KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 7 of 7

Some of the metrics are blocked by your
consent settings
Recognition of archaic Lanna handwritten manuscripts using a hybrid bio-inspired algorithm
(2015-03-01)
Thammano, Arit
;
Pravesjit, Sakkayaphop
This paper proposes a new character recognition system for archaic Lanna handwritten characters. The proposed system consists of two main processes: the segmentation process and the recognition process. The segmentation process decomposes the touching or overlapping characters, commonly found in Lanna manuscripts, into isolated characters. In the recognition process, the proposed system uses the self-organizing map to divide the input images into several clusters. The clonal selection algorithm of the artificial immune system is then used to build a recognition model for each cluster created by the self-organizing map. Finally, the particle swarm optimization is employed as a local search mechanism. The proposed system was evaluated and compared to several state-of-the-art approaches. The experimental results demonstrate that the proposed system is very effective in recognizing not only Lanna characters but also the handwritten numerals of the five most popular scripts.
Some of the metrics are blocked by your
consent settings
On-line writer dependent character recognition for Khmer based on FIR system characterizing handwriting motion
(2008-12-01)
Thumwarin, Pitak
;
Khem, Sochenda
;
Janchitraponvej, Kanok
;
Matsuura, Takenobu
We propose an on-line writer dependent Khmer recognition method based on time-frequency characteristics of handwriting motion. The handwriting motion can be described by two features, barycenter trajectory and its velocity of pen-tip movement. Then the barycenter trajectory and its velocity of the pen-tip movement are expanded into wavelet series to extract the time-frequency characteristics of the handwriting motion. After that, The fluctuation of the wavelet coefficients can be reduced by the FIR(Finite Impulse Response) Wiener filter. Moreover, the FIR system characterizing the time-frequency characteristics of the handwriting motion is introduced by using wavelet coefficients of the velocity and trajectory of the barycenter with fluctuation reduced as the input and output of the FIR system, respectively. The obtained impulse response of the FIR system is considered as the individual feature for a particular character. Finally, Khmer alphabets can be recognized by using the Euclidean distance between the impulse responses obtained from the reference alphabets and those of the alphabets to be recognized. Khmer character recognition experiments were performed on a database consisting of 6,750 of numerals and alphabets written by 17 people. As the results, the average of the recognition rate was 98.17%. © 2008 SICE.
Some of the metrics are blocked by your
consent settings
Language-support system using character recognition
(2004-01-01)
Asai, Kikuo
;
Sugimoto, Yuji Y.
;
Hemmakorn, Narong
;
Osawa, Noritaka
;
Kondo, Kimio
We have developed a prototype system that supports language translation for distance lecturing and local learning. The system automatically detects a word displayed in a video image by matching the identified character markers; it then translates the word into the other language and displays the translated word using text, image, and/or sound. The ARToolkit is used as an image-processing tool to search the video image for character markers, which are composed of square .frames and registered characters. The use of image processing enables any language to be used as the source language; Japanese is used in the prototype. The use of video makes it easy to implement the system in existing telecommunication systems, because a user just prepares a PC for capturing video images. A Web browser is used as the presentation tool; it can handle any content format normally supported by a Web browser: images, sounds, 3D models, and characters. An automatic voice function automatically reads out words on user demand. The software consists of a character-recognition part (server) and a display part (client); the server controls the client through socket communications. This software architecture makes it possible for one lecturer to simultaneously teach students with various mother tongues, when the clients treat multi-language, because there is a different client for each student. Testing of this language-support system for distance lecturing between Thailand and Japan using a satellite communication system showed that it could be used for both fundamental language lecturing and language translation during international exchange.
Some of the metrics are blocked by your
consent settings
Thai handwriting legal amounts recognition
(2003-01-01)
Chatwiriya, W.
;
Klinkhachorn, P.
;
Lass, N.
A legal amount, or the word amount, is defined as the expression of a numerical value into words. It is used mostly in monetary documents. Thai legal amounts consist of a set of 17 words of 25 characters. This paper's objective is to develop a system to recognize Thai handwriting legal amounts. The strategies to improve the character recognition rates at the preprocessing and recognition process are discussed including the division of input characters into smaller subgroups: the body and high characters group, using multiple feature combinations and using the classifiers that give answers as a list of possible answers with scores. And at the post processes, we improved overall recognition rate using prior knowledge such as lexicon and syntax. We proposed the concept of partial lexicon matching mainly relying on the recognition results of the body group characters. The syntax verification algorithms are based on the sequences of word types, the joining between word types, and special rules for specific words. Irregular handwriting, i.e., touching and broken characters, are detected. Also holistic word recognition is employed to recognize these characters without explicit character segmentation.
Some of the metrics are blocked by your
consent settings
Printed Thai character recognition using the hybrid approach
(2002-01-01)
Thammano, Arit
;
Ruxpakawong, Phongthep
Many researchers have been conducted on the recognition of Thai characters. Different approaches, such as neural network, syntactic, and structural methods, have been proposed. However, the success in recognizing Thai characters is still limited, compared to English characters. This paper proposes an approach to recognize the printed Thai characters using the hybrid of global feature, local features, fuzzy membership function and the neural network. The global feature classifies all characters into seven main groups. Then the local features and the neural network are applied to identify the characters.
Some of the metrics are blocked by your
consent settings
Handwritten Thai character recognition using Fourier descriptors and genetic neural networks
(2002-01-01)
Phokharatkul, Pisit
;
Kimpan, Chom
This article presents a method to solve the rotated and scaling character recognition problem using Fourier descriptors and genetic neural networks. The contours of character image are extracted and separated between the outer contour and inner or loop contours. The loop contours are a special characteristic of Thai characters, called the head of the character. The special features of Thai characters (loop contours) are used at the rough classification stage, and Fourier descriptors with genetic neural networks are used at the fine classification stage. The Fourier descriptors detect the outer contour of a character and it is fed to network. These features are recognized by a multilayer neural network. Genetic algorithms (GAs) are utilized to help compute the weights of the neural network optimally and reduce uncertain states in the neural networks output. Experimental results have shown that the combination of the Fourier descriptors with genetic neural networks, loop features, and local curvature charateristics of similar characters are powerful tools for successfully classifying Thai characters. The recognition rate by this method is 99.12% for 1200 examples of handwritten Thai words (a total of 13,500 characters) written by 60 persons.
Some of the metrics are blocked by your
consent settings
On-line writer recognition for Thai numeral
(2002-01-01)
Thumwarin, P.
;
Tangtisanon, P.
;
Murata, S.
;
Matsuura, T.
This paper discusses an on-line writer recognition for Thai numerals. First, a center point of a Thai numeral is determined after normalization. Then the center point is shifted to an origin. Secondly, the distance from the origin to the pen-point position in the handwriting process is measured and the area of a triangle determined from the origin and the two adjacent points of the pen-point position is calculated. Thirdly, the features of pen-point movement in the process are extracted by expanding the time sequence of the above distance and area into Fourier series. Then the features of pen-point movement are represented in terms of Fourier coefficients. Fourthly, in order to describe the handwriting process features, the FIR system with the above coefficients as input and output of the system is introduced. Then the impulse response of the FIR system is used as the feature of handwriting process. Furthermore in order to recognize the writer, K-L expansion (J.T. Tou and R.C. Gonzalez "Pattern Recognition Principles", Addison-Wesley, pp. 269, 1974) of the obtained impulse response is used. In the experiments, the average rates of type I (false rejection) and type II (false acceptance) error were 17.14% and 9.26%, respectively.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results