Off-line Recognition Research Articles

Directed connection measurement for evaluating reconstructed stroke sequence in handwriting images

Generally speaking, the performance of on-line handwriting recognition systems is better than that of off-line handwriting recognition systems. Many researchers then started to extract/estimate on-line/temporal information from a static handwritten image to improve the performance of off-line recognition systems. More specifically, various algorithms have been developed to extract the stroke sequence, which is an important temporal information, from a static handwritten image in the last decade. However, to the best of our knowledge, there are no methods to evaluate the performance of these algorithms. In view of this limitation, this paper presents a directed connection measurement for evaluating the performance of these methods. The measurement is designed in terms of direction and connection. The direction measurement is based on the consecutive arrangements of the items, while the connection measurement is based on both the number of disconnections and the Feigin and Cohen model (FCM) in ranking analysis. We replace the Kendall's distance, the key metric component in FCM, by a new connection metric. We have also proved that the new direction and connection metrics satisfy all axioms for being a valid metric. The two correlations are combined to form a directed connection correlation, which is applicable in comparing the performance of different stroke sequence recovery algorithms.

Offline recognition of unconstrained handwritten texts using HMMs and statistical language models

This paper presents a system for the offline recognition of large vocabulary unconstrained handwritten texts. The only assumption made about the data is that it is written in English. This allows the application of Statistical Language Models in order to improve the performance of our system. Several experiments have been performed using both single and multiple writer data. Lexica of variable size (from 10,000 to 50,000 words) have been used. The use of language models is shown to improve the accuracy of the system (when the lexicon contains 50,000 words, the error rate is reduced by approximately 50 percent for single writer data and by approximately 25 percent for multiple writer data). Our approach is described in detail and compared with other methods presented in the literature to deal with the same problem. An experimental setup to correctly deal with unconstrained text recognition is proposed.

Off-line recognition of handwritten Arabic words using multiple hidden Markov models

A complete scheme for unconstrained Arabic handwritten word recognition based on a multiple hidden Markov models (HMM) is presented. The overall engine of this combination of a global feature scheme with a HMM module, is a system able to classify Arabic-handwritten words. The system first removes some of the variation in the images. Next, it codes the skeleton and edge of the word such that features are extracted. Then, a rule-based classifier is used as a global recognition engine. Finally, for each group, the HMM approach is used for trial classification. The output is a word in the lexicon

Recognising handwritten Arabic manuscripts using a single hidden Markov model

This paper presents a new method on off-line recognition of handwritten Arabic script. The method does not require segmentation into characters, and is applied to cursive Arabic script, where ligatures, overlaps and style variation pose challenges to the recognition system. The method trains a single hidden Markov model (HMM) with the structural features extracted from the manuscript words. The HMM is composed of multiple character models where each model represents one letter from the alphabet. The performance of the proposed method is assessed using samples extracted from a historical handwritten manuscript.

An Approach to Off-Line Handwritten Chinese Character Recognition Based on Hierarchical Radical Decomposition

Off-line recognition of handwritten Chinese characters is of considerable practical importance as well as representing a very hard pattern recognition problem. A popular approach is to decompose characters into their component or ‘primitive’ parts – most usually strokes. Here, however, we take the less usual approach of decomposing into radicals. Active shape modelling is applied and developed into active radical modelling. In training, 60 examples of each radical are represented by ‘landmark’ points, labelled semi-automatically, with radicals in different characteristic positions treated as distinctly different radicals. Principal component analysis then captures the main variation around the mean radical. In recognition, the dynamic tunnelling algorithm is incorporated with gradient descent to search for optimal shape parameters in terms of chamfer distance minimisation. Although prior landmark labelling is time-consuming and gradient descent search during recognition is computationally expensive, the method is theoretically well motivated, incorporates prior knowledge about the structure of Chinese characters in an appropriate way, and avoids problems implicit in stroke extraction. Experiments are conducted on 280,000 loosely constrained characters from 200 writers. There are 98 different categories of radical included in 1400 character categories, and approximately 590,000 radicals in total. The matching rate on this large test set is 94.2% radicals correct (writer-independent), greatly superior to existing radical approaches. Assuming character composition to be a Markov process in which up to four radicals are combined in some assumed sequential order, we can recognise complete, hierarchically composed characters using the Viterbi algorithm. This results in a character recognition rate of 92.6%.

Vector-to-Image Transformation of Character Patterns for On-line and Off-line Recognition

This paper proposes a method to generate realistic character images from on-line patterns. From the pen trajectory of an on-line pattern, the proposed method generates images of various stroke shapes using four painting modes: constant line mode, proportional mode and two calligraphic modes. Particularly, the calligraphic modes combine the pen trajectory with real stroke-shape images so that the generated images resemble the characters produced with brush pen. In the calligraphic painting mode based on primitive stroke identification (PSI), the strokes of on-line patterns are classified into different classes and each class of strokes is painted with the corresponding stroke shape template; while in the calligraphic painting mode based on stroke component classification (SCC), each stroke is decomposed into ending, bending and connecting parts, and each part is painted with a stroke shape template. Decomposing strokes into parts is helpful to deal with connected strokes in cursive writing. Our method of image transformation serves two purposes: supplying image samples for off-line recognition, and application of off-line recognition methods to on-line recognition. The experimental results show that our methods are well suited for both purposes. We show that calligraphic painting mode is appropriate for offline recognition while constant line mode and proportional mode are appropriate for on-line recognition.

Handwritten Chinese character recognition using nonlinear active shape models and the Viterbi algorithm

Since Chinese characters are composed from a small set of fundamental shapes (radicals) the problem of recognising large numbers of characters can be converted to that of extracting a small number of radicals and then finding their optimal combination. In this paper, radical extraction is carried out by nonlinear active shape models, in which kernel principal component analysis is employed to capture the nonlinear variation. Treating Chinese character composition as a discrete Markov process, we also propose an approach to recognition with the Viterbi algorithm. Our initial experiments are conducted on off-line recognition of 430,800 loosely-constrained characters, comprised of 200 radical categories covering 2154 character categories from 200 writers. The correct recognition rate is 93.5% characters correct (writer-independent). Consideration of published figures for existing radical approaches suggests that our method achieves superior performance.

Off-Line Arabic Character Recognition – A Review

Off-line recognition requires transferring the text under consideration into an image file. This represents the only available solution to bring the printed materials to the electronic media. However, the transferring process causes the system to lose the temporal information of that text. Other complexities that an off-line recognition system has to deal with are the lower resolution of the document and the poor binarisation, which can contribute to readability when essential features of the characters are deleted or obscured. Recognising Arabic script presents two additional challenges: orthography is cursive and letter shape is context sensitive. Certain character combinations form new ligature shapes, which are often font-dependent. Some ligatures involve vertical stacking of characters. Since not all letters connect, word boundary location becomes an interesting problem, as spacing may separate not only words, but also certain characters within a word. Various techniques have been implemented to achieve high recognition rates. These techniques have tackled different aspects of the recognition system. This review is organised into five major sections, covering a general overview, Arabic writing characteristics, Arabic text recognition system, Arabic OCR software and conclusions.

An off-line oriental character recognition system (OOCRS): synergy of distortion modeling, hidden Markov models and vector quantization

Off-line handwritten oriental character recognition is a difficult task due to the large category and stroke variety. These oriental characters are made up of components known as radicals, which are often written in a distorted proportion and size. All these factors lead to a difficult recognition problem, which unfortunately cannot be solved using direct classification approach like the neural network classifier and a preprocessing module. This paper proposes several novel preprocessing approaches and synergy of classifiers to achieve good performance. Novel classification approaches, comprising rough and coarse classification modules are proposed which when combined appropriately produced a high-performance recognition system capable of producing high accuracy classification in off-line oriental character recognition. The recognition accuracy of the system is a high of 97% and a 99% for the top 5 candidate selection scores.

AN OFF-LINE HANDWRITTEN THAI ALPHABET RECOGNITION SYSTEM FOR UNCONSTRAINED HANDWRITING STYLES

This paper presents a new off-line character recognition system for handwritten Thai alphabets. Unlike many existing techniques, the proposed system has been developed without any assumption on the way that the characters are written. As the results, the reported recognition rate of the proposed system correlates well with the results from real-world tests.The system consists of two classification phases where the techniques that are appropriate for each phase have been independently developed. The first phase classifies character images based on the overall character structure by a technique that emphasizes high tolerance to noise and shape distortion. The second phase performs local contour analysis using a fuzzy system in order to extract detailed features that distinguishes among similar characters. The recognition engine for both phases are neural network-type enhanced with technique called outpost vector for improved generalization. The experimental results on arbitrary handwriting confirms the effectiveness and practicality of the technique.

An Off-Line Handwritten Thai Alphabet Recognition System for Unconstrained Handwriting Styles

This paper presents a new off-line character recognition system for handwritten Thai alphabets. Unlike many existing techniques, the proposed system has been developed without any assumption on the way that the characters are written. As the results, the reported recognition rate of the proposed system correlates well with the results from real-world tests.The system consists of two classification phases where the techniques that are appropriate for each phase have been independently developed. The first phase classifies character images based on the overall character structure by a technique that emphasizes high tolerance to noise and shape distortion. The second phase performs local contour analysis using a fuzzy system in order to extract detailed features that distinguishes among similar characters. The recognition engine for both phases are neural network-type enhanced with technique called outpost vector for improved generalization. The experimental results on arbitrary handwriting confirms the effectiveness and practicality of the technique.

A discrete contextual stochastic model for the off-line recognition of handwritten Chinese characters

We study a discrete contextual stochastic (CS) model for complex and variant patterns like handwritten Chinese characters. Three fundamental problems of using CS models for character recognition are discussed, and several practical techniques for solving these problems are investigated. A formulation for discriminative training of CS model parameters is also introduced and its practical usage investigated. To illustrate the characteristics of the various algorithms, comparative experiments are performed on a recognition task with a vocabulary consisting of 50 pairs of highly similar handwritten Chinese characters. The experimental results confirm the effectiveness of the discriminative training for improving recognition performance.

A stroke-based neuro-fuzzy system for handwritten chinese character recognition

In this article, a stroke-based neuro-fuzzy system for off-line recognition of handwritten Chinese characters is proposed. The system consists of three main components: stroke extraction, feature extraction, and recognition. Stroke extraction applies a run-length-based method to extract strokes from the image of a given character. Various fuzzy features of the extracted strokes, including slope, length, location, and cross relation, are obtained by the feature extraction module. An ART-based neural network, using a two-stage training algorithm, is used to recognize characters. This system extracts strokes in only two passes, and is free from the presence of spurious and thick strokes. The neural model used provides a fast convergence rate. Nodes are allowed to be shared to reduce the size of the resulting network. Features need not be classified in advance by the user. Furthermore, the architecture of the network is self-constructed without the intervention of the user. Experiments have shown that this system is effective.

Off-Line Signature Recognition Using Feature Extraction and Multilayer Neural Networks

Off-Line Signature Recognition Using Feature Extraction and Multilayer Neural Networks

A lexicon-driven approach for optimal segment combination in off-line recognition of unconstrained handwritten Korean words

We propose a new method for off-line recognition of unconstrained handwritten words consisting of Korean and numeric characters. To overcome the difficulty in separating touching characters, we adopt an over-segmentation strategy. Given a slice of the input word image, we find the optimal segment combination using a lexicon-driven word scoring technique and a nearest-neighbor classifier. The optimal combination gives the final segmentation positions for individual characters, along with the best matching word in the lexicon. Superiority of the proposed system has been proven by testing it with 908 images of unconstrained words handwritten on live mail pieces.

Unconstrained word-based approach for off-line script recognition using density-based random-vector functional-link net

This paper reports on the performance of a holistic-styled word-based approach to off-line recognition of English language script. Neural-net computing is used in the implementation of the approach. The thrust of the investigation is straightforward. The objective of the work is to determine to what extent it is possible to build a ‘filter’ which would recognize a word holistically without segmentation, regardless of how ill formed it might be. Of interest also are the failure modes. This investigation required the introduction of innovation in neural-net computing consisting principally of combining the practices of radial basis function neural net (RBNN) approach and the random-vector functional-link net approach (RVFLN). The combined method is called the density-based random-vector functional-link net (DBRVFLN) and it is helpful in improving the performance of the word recognition filter. The CEDAR database was used for training and validation. That database is a real-world database and we used the script images written by 2490 anonymous writers. In-depth investigation on one word yielded results of a very low substitution error rate of 0.74%, with a rejection rate of 11.48%.

Computer Vision and robotics in postal automation

The paper is intended to provide an overview of advanced robotic technologies within the context of Postal Automation services. The main functional requirements of the application are briefly referred, as well as the state of the art and new emerging solutions. Image Processing and Pattern Recognition have always played a fundamental role in Address Interpretation and Mail sorting and the new challenging objective is now off-line handwritten cursive recognition, in order to be able to handle all kind of addresses in a uniform way. On the other hand, advanced electromechanical and robotic solutions are extremely important to solve the problems of mail storage, transportation and distribution, as well as for material handling and logistics. Finally a short description of new services of Postal Automation is referred, by considering new emerging services of hybrid mail and paper to electronic conversion.

Analysis of stroke structures of handwritten Chinese characters

Most handwritten Chinese character recognition systems suffer from the variations in geometrical features for different writing styles. The stroke structures of different styles have proved to be more consistent than geometrical features. In an on-line recognition system, the stroke structure can be obtained according to the sequences of writing via a pen-based input device such as a tablet. But in an off-line recognition system, the input characters are scanned optically and saved as raster images, so the stroke structure information is not available. In this paper, we propose a method to extract strokes from an off-line handwritten Chinese character. We have developed four new techniques: 1) a new thinning algorithm based on Euclidean distance transformation and gradient oriented tracing, 2) a new line approximation method based on curvature segmentation, 3) artifact removal strategies based on geometrical analysis, and 4) stroke segmentation rules based on splitting, merging and directional analysis. Using these techniques, we can extract and trace the strokes in an off-line handwritten Chinese character accurately and efficiently.

Off-line recognition of Chinese handwriting by multifeature and multilevel classification

In this paper, an off-line recognition system based on multifeature and multilevel classification is presented for handwritten Chinese characters. Ten classes of multifeatures, such as peripheral shape features, stroke density features, and stroke direction features, are used in this system. The multilevel classification scheme consists of a group classifier and a five-level character classifier, where two new technologies, overlap clustering and Gaussian distribution selector are developed. Experiments have been conducted to recognize 5,401 daily-used Chinese characters. The recognition rate is about 90 percent for a unique candidate, and 98 percent for multichoice with 10 candidates.

Comparison of Feature Performance and its Application to Feature Combination in Off-Line Handwritten Korean Alphabet Recognition

This paper presents (1) a quantitative performance comparison of the features used in off-line handwritten Korean alphabet recognition, and (2) some experimental results based on the feature combination approach. We classified the features into three: geometrical/topological, statistical and global. For each class, we selected four or five features and performed recognition experiments with PE92 database using neural network classifiers. We compared their recognition performances and selected the top four among those thirteen features: Up Down Left Right Hole, Gradient, Structure, and Mesh. With the combination of some or all of the four features, we repeated recognition experiments. Experimental results show that (1) as a whole, geometrical/topological features outperform the other two feature classes in terms of the recognition rate, and (2) UDLRH and Gradient features in geometrical/topological feature class outperform the other features, (3) the feature combination approach can result in performance improvement.