Последнее обновление: 25.10.2006   Статьи / Алгоритмы

Алгоритмы

A Format-Driven Handwritten Word Recognition System Авторы: Xia Liu & Zhixin Shi
Организация: Center of Excellence for Document Analysis and Recognition State University of New York at Buffalo, Buffalo, NY14260, U.S.A.
Дата: ориентировочно 2000-2003 год
Кол-во страниц: 5
A format-driven word recognition system is proposed for recognition of handwritten words. Unlike most traditional handwritten word recognizers being given a set of target words as lexicon, we assume that our system is given a set of format descriptions other than lexicon words. Applications of the proposed system include recognition of relatively more important keywords such as postal codes, titles or trademarks. The format descriptions are in terms of the lengths of the keywords, the types of the characters in the keywords and positional informations. Due to the important role of the keywords in the applications,the recognition expectations in terms of recognition rate and accuracy are usually higher then lexicon-driven wordrecognizers.
 Скачать файл (85 Кб)
 Скачать файл (59 Кб)

Empirical Error based Optimization of SVM Kernels: Авторы: N.E. Ayat & M. Cheriet & C.Y.Suen
Организация: LIVIA, Ecole de TechnologieSuperieure,' 1100, rue Notre Dame Ouest, Montreal, H3C 1K3, Canada CENPARMI,ConcordiaUniversity,1455 de MaisonneuveBlvd West,Montreal, H3G 1M8,Canada
Дата: ориентировочно 2001-2003 год
Кол-во страниц: 6
We address the problem of optimizing kernel parameters in Support Vector Machine modeling, especially when the number of parameters is greater than one as in polynomial kernels and KMOD, our newly introduced kernel. The present work is an extended experimental study of the framework proposed by Chapelle et al. for optimizing SVM kernels using an analytic upper bound of the error. However, our optimization scheme minimizes an empirical error estimate using a Quasi-Newton optimization method. Toassess our method, the approach is further used for adapting KMOD, RBF and polynomial kernels on synthetic data and NIST database. The method shows a much faster convergence with satisfactory results in comparison with the simple gradient descent method.
 Скачать файл (103 Кб)

AN OBJECT-ORIENTED CHARACTER RECOGNITION ENGINE Авторы: B. Klauer & K. Waldschmidt & J.W.Goethe
Организация: Universitat Frankfurt, Frankfurt am Main, INFOS GmbH Ludwigstr. 78, D - 63067 Offenbach
Дата: 1993 год
Кол-во страниц: 14
This paper introduces the object-oriented character recognition engine AQUIRE which was originally designed as a general pattern-classifier under the academic aspects of object-orientation and parallelism together with low space and time complexity. When industry ex- pressed interest in using AQUIRE to provide pen-based computers with a powerful character-recognition engine, AQUIRE was revised to meet the commercial requirements. Together with a brief discussion of some special pen-related problems, this paper contains the theoretical background for the recognition mechanism and a description of the methods used to perform a high quality and high speed recognition-process with a minimum of executable code. The proposed method has been implemented on an INFOS NotePad 386-SX pen-computer and on the associative processor AM developed within the PROMETHEUS project at the Department for Technical Computer Sciences at J.W.Goethe-University. The acceleration of the recognition mechanism by using the associative type of parallelism will be shown.
 Скачать файл (53 Кб)

Imaged Document Text Retrieval without OCR Авторы: Chew Lim Tan , Weihua Huang, Zhaohui Yu, Yi Xu
Организация: School of Computing, National University of Singapore Kent Ridge, Singapore 117543
Дата: ориентировочно 2002-2004 год
Кол-во страниц: 16
We propose a method for text retrieval from document images without the use of OCR. Documents are segmented into character objects. Image features, namely the Vertical Traverse Density (VTD) and Horizontal Traverse Density (HTD), are extracted. An n-gram based document vector is constructed for each document based on these features. Text similarity between documents is then measured by calculating the dot product of the document vectors. Testing with seven corpora of imaged textual documents in English and Chinese as well as images from UW1 database confirms the validity of the proposed method.
 Скачать файл (221 Кб)

Localization and Recognition of Traffic Signs for Automated Vehicle Control Systems Авторы: M.M. Zadeh, T. Kasvand, C.Y. Suen
Организация: Concordia University, Computer Science Department, CENPARMI, GM 606 1455 De Maisonneuve Blvd west, Montreal, PQ, H3G 1M8, Canada
Дата: ориентировочно 2000-2003 год
Кол-во страниц: 14
We present a computer vision system for detection and recognition of traffic signs. Such systems are required to assist drivers and for guidance and control of autonomous vehicles on roads and city streets. For experiments we use sequences of digitized photographs (to be replaced by video tapes taken from a moving car) and offнline analysis. The system contains four stages. First, region segmentation based on colour pixel classification called SRSM (Supervised Region Segmentation Method). SRSM limits the search to regions of interest in the scene (image). Second, we use edge tracing to find parts of outer edges of signs which are circular or straight, corresponding to the geometrical shapes of traffic signs (circle, triangle, rectangle, octagon). The third step is geometrical analysis of the outer edge and preliminary recognition of each candidate region, which may be a potential traffic sign. The final step in recognition uses colour combinations within each region and model matching. This system may be used for recognition of other types of objects, provided that the geometrical shape and colour content remain reasonably constant. The method is reliable, easy to implement, and fast. This differs from the road signs recognition method in the PROMETEUS (EU45), [1]. The overall structure of the approach is sketched in Fig. 1.
 Скачать файл (612 Кб)

Optical Character Recognition Авторы: Line Eikvil
Организация: Norsk Regnesentral, P.B. 114 Blindern, N 0314
Дата: ориентировочно 1993 год
Кол-во страниц: 35
Хорошее описание истории методов оптического распознавания символов. К сожалению, уже несколько устаревшее. Тем не менее, в работе представлен довольно интересный метод фильтрации изображения на этапе предобработки.
 Скачать файл (146 Кб)

A moving window classifier for off-line character recognition Авторы: M.S. HOQUE, M.C. FAIRHURST
Организация: Electronic Engineering Laboratory, University of Kent, Canterbury, UK
Дата: ориентировочно 2001-2003 год
Кол-во страниц: 6
A new classification scheme, primarily aimed at applications in document image processing is presented. Features are extracted from a partial image and a subclassifier generates strokes based on likelihood of the sub-image belonging to the candidate classes. This partial classification is carried out for several overlapping image segments and scores are combined to make the final classification. The scheme shows promising results in OCR applications where high processing speeds are achievable with minimal compromise in the recognition accuracy.
 Скачать файл (55 Кб)

Character recognition by matching sequences of pseudo-stroke positions and directions Авторы: Hanhong Xue & Venu Govindaraju
Организация: CEDAR, State University of New York at Buffalo
Дата: ориентировочно 2001-2003 год
Кол-во страниц: 6
Chain-coded contours are informative in off-line character recognition. As approximations to contours, sequences of pseudo-strokes consisting of both positional and directional information make up feature vectors for character images. In order to carry out fast pattern matching, a scheme of generating fixed-length feature vectors that combine information about outer contour and inner contours into a uniform data structure is proposed and tested on CEDAR databases.
 Скачать файл (61 Кб)

Segmentation and Recognition of Handwritten Dates Авторы: M.Morita , R. Sabourin , F. Bortolozzi , and C. Y. Suen
Организация: Ecole de Technologie Sup'erieure - Montreal,Canada RecognitionandMachine Intelligence- Montreal,Canada Pontif'?ciaUniversidadeCat'olica do Paran'a - Curitiba, Brazil
Дата: ориентировочно 2003-2005 год
Кол-во страниц: 6
This paper presents an HMM-MLP hybrid system to recognize complex date images written on Brazilian bank cheques. The system first segments implicitly a date image into sub-fields through the recognition process based on an HMM-based approach. Afterwards, the three obligatory date sub-fields are processed by the system (day, month and year). A neural approach has been adopted to work with strings of digits and a Markovianstrategyto recognize and verify words. We also introduce the concept of meta-classes of digits, which is used to reduce the lexicon size of the day and year and improve the precision of their segmentation and recognition. Experiments show interesting results on date recognition.
 Скачать файл (84 Кб)

Uncinstrained handwriting recognition applied to the processing of the bank cheques Авторы: Didier Guillevic
Организация: Concordia University Montreal, Canada
Дата: 1995 год
Кол-во страниц: 198
A method for recognizing unconstrained handwritten words belonging to a small static lexicon is proposed. Previous approaches typically attempt to recognize charн acters or parts of characters in order to recognize words. Our approach, in its first step, bypasses the notion of characters. In addition to language independence, our method is more context oriented and should prove to be more robust against poor handwriting, spelling mistakes, noise and the like. Our computational theory is based on a psychological model of the reading process of a fast reader. First a few graphical clues such as ascenders, descenders and their relative positions are extracted from the word. If these prove not to be sufficient to clearly identify the word, then details (secн ondary features including first and last characters of words) are extracted to enhance the word recognition. We designed and collected a database of bank cheques both in English and French. This resulted in a one of its kind database in a university setting dealing with handн written information from bank cheques, both in terms of the size of the database as well as the number of different writers involved. We further designed an innovaн tive, simple yet powerful in place tagging procedure for our database. It enables us to extract at will not only the bitmaps of words, characters, digits, lines, commas, etc... but also all kinds of contextual information. We developed a fully trainable word recognizer with the requirement that the switch to a different database and/or language shall not require any redesign nor any extensive retraining time. The number of parameters within the system has been kept to a minimum and whenever possible we designed algorithms that require no parameters and therefore no training. Such an example is our slant correction algorithm that shines by its simplicity and robustness. Whenever parameters might need to be adjusted to a specific database, it is done automatically by running some genetic algorithms. We tested the generality and adaptability of our system on 2 different databases of bank cheques (respectively English and French). We noticed that the system's parameters did not need to be readjusted for it to perform satisfactorily when the switch was made from one database to the other. At the time of this dissertation, our survey indicates that this research is the only one in the literature which can handle English cheques and our results are comparable to those published on the processing of French cheques. Our preliminary results on the French database report a recognition rate of 98.9% and 94.3% on the word and the full legal amount respectively among the top 5 choices.
 Скачать файл (4059 Кб)

Recognition of Handwritten Numerals Using Elastic Matching Авторы: Patrice Scattolin
Организация: Concordia University Montreal, Canada
Дата: 1995 год
Кол-во страниц: 174
Elastic matching has been used for the recognition of handwritten characters for two decades. It is usually only used for writerнdependent systems with onнline data. We attempt to use this method in a multiнwriter environment for both onнline and offнline recognition of handwritten numerals. By its nature, elastic matching is best suited to single writer onнline systems. Two challenges present themselves to attain reasonable results under these conditions. First, the algorithm must be modified to better generalize the models, to recognize a wider variety of patterns with a given number of models. Secondly the offнline data is not in a suitable format as the patterns are not represented by a sequence of ordered points. We will apply two modifications to the typical elastic matching system to adapt it to the multiнwriter environment and for the offнline data. To process the offнline data, we use a stroke reconnection heuristic to create data nearly identical to typical onнline data. To adapt to a multiн writer environment, we modify the elastic matching algorithm to add weights to each point of the models. These weights allow portions of the characters that better discriminate between confusing classes to take on more importance. In so doing confusing classes are better separated, increasing the recognition rate.
 Скачать файл (777 Кб)


Поддержите проект материально!

рублей Яндекс.Деньгами
на счёт 41001275511292 (cr-online.ru)

Вы используете мой ресурс? Буду вам очень благодарен, если накоплю хотя бы на оплату хостинга.