• IEIE Transactions on Smart Processing and Computing
• /
• v.5 no.1
• /
• pp.49-54
• /
• 2016

Regarding global and local factors of a set of features, a given single image or multiple images is a common approach in image processing. This paper introduces an application of an adaptive version of non-local filter whose original version searches non-local similarity for removing noise. Since most images involve texture partner in both foreground and background, extraction of signified regions with texture is a challenging task. Aiming to the detection of visual attention regions for images with texture, we present the contrast analysis of image patches located in a whole image but not nearby with assistance of the adaptive filter for estimation of non-local divergence. The method allows extraction of signified regions with texture of images of wild life. Experimental results for a benchmark demonstrate the ability of the proposed method to deal with the mentioned challenge.

• The Journal of the Korea Contents Association
• /
• v.9 no.4
• /
• pp.121-130
• /
• 2009

In this paper, feature extraction methods, which is one field of reducing dimensions of high-dimensional data, are empirically investigated. We selected the traditional PCA(Principal Component Analysis), ICA(Independent Component Analysis), NMF(Non-negative Matrix Factorization), and sNMF(Sparse NMF) for comparisons. ICA has a similar feature with the simple cell of V1. NMF implemented a "parts-based representation in the brain" and sNMF is a improved version of NMF. In order to visually investigate the extracted features, handwritten digits are handled. Also, the extracted features are used to train multi-layer perceptrons for recognition test. The characteristic of each feature extraction method will be useful when applying feature extraction methods to many real-world problems.

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.1
• /
• pp.79-85
• /
• 2013

Feature extraction of CASA system uses time continuity and channel similarity and makes correlogram of auditory elements for the use. In case of using feature extraction with cross correlation coefficient for channel similarity, it has much computational complexity in order to display correlation quantitatively. Therefore, this paper suggests feature extraction method using non-parametric correlation coefficient in order to reduce computational complexity when extracting the feature and tests to segregate target speech by CASA system. As a result of measuring SNR (Signal to Noise Ratio) for the performance evaluation of target speech segregation, the proposed method shows a slight improvement of 0.14 dB on average over the conventional method.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.41 no.6
• /
• pp.175-186
• /
• 2004

We present a hybrid approach of texture-based method and connected component (CC)-based method for text extraction in complex images. Two primary methods, which are mainly utilized in this area, are sequentially merged for compensating for their weak points. An automatically constructed MLP-based texture classifier can increase recall rates for complex images with small amount of user intervention and without explicit feature extraction. CC-based filtering based on the shape information using NMF enhances the precision rate without affecting overall performance. As a result, a combination of texture and CC-based methods leads to not only robust but also efficient text extraction. We also enhance the processing speed by adopting appropriate region marking methods for each input image category.

• Cross-Cultural Studies
• /
• v.32
• /
• pp.109-132
• /
• 2013

This paper compares sequential and non-sequential verbal conjunctions in Korean, English, and Japanese by looking at how sequential verbal conjunction is treated in each language. It frist reviews verbal conjunctions in Korean, where sequential conjunction is treated as subordination and non-sequential conjunction is treated as coordination, and looks at verbal conjunctions in English and Japanese to see whether or not sequential conjunction in those languages is subordination. According to Oh (2010), sequential and non-sequential conjunctions in Korean behave quite differently with respect to the tense and negation in the final conjunct. Also, Cho (1995, 2005) and Kwon (2004) show that syntactic operations such as extraction and scrambling clearly distinguish sequential conjunction from non-sequential conjunction. The purpose of this paper is to see how sequential and non-sequential conjunctions are analyzed in English and Japanese and to compare those languages with Korean, especially focusing on whether or not sequential conjunctions in English and Japanese are treated as subordination. For this purpose, I first investigate how tense and negation, which provided crucial evidence for concluding that Korean sequential conjunction is subordination, is interpreted in sequential and non-sequential verbal conjunctions in English and Japanese. Also, I investigate the syntactic properties of sequential and non-sequential conjunctions with respect to syntactic operations such as extraction and scrambling in those languages. The results of the investigation show that in Japanese, which is considered typologically similar to Korean, the sequential conjunction is a case of subordination, while in English, which is considered typologically different from Korean, both sequential and non-sequential conjunctions are treated as coordination.

• Proceedings of the IEEK Conference
• /
• 2003.11a
• /
• pp.49-52
• /
• 2003

In this paper, we propose new speech feature parameter using NMf(Non-Negative Matrix Factorization). NMF can represent multi-dimensional data based on effective dimensional reduction through matrix factorization under the non-negativity constraint, and reduced data present parts-based features of input data. In this paper, we verify about usefulness of NMF algorithm for speech feature extraction applying feature parameter that is got using NMF in Mel-scaled filter bank output. According to recognition experiment result, we could confirm that proposal feature parameter is superior in recognition performance than MFCC(mel frequency cepstral coefficient) that is used generally.

• 대한치과교정학회:학술대회논문집
• /
• 1999.11a
• /
• pp.52.3-53
• /
• 1999

• The korean journal of orthodontics
• /
• v.21 no.1 s.33
• /
• pp.223-238
• /
• 1991

This study was carried out in order to findout the amount of tooth movement, the changes arch size and the changes in arch morphology following orthodontic treatment and to provide a guideline for to predict post-treatment arch morphology. The sample group for this study consists of 15 males and 22 females, totalling in 37 persons, who received orthodontic treatment at Orthodontic Department of Dankook Univ. Dental Hospital. They are classified into Extraction Class I treatment group (E I), Non-extraction Class I treatment group (N I), and Non-extraction Class III treatment group (N III), according to their pre-treatment malocclusion state and methods of treatment. Following conclusions and averaged dental arch form for each group were obtained by cephalometric linear measurements and dental arch measurements using pre- and post-treatment lateral cephalograms and plaster study models. 1. Intercanine width were reduced in max. of both EI and NI during the period of treatment, 2. Intermolar width were reduced in max. of EI and increased in max. of NI. Therefore although there was no difference between these two groups before the treatment, intermolar width of the max, of NI was wider than that of E1 after the treatment. 3. PMV-incisor distance and PMV-canine distance were decreased in both max. and mand. of EI and that of NI, during the period of treatment. PMV-molar distance was decreased in both max. and mand. of NI and in mand. of NIII. 4. Items that showed stability during the treatment were: max. & mand. PMV-molar distance, mand. intercanine and intermolar width in EI; mand. intercanine and intermolar width in NI; mand. & max. PMV-incisor distance, PMV-canine distance, max. PMV-molar distance and max. & mand. intercanine and intermolar width in NIII. 5. The differences in averaged canine and molar variances to post-treatment dental arch form were present only in EI and in NI. There was no variance between maxilla and mandible in each group.

• Herbal Formula Science
• /
• v.19 no.1
• /
• pp.175-182
• /
• 2011

Objectives : This study was performed to compare the differences between decoctions extracted by different extraction method and extraction time. Methods : Decoctions were prepared with pressed or non-pressed extraction conditions for 60, 120 and 180min. The yields of extracts, sugar contents, hydrogen ion concentrations(pH), the contents of reference compounds in Yukmijihwang-tang(Liuweidihuang-tang) were investigated. Results : The yields and the dissolved solid content containing sugar content of decoctions were more in pressed extraction method than unpressed extraction method, and they tended to be increased as extraction time increased. The pH values of decoctions methods did not show significant differences between pressed and unpressed extraction methods or extraction times. Most of reference compounds showed higher contents in pressed extraction method than unpressed extraction method and their contents were augmented according to increase of extraction time. The content of paeonol was decreased when extracted in more than 120min with pressed extraction method and tended to be decreased as extraction time increased in unpressed extraction method. Conclusions : The pressed extraction with long extraction time could be useful for decoction of Yukmijihwang-tang(Liuweidihuang-tang). However, another ingredients possible to decrease in such condition need to be considered to determine suitable extraction condition.

• Journal of Multimedia Information System
• /
• v.3 no.2
• /
• pp.13-20
• /
• 2016

Biometric authentication for account-based mobile payment continues to gain attention because of improvements on sensors that can collect biometric information. We propose an enhanced method for mobile payment security based on biometric authentication. In this mobile payment system, the communication between the user and the relying party is based on public key infrastructure. This method secures both the key and the biometric template in the user side using fuzzy vault biometric cryptosystems, which is based on non-random chaff point generator. In this paper, we consider an important process for the common fuzzy vault system, that is, the feature extraction method. We evaluate various feature extraction methods to enhance the accurate performance of the system.