Showing papers by "Ching Y. Suen published in 1998"

A recursive thresholding technique for image segmentation

Abstract: In this correspondence, we present a general recursive approach for image segmentation by extending Otsu's (1978) method. The new approach has been implemented in the scope of document images, specifically real-life bank checks. This approach segments the brightest homogeneous object from a given image at each recursion, leaving only the darkest homogeneous object after the last recursion. The major steps of the new technique and the experimental results that illustrate the importance and the usefulness of the new approach for the specified class of document images of bank checks is presented.

Automatic reading of cursive scripts using a reading model and perceptual concepts

Abstract: This paper presents a model for reading cursive scripts which has an architecture inspired by the behavior of human reading and perceptual concepts. The scope of this study is limited to offline recognition of isolated cursive words. First, this paper describes McClelland and Rumelhart's reading model, which formed the basis of the system. The method's behavior is presented, followed by the main original contributions of our model which are: the development of a new technique for baseline extraction, an architecture based on the chosen reading model (hierarchical, parallel, with local representation and interactive activation mechanism), the use of significant perceptual features in word recognition such as ascenders and descenders, the creation of a fuzzy position concept dealing with the uncertainty of the location of features and letters, and the adaptability of the model to words of different lengths and languages. After a description of our model, new results are presented.

Distance features for neural network-based recognition of handwritten characters

Abstract: Features play an important role in OCR systems. In this paper, we propose two new features which are based on distance information. In the first feature (called DT, Distance Transformation), each white pixel has a distance value to the nearest black pixel. The second feature is called DDD (Directional Distance Distribution) which contains rich information encoding both the black/white and directional distance distributions. A new concept of map tiling is introduced and applied to the DDD feature to improve its discriminative power. For an objective evaluation and comparison of the proposed and conventional features, three distinct sets of characters (i.e., numerals, English capital letters, and Hangul initial sounds) have been tested using standard databases. Based on the results, three propositions can be derived to confirm the superiority of both the DDD feature and the map tilings.

HMM-KNN word recognition engine for bank cheque processing

Abstract: Describes the mixed HMM-KNN word recognition module of a bank cheque processing system developed at CENPARMI. It uses a combination of 2 segmentation free word recognition schemes. The first scheme uses a set of global features associated to a modified K nearest neighbour classifier; while the second one uses a set of directional contour features as input to an HMM. The system has been designed to be modular and independent of specific languages as in Canada one has to deal with at least 2 languages, namely English and French. It can be easily adapted to read other European languages based on the Roman alphabet. The system is continuously tested on data from the local phone company, and we report here the results on a database of approximately 4,500 cheques.

Veinerization: a new shape description for flexible skeletonization

Abstract: We introduce the new concept of "veinerization", which produces a graph that contains all the "topological" information needed to derive a wide variety of skeletons. Theoretically, the main contribution is to provide a homogeneous framework for integration of the major concepts described in other related works on digital skeletonization. In practice, the new aspect of this approach is to provide the user with different criteria for selecting the most suitable skeleton for a given application, e.g., the user can select a suitable threshold for obtaining the desirable balance between " having a skeleton without noisy prunes" and "having a skeleton that reflects the initial shape". This algorithm has been tested on numerous kinds of patterns, including pathological ones like fractal sets well-known for the complexity of their shapes.

Sorting and Recognizing Cheques and Financial Documents

Abstract: In collaboration with financial institutions and utility companies, we have carried out substantial research on document analysis and handwriting recognition. This paper describes our prototype which can differentiate between cheques and remittance slips, between English and French cheques, and recognize their contents. A new technique of sorting handwritten cheques and financial documents will be described. It is based on the detection of the structural properties printed on such documents. Handwritten numeric amounts are recognized by a multiple-expert system. These systems have been applied to read handwritten cheques and numerous financial documents with a great variety of backgrounds, colours, and designs in real-life environments. Their performance will be presented and analyzed.

Recognition of legal amounts on bank cheques

Abstract: This article describes the recognition of legal amounts of a bank cheque processing system developed at CENPARMI. The preprocessing, sentence to word segmentation and word recognition approaches are presented along with some critical reviews. The overall engine is a combination of a global feature scheme with an HMM module. The global features consist of the encoding of the relative position of the ascenders, descenders and loops within a word. The HMM uses one feature set based on the orientation of contour points as well as their distance to the baselines. Our system is fully trainable, reducing to a strict minimum the number of hand-set parameters. The system is also modular and independent of specific languages as we have to deal with at least two languages in Canada, namely English and French. The system can be easily adapted to read other European languages based on the Roman alphabet. The system is continuously tested on data from the local phone company, and we report here the results on a balanced French database of approximately 2000 cheques with specified amounts.

Skew detection, page segmentation, and script classification of printed document images

Abstract: Automatic processing of international documents presents a number of challenging problems because Optical Character Recognition (OCR) techniques are not available for all languages and all script classes. Document images must be categorized according to their script type first, in our case Roman, Ideographic, or Arabic. We present a set of statistical methods that first detect and correct the skew of a document image. Next, the page is segmented into text and graphical components. The textual components are then segmented into paragraphs and lines; and finally we classify the script type into one of three categories. The system predicts the correct script category in 91% of cases when tested on real-life documents of varying kinds, diverse formats and qualities from many sources.

Differentiating between oriental and european scripts by statistical features

Abstract: Two types of techniques are usually adopted in language differentiation: token matching and statistical analysis. In this paper we present a method which uses a combined analysis of several discriminating statistical features, for the differentiation between European and oriental language scripts. When applied to more than 23 languages, it has proved to be effective in differentiating between documents printed in these different scripts.

Improved model architecture and training phase in an off-line HMM-based word recognition system

Abstract: Describes the latest developments to enhance the performance of our HMM-based handwritten word recognition system. These methods only deal with the recognition phase and involve the improvement of the HMM architecture as well as the optimization of the training phase. Experiments carried out on real data show that the proposed approaches lead to significant improvements in the accuracy of the system.

Using class separation for feature analysis and combination of class-dependent features

Abstract: We analyze the class separation of the features in handwriting recognition. Behaviors of measurement tools are studied with a partial and full classifications. A new scheme of selecting and combining class-dependent features is proposed. In this scheme, a class is considered to have its own optimal feature vector for discriminating itself from the other classes. Using an architecture of modular neural networks as the classifier, a series of experiments have been conducted on totally unconstrained handwritten numerals. The results indicate that the selected features are effective in separating pattern classes and the new feature vector derived from a combination of two types of such features further improves the recognition rate.

Intriguing aspects of oriental languages

Abstract: This paper includes a description of 3 affiliated oriental languages: Chinese, Japanese, and Korean. It includes a description of the origins of these 3 languages and the inter-relationship among them. Drawn from the viewpoints of several experienced researchers in the field of OCR (Optical Character Recognition) and computational linguistics, it attempts to bring out the intriguing aspects of these 3 ideographic languages, including the formation and composition of pictograms, special features, learning, understanding, contextual information, and recognition of characters and words, and their relations to poetic expressions and pattern recognition techniques. Numerous references are given and comments on future trends are also presented.

Automatic processing of documents and bank cheques

Abstract: Automatic processing of documents with the purpose to scan different documents, recognize them, extract, and process different data items obtained from them could be achieved by top-down and bottom-up approaches. The former processes documents starting from the document class and ends with the pixel representation of the different items that should be extracted. The latter, however, processes documents in a reversed manner. In this thesis, a top-down formal approach for automatic processing of documents and bank cheques is proposed. This approach will view a document as a hierarchy of related items: (a) the background which contains simple or complex scenes that should be eliminated, and (b) the foreground which contains (i) base lines that must be removed and (ii) handwritten data, such as the date, the legal amount, and the courtesy amount, that should be extracted with minimum distortion. The novelty of this new approach is to eliminate the background, first, by introducing a new recursive dynamic thresholding technique that could be used globally or locally on a given cheque image. As a second step, base lines that intersect the handwritten data are recognized and removed with the challenge of minimizing the distortion on the extracted items. Two methods are proposed to tackle this difficulty. The first method detects the handwritten data that intersects with the base lines that should be eliminated and uses morphological and topological processing to identify and fill the gaps resulting from the elimination of the detected base lines. The second method proposed a new dynamic morphological processing technique which acts as a detector and a preserver of the handwritten data that intersect, with the base lines. The second method highly increased the efficiency of item extraction by more than 80% and enhanced the quality of the extracted items when combined with local processing techniques. In a step to study the reliability of the proposed top-down automatic item extraction system, a quantitative analysis technique is investigated and an experimental study is performed comparing the top-down formal approach with another newly developed bottom-up approach using the same training set of 500 real-life bank cheques and two testing sets of 200 bank cheques obtained from the CENPARMI database. The purpose of the quantitative performance analysis technique is to subject the extracted items of the top-down and the bottom-up approaches to the same item processing system that is able to recognize these corresponding items and provide quantitative results to indicate the reliability of both approaches. The experimental results showed that the reliability of the top-down approach on the training set. first testing set, and second testing set are 89.20%, 87.91%, and 90.10% respectively while those on the bottom-up approach are 91.35%, 91.30%, and 93.10%, respectively. Finally, in a step towards the construction of a highly reliable system, a feasibility study has been conducted by combining both approaches. The result is quite encouraging and a reliability of 97.09% has been achieved when these two systems are combined.

Rejection Versus Error in a Multiple Expert Environment

Abstract: The combination of classifiers has become a very active research area in recent years, and many results have been obtained through various methods. This paper presents some of our theoretical and experimental work in this domain.

A Reliability Design Methodology for Chinese Character Recognition

Abstract: This paper proposes a novel method which enables a Chinese character recognition system to obtain reliable recognition. In this method, two thresholds, i.e. class region thresholdRk and disambiguity thresholdAk, are used by each Chinese character k when the classifier is designed based on the nearest neighbor rule, where Rk defines the pattern distribution region of character k, and Ak prevents the samples not belonging to character k from being ambiguously recognized as character k. A novel algorithm to derive the appropriate thresholds Ak and Rk is developed so that a better recognition reliability can be obtained through iterative learning. Experiments performed on the ITRI printed Chinese character database have achieved highly reliable recognition performance (such as 0.999 reliability with a 95.14% recognition rate), which shows the feasibility and effectiveness of the proposed method.

Using Class Separation for Feature Analysis and Combination Features of Class-dependent

Abstract: In tkis paper, we analyze the class separation ofthefeatures in handwriting recognition. Behaviors of measurement tools are studied with partial and full class$cations. A new scheme ofselecting and combining class-dependentfeatures is proposed. In this scheme, a class is considered to have its own optimalfeature vectorfOr discriminating irselfSrom tke other classes. Using an architecture of modular neural nehvorks as tke classifier, a series of experiments kave been conducted on totally unconstrained handwritten numerals. The results indicate that the selectedfeatures are efective in separating pattern classes and the new feature vector derived from a combination of two types of suck featuresfurther improves the recognition rate.