and by 35.7% relatively compared with state-of-the-art methods. In addition, the shared hidden layers learned are also useful for unseen image character recognition tasks.
56 citations
24 Aug 2014
TL;DR: The Handwritten Online Musical Symbols (HOMUS) dataset is presented, which consists of 15200 samples of 32 types of musical symbols from 100 different musicians and can establish a binding point in the field of recognition of online handwritten music notation and serve as a baseline for future developments.
Abstract: A profitable way of digitizing a new musical composition is by using a pen-based (online) system, in which the score is created with the sole effort of the composition itself. However, the development of such systems is still largely unexplored. Some studies have been carried out but the use of particular little datasets has led to avoid objective comparisons between different approaches. To solve this situation, this work presents the Handwritten Online Musical Symbols (HOMUS) dataset, which consists of 15200 samples of 32 types of musical symbols from 100 different musicians. Several alternatives of recognition for the two modalities -online, using the strokes drawn by the pen, and offline, using the image generated after drawing the symbol- are also presented. Some experiments are included aimed to draw main conclusions about the recognition of these data. It is expected that this work can establish a binding point in the field of recognition of online handwritten music notation and serve as a baseline for future developments.
53 citations
01 Dec 2014
TL;DR: This paper presents a technique based on Multi Layer Perceptron (MLP) Neural Network model that detects graphical symbols by identifying lines and characters from the image and analyzes the symbols by training the network using feed forward topology for a set of desired unicode characters.
Abstract: Machine vision researchers are working on the area of recognition of handwritten or printed text from scanned images for the purpose of digitizing documents and for reducing the errorless data entry cost. The classic difficulty of being able to correctly recognize language symbols is the complexity and the irregularity among the pictorial representation of characters due to variation in writing styles, size of symbols etc. Character recognition process depends on, how the input data is given to the system. Input data may be categorized as Online data or Offline data. Both the forms of data input have their own issues. In this paper, we are focusing on the Offline Gurmukhi character recognition from text image. There are lot of complexities associated with Gurmukhi Script. In this paper, we present a technique based on Multi Layer Perceptron (MLP) Neural Network model. Here we consider isolated handwritten Gurmukhi characters for recognition. MLP is used because it uses generalized delta learning rules and easily gets trained in less number of iterations. The proposed method in this paper detect graphical symbols by identifying lines and characters from the image. After that it analyzes the symbols by training the network using feed forward topology for a set of desired unicode characters. We achieve the performance rate of proposed system maximum up to 98.96% for recognition of symbols by using MLP neural network.
46 citations
23 May 2014
TL;DR: This work applied Sparse Representation Classifier on the image zone density, an image domain statistical feature extracted from the character image, to classify the Bangla numerals and demonstrates an excellent accuracy of 94% on the off-line handwritten Bangla numeral database CMATERdb 3.1.
Abstract: We present a framework for handwritten Bangla digit recognition using Sparse Representation Classifier. The classifier assumes that a test sample can be represented as a linear combination of the train samples from its native class. Hence, a test sample can be represented using a dictionary constructed from the train samples. The most sparse linear representation of the test sample in terms of this dictionary can be efficiently computed through l 1 -minimization, and can be exploited to classify the test sample. We applied Sparse Representation Classifier on the image zone density, an image domain statistical feature extracted from the character image, to classify the Bangla numerals. This is a novel approach for Bangla Optical Character Recognition, and demonstrates an excellent accuracy of 94% on the off-line handwritten Bangla numeral database CMATERdb 3.1.1. This result is promising, and should be investigated further.
41 citations
TL;DR: A novel scene text-recognition method integrating structure-guided character detection and linguistic knowledge is proposed, using part-based tree structure to model each category of characters so as to detect and recognize characters simultaneously.
Abstract: Scene text recognition has inspired great interests from the computer vision community in recent years. In this paper, we propose a novel scene text-recognition method integrating structure-guided character detection and linguistic knowledge. We use part-based tree structure to model each category of characters so as to detect and recognize characters simultaneously. Since the character models make use of both the local appearance and global structure informations, the detection results are more reliable. For word recognition, we combine the detection scores and language model into the posterior probability of character sequence from the Bayesian decision view. The final word-recognition result is obtained by maximizing the character sequence posterior probability via Viterbi algorithm. Experimental results on a range of challenging public data sets (ICDAR 2003, ICDAR 2011, SVT) demonstrate that the proposed method achieves state-of-the-art performance both for character detection and word recognition.
36 citations
10 Jul 2014
TL;DR: A review of the OCR history and the various techniques used for OCR development in the chronological order is being done.
Abstract: Many researches are going on in the field of optical character recognition (OCR) for the last few decades and a lot of articles have been published. Also a large number of OCR is available commercially. In this literature a review of the OCR history and the various techniques used for OCR development in the chronological order is being done.
TL;DR: The novelty of the approach lies in the formulation of appropriate rules of character decomposition for segmenting the character skeleton into stroke segments and then grouping them for extraction of meaningful shape components.
24 Aug 2014
TL;DR: The performance of several well-known offline features for handwritten math symbol recognition are assessed and a novel set of features based on polar histograms and the vertical repositioning method for feature extraction is tested.
Abstract: In mathematical expression recognition, symbol classification is a crucial step. Numerous approaches for recognizing handwritten math symbols have been published, but most of them are either an online approach or a hybrid approach. There is an absence of a study focused on offline features for handwritten math symbol recognition. Furthermore, many papers provide results difficult to compare. In this paper we assess the performance of several well-known offline features for this task. We also test a novel set of features based on polar histograms and the vertical repositioning method for feature extraction. Finally, we report and analyze the results of several experiments using recurrent neural networks on a large public database of online handwritten math expressions. The combination of online and offline features significantly improved the recognition rate.
19 Dec 2014
TL;DR: A novel technique to recognize handwritten Bangla word is proposed using Histograms of Oriented Gradients to represent each word sample at the feature space and a neural network based classifier is applied to classify the word images.
Abstract: The holistic approaches for handwritten word recognition treat the words as single, indivisible entity and attempt to recognize words from their overall shape. In the present work, a novel technique to recognize handwritten Bangla word is proposed. Histograms of Oriented Gradients (HOG) are used as the feature set to represent each word sample at the feature space and a neural network based classifier is applied to classify the word images. On the basis of the HOG feature set, the performance achieved by the technique on a small dataset is quite satisfactory.
14 Nov 2014
TL;DR: A comparison among 5 well known classifiers is carried out in terms of their accuracies to select the suitable classifier for evaluating the present work and a neural network based classifier is chosen.
Abstract: In the present work, a holistic word recognition technique is proposed for the recognition of the handwritten Bangla words. Holistic word recognition technique assumes a word as a single and indivisible entity and extracts features from the entire word to recognize it. In this work, a set of elliptical features is extracted from handwritten word images to represent them in the feature space. Then, a comparison among 5 well known classifiers is carried out in terms of their accuracies to select the suitable classifier for evaluating the present work. Based on that, finally, a neural network based classifier is chosen for the recognition task. Using the elliptical features, the proposed system provides a satisfactory result on a small dataset.
15 Dec 2014
TL;DR: A preliminary experiment is performed on a dataset of 10,120 Bangla handwritten words and it is found that the proposed approach outperforms the custom way of HMM based recognition.
Abstract: This paper presents a novel approach for offline Bangla (Bengali) handwritten word recognition by Hidden Markov Model (HMM). Due to the presence of complex features such as headline, vowels, modifiers, etc., character segmentation in Bangla script is not easy. Also, the position of vowels and compound characters make the segmentation task of words into characters very complex. To take care of these problems we propose a novel method considering a zone-wise break up of words and next perform HMM based recognition. In particular, the word image is segmented into 3 zones, upper, middle and lower, respectively. The components in middle zone are modeled using HMM. By this zone segmentation approach we reduce the number of distinct component classes compared to total number of classes in Bangla character set. Once the middle zone portion is recognized, HMM based forced alignment is applied in this zone to mark the boundaries of individual components. The segmentation paths are extended later to other zones. Next, the residue components, if any, in upper and lower zones in their respective boundary are combined to achieve the final word level recognition. We have performed a preliminary experiment on a dataset of 10,120 Bangla handwritten words and found that the proposed approach outperforms the custom way of HMM based recognition.
01 Jan 2014
TL;DR: In this chapter, an overview of the transcription of written text images is given and the steps along the processing chain from the text line image to the final output are explained, starting with image normalization and feature representation.
Abstract: The transcription of written text images is one of the most challenging tasks in document analysis since it has to cope with the variability and ambiguity encountered in handwritten data. Only in a very restricted setting, as encountered in postal addresses or bank checks, transcription works well enough for commercial applications. In the case of unconstrained modern handwritten text, recent advances have pushed the field towards becoming interesting for practical applications. For historic data, however, recognition accuracies are still far too low for automatic systems. Instead, recent efforts aim at interactive solutions in which the computer merely assists an expert creating a transcription. In this chapter, an overview of the field is given and the steps along the processing chain from the text line image to the final output are explained, starting with image normalization and feature representation. Two recognition approaches, based on hidden Markov models and neural networks, are introduced in more detail. Finally, databases and software toolkits are presented, and hints to further material are provided.
01 Oct 2014
TL;DR: Zonal based feature extraction is used in the present proposed method and using this Zoning method the recognition accuracy is found to be 78%.
Abstract: Character recognition is one of the oldest applications of pattern recognition. Recognizing Hand-Written Characters (HWC) is an effortless task for humans, but for a computer it is a difficult job. Research in character recognition is very popular for various potential applications such as in banks, post offices, defense organizations, reading aid for the blind, library automation, language processing and multi-media design. Optical Character Recognition (OCR) is based on optical mechanism which consists of a machine to recognize scanned and digitized character automatically. Automatic recognition of handwritten text can be done either Offline or Online. Offline handwritten recognition is the task of recognizing the image of a hand written text, in contrast to Online recognition where the dynamic characteristics of the writing are available and recorded while the scriber is writing on a special screen with a pen/stylus made for this application. Zonal based feature extraction is used in the present proposed method. The character image is divided into predefined number of zones and a statistical feature is computed from each of these zones. Usually, this feature is based on the pixels contained in that zone. The gray values of the pixels in that selected zone are summed up to form a feature for that zone in that image. The features of all the zones in the image form a feature vector which is used for handwritten character recognition. In this work, using this Zoning method the recognition accuracy is found to be 78%.
14 Apr 2014
TL;DR: This work proposes a hybrid feature extraction technique using Kirsch gradient operator and curvature properties of handwritten numerals, followed by a feature dimension reduction using Principal Component Analysis (PCA), which uses Modified Quadratic Discriminant Function (MQDF), Discriminative Learning quadratic discriminant function (DLQDF) classifiers as they provide high accuracy of recognition and compares both the classifier performances.
Abstract: Unconstrained handwritten character recognition is a major research area where there is a lot of scope for improving accuracy. There are many statistical, structural feature extraction techniques being proposed for different languages. Many classifier models are combined with these features to obtain high recognition rates. There still exists a gap between the recognition accuracy of printed characters and unconstrained handwritten scripts. Odia is a popular and classical language of the eastern part of India. Though the research in Optical Character Recognition (OCR) has advanced in other Indian languages such as Devanagari and Bangla, not much attention has been given to Odia character recognition. We propose a hybrid feature extraction technique using Kirsch gradient operator and curvature properties of handwritten numerals, followed by a feature dimension reduction using Principal Component Analysis (PCA). We use Modified Quadratic Discriminant Function (MQDF), Discriminative Learning Quadratic Discriminant Function (DLQDF) classifiers as they provide high accuracy of recognition and compare both the classifier performances. We verify our results using the Odia numerals database of ISI Kolkata. The recognition accuracy for Odia numerals with our proposed approach is found to be 98.5%.
07 Apr 2014
TL;DR: A modification of a finite-state-transducer (fst) n-gram that enables the creation of a static transducer, i.e. when it is not possible to perform on-demand composition, is presented and it is shown that this model is competitive with state-of-the-art solutions.
Abstract: Hybrid statistical grammars both at word and character levels can be used to perform open-vocabulary recognition. This is usually done by allowing the special symbol for unknown-word in the word-level grammar and dynamically replacing it by a (long) n-gramat character-level, as the full transducer does not fit in the memory of most current computers. We present a modification of a finite-state-transducer (fst) n-gram that enables the creation of a static transducer, i.e. when it is not possible to perform on-demand composition. By combining paths in the "LG" transducer (composition of lexicon and n-gram)making it over-generative with respect to the n-grams observed in the corpus, it is possible to reduce the number of actual occurrences of the character-level grammar, the resulting transducer fits the memory of practical machines. We evaluate this model for handwriting recognition using the RIMES and the IAM dabases. We study its effect on the vocabulary size and show that this model is competitive with state-of-the-art solutions.
07 Apr 2014
TL;DR: A web based OCR system which follows a unified architecture for seven Indian languages, is robust against popular degradations, follows a segmentation free approach, addresses the UNICODE re-ordering issues, and can enable continuous learning with user inputs and feedbacks is proposed.
Abstract: The current Optical Character Recognition OCR systems for Indic scripts are not robust enough for recognizing arbitrary collection of printed documents. Reasons for this limitation includes the lack of resources (e.g. not enough examples with natural variations, lack of documentation available about the possible font/style variations) and the architecture which necessitates hard segmentation of word images followed by an isolated symbol recognition. Variations among scripts, latent symbol to UNICODE conversion rules, non-standard fonts/styles and large degradations are some of the major reasons for the unavailability of robust solutions. In this paper, we propose a web based OCR system which (i) follows a unified architecture for seven Indian languages, (ii) is robust against popular degradations, (iii) follows a segmentation free approach, (iv) addresses the UNICODE re-ordering issues, and (v) can enable continuous learning with user inputs and feedbacks. Our system is designed to aid the continuous learning while being usable i.e., we capture the user inputs (say example images) for further improving the OCRs. We use the popular BLSTM based transcription scheme to achieve our target. This also enables incremental training and refinement in a seamless manner. We report superior accuracy rates in comparison with the available OCRs for the seven Indian languages.
01 Oct 2014
TL;DR: A new strategy for the segmentation of conjuncts, and overlapping characters in Devanagari script on Hindi language is shown, focused around Cluster Detection technique and gives 95% correctness for segmenting touching, conjunct characters and 88% effectiveness for overlapping characters.
Abstract: Optical Character Recognition alludes to the methodology of taking images or photos of letters or typewritten content and changing over them into information that a machine can easily interpret, e.g. organizations and libraries taking physical duplicates of books, magazines, or other old printed material and utilizing OCR to put them into computers. Segmentation is the indispensable and most difficult part of OCR process, and it gets to be additionally difficult with handwritten text due to varieties in writing styles and presence of abnormalities. This paper shows a new strategy for the segmentation of conjuncts, and overlapping characters in Devanagari script on Hindi language. The proposed algorithm is focused around Cluster Detection technique and gives 95% correctness for segmenting touching, conjunct characters and 88% effectiveness for overlapping characters.
01 Aug 2014
TL;DR: An efficient Arabic handwritten characters recognizer aimed at facilitating real-time handwritten script analysis tasks and achieving fast yet accurate recognition results in a dictionary-free environment is proposed.
Abstract: Delaying the analysis launch until the completion of the handwritten word scribing, restricts on-line recognition systems to meet the highly responsiveness demands expected from such applications, and prevents implementing advanced features of input typing such as automatic word completion and real-time automatic spelling. This paper proposes an efficient Arabic handwritten characters recognizer aimed at facilitating real-time handwritten script analysis tasks. The fast classification is enabled by employing an efficient embedding of the feature vectors into a normed wavelet coefficients domain in which the Earth Movers Distance metric is approximated using the Manhattan distance. A sub-linear time character classification is achieved by utilizing metric indexing techniques. Using the results of the top ranked shapes of each predicted character, a list of candidate shapes of Arabic word parts is generated in a filter and refine approach to enable fast yet accurate recognition results in a dictionary-free environment. The system was trained and tested on characters and word parts extracted from the ADAB database, and promising accuracy and performance results were achieved.
15 Dec 2014
TL;DR: The recent study of a novel combination of two feature vectors for holistic recognition of offline handwritten word images shows sharp improvement in recognition accuracy over the use of any of the individual feature representation schemes.
Abstract: In this article, we describe our recent study of a novel combination of two feature vectors for holistic recognition of offline handwritten word images. In the literature, both contour and skeleton based feature representations have been studied for offline handwriting recognition purpose. However, to the best of our knowledge, there is no such study in which combination of the two feature representations have been considered for the purpose. In the proposed recognition scheme, we use multiclass SVM as the classifier. We have implemented the proposed approach for holistic recognition of Devanagari handwritten town names and tested its performance on a large handwritten word sample database of 100 Indian town names written in Devanagari. Experimental results show sharp improvement in recognition accuracy over the use of any of the individual feature representation schemes. The proposed approach is script independent and can be used for development of a holistic handwritten word image recognition of any script.
21 Jul 2014
TL;DR: An intelligent feature-based technique for word-level script identification in multi-script handwritten document pages for multilingual optical character recognition (OCR) system and statistical significance tests declare MLP to be the best performing one.
Abstract: Script identification for handwritten document image is an open document analysis problem especially for multilingual optical character recognition (OCR) system. To design the OCR system for multi-script document pages, it is essential to recognise different scripts before running a particular OCR system of a script. The present work reports an intelligent feature-based technique for word-level script identification in multi-script handwritten document pages. At first, the text lines and then the words are extracted from the document pages. A set of 39 distinctive features have been designed of which eight features are topological and the rest (31) are based on convex hull for each word image. For selection of a suitable classifier, performances of multiple classifiers are evaluated with the designed feature set on multiple subsets of freely available database CMATERdb1.5.1 (http://www.code.google.com/p/cmaterdb), which comprises of 150 handwritten document pages containing both Devnagari and Roman script words. Statistical significance tests on these performance measures declare MLP to be the best performing one. The overall word-level script identification accuracy with MLP classifier on the said database is observed as 99.74%.
06 Oct 2014
TL;DR: A novel efficient approach for the recognition of off-line Arabic handwritten characters based on novel preprocessing operations, structural statistical and topological features from the main body of the character and also from the secondary components is proposed.
Abstract: There are many difficulties facing a handwritten Arabic recognition system such as unlimited variation in character shapes. This paper describes a new method for handwritten Arabic character recognition. We propose a novel efficient approach for the recognition of off-line Arabic handwritten characters. The approach is based on novel preprocessing operations, structural statistical and topological features from the main body of the character and also from the secondary components. Evaluation of the importance and accuracy of the selected features was made. Our method based on the selected features and the system was built, trained and tested by CENPRMI dataset. We used SVM (RBF) and KNN for classification to find the recognition accuracy. The proposed algorithm obtained promising results in terms of accuracy; with recognition rates of 89.2% for SVM. Compared with other related works and also our recently published work we find that our result is the highest among them.
17 Dec 2014
TL;DR: Back propagation neural network classifier for single stroke characters in initial half form is designed with overall accuracy of 91.3% and stored as a binary file containing x & y coordinates and pressure values rather than in an image form to reduce complexity of the recognition problem.
Abstract: Necessity of unfolding the enticing field of handwritten character recognition is revealed with the mushroom growth of portable devices. Effective human machine interaction insists the development of a reliable and efficient online handwritten character recognition system. The quest becomes more challenging when it involves Urdu script based languages especially written in Nastalique font. Urdu, in Nastalique style, is a context sensitive and a highly cursive language. Difficulty arises in this style of writing as the shape of a character depends whether it is written in isolated, initial, medial or terminal position in a word. In this paper, online recognition of Urdu characters in their initial half form have been studied. Data is collected using a pen-tablet and the signal is stored as a binary file containing x a y coordinates and pressure values rather than in an image form to reduce complexity of the recognition problem. Wavelets transform is applied to analyze the character signal. Back propagation neural network classifier for single stroke characters in initial half form is designed with overall accuracy of 91.3%.
04 May 2014
TL;DR: Grayscale based Chinese Image Text Recognition (gCITR), implements the recognition directly on grayscale pixels via the following steps: image text over-segmentation, building recognition graph, Chinese character recognition and beam search determination.
Abstract: This paper presents a novel scheme for Chinese text recognition in images and videos. It's different from traditional paradigms that binarize text images, fed the binarized text to an OCR engine and get the recognized results. The proposed scheme, named grayscale based Chinese Image Text Recognition (gCITR), implements the recognition directly on grayscale pixels via the following steps: image text over-segmentation, building recognition graph, Chinese character recognition and beam search determination. The advantages of gCITR lie in: (1) it does not heavily rely on the performance of binarization, which is not robust in practical and thus severely affects the performance of OCR, (2) grayscale image retains more information of the text thus facilitates the recognition. Experimental results on text from 13 TV news videos demonstrate the effectiveness of the proposed gCITR, from which significant performance gains are observed.
Posted Content•
TL;DR: New methods for handwritten Arabic character recognition which is based on novel preprocessing operations including different kinds of noise removal also different kind of features like structural, Statistical and Morphological features from the main body of the character and also from the secondary components are proposed.
Abstract: There are many difficulties facing a handwritten Arabic recognition system such as unlimited variation in human handwriting, similarities of distinct character shapes, interconnections of neighbouring characters and their position in the word. The typical Optical Character Recognition (OCR) systems are based mainly on three stages, preprocessing, features extraction and recognition. This paper proposes new methods for handwritten Arabic character recognition which is based on novel preprocessing operations including different kinds of noise removal also different kind of features like structural, Statistical and Morphological features from the main body of the character and also from the secondary components. Evaluation of the accuracy of the selected features is made. The system was trained and tested by back propagation neural network with CENPRMI dataset. The proposed algorithm obtained promising results as it is able to recognize 88% of our test set accurately. In Comparable with other related works we find that our result is the highest among other published works.
01 Dec 2014
TL;DR: This paper describes the handwritten Meitei Mayek (Manipuri script) alphabets recognition (HMMAR) using a neural network approach and focuses on the process of character segmentation from a whole document i.e. isolating a single character image from a complete scanned document.
Abstract: Handwritten character recognition is a part of optical character (OCR) system. OCR can be applied to both printed text and handwritten documents. In this paper we discussed the handwritten character recognition of Meitei Mayek (Manipuri script). Although OCR has been studied and developed for many Indian script very few works have been reported so far for Meitei-Mayek. This paper describes the handwritten Meitei Mayek (Manipuri script) alphabets recognition (HMMAR) using a neural network approach. The alphabet database is pre-processed and the extracted feature is sent to a neural network system for training. The trained neural network is further tested and performance analysis is observed. The emphasis is given on the process of character segmentation from a whole document i.e. isolating a single character image from a complete scanned document.