• 한국통신학회논문지
• /
• 제19권6호
• /
• pp.1140-1148
• /
• 1994

본 논문에서는 음성신호의 프랙탈 차원을 이용하여 한국어 모음인식 실험을 수행하였다. 프랙탈 차원은 Minkowski-Bouligand 차원을 사용하였으며, 형태학적 커버링(morphological covering) 방법을 이용하여 구하였다. 프렉탈 차원과 더불어 기존에 우수한 음성 인식 파라메타로 알려져 있는 LPC 켐스트럼(cepstrum)을 함께 사용하였으며, 프랙탈 차원의 음성인식에의 유용성 여부를 조사하였다. 다양한 자음환경에서의 모음인식 실험결과, LPC 켐스트럼 만을 사용하는 경우 및 프렉탈 차원과 LPC 켐스트럼을 함께 사용하는 경우의 모음 오인식율이 각각 5.6% 및 3.2%로 얻어졌다. 이는 LPC 켑스트럼에 프렉탈 차원을 추가함으로써 오인식되는 데이터가 40%이상 감소되는 결과이며, 프랙탈 차원이 음성인식에 있어서 유용한 특징 파라메터임을 보여준다.

• 한국전자통신학회논문지
• /
• 제7권4호
• /
• pp.783-787
• /
• 2012

한 어절을 중심으로 검사가 이루어지는 맞춤법 검사는 문맥상 어울리지 않는 단어로 인하여 생기는 오류는 찾기 어렵다. 맞춤법 검사기는 현재 어절 단위로 오류 여부를 판단하는 것이기 때문에 어휘적 중의성을 고려하지 않아도 된다. 그러나 문법 검사기는 문장 분석을 해야 하므로 어휘적 중의성을 제거하지 않고는 정확한 검사가 어렵다. 본 논문에서는 어휘적 중의성을 해결하기 위하여 몇 가지 규칙을 만들고 이를 활용하여 문장에 존재하는 어휘적 중의성을 해결할 수 있는 방법을 보인다. 또한 실험을 통하여 그 결과를 분석하였다.

• Journal of Electrical Engineering and Technology
• /
• 제8권5호
• /
• pp.1194-1201
• /
• 2013

This paper introduces a vision-based automatic system (VAMS) for non-contact measurement of morphometric characteristics of flatfish, such as total length (TL), body width (BW), height (H), and weight (W). The H and W are simply measured by a laser displacement and a load cell, respectively. The TL and BW are measured by a proposed morphological image processing algorithm. The proposed algorithm cans measurement, when the tail of flatfish is deformed, and when it is randomly oriented. In the experiment, the average and maximum measurement errors were recorded, and standard deviations and coefficients of variation (CVs) for the measurements were calculated. From those results, when flatfish the TL measurements had an average of 266.844 mm, a standard deviation of 0.351 mm, a CV of 0.131%, and a maximum error of 0.87 mm with straightened flatfish ($TL_A$ : 267 mm, $BW_A$ : 141 mm), and when flatfish of different sizes were measured, the errors in the TL and BW measurements were both about 0.2 %. Using a single conveyor, the VAMS can process up to 900 fishes per hour. Moreover, it can measure morphometric characteristics of flatfish with a TL of up to 500 mm.

• 한국연안방재학회지
• /
• 제6권3호
• /
• pp.111-120
• /
• 2019

In coastal morphological modelling, there are a number of input factors: wave height, water depth, sand particle size, bed friction coefficients, coastal structures and so forth. Measurements or estimates of these input data may include uncertainties due to errors by the measurement or hind-casting methods. Therefore, it is necessary to consider the uncertainty of each input data and the range of the uncertainty during the evaluation of numerical results. In this study, three uncertainty factors are considered with regard to the prediction of coastal erosion in Ilsan beach located in Ilsan-dong, Ulsan metropolitan city. Those are wave diffraction effect of XBeach model, wave input scenario and the specification of the coastal structure. For this purpose, the values of mean wave direction, significant wave height and the height of the submerged breakwater were adjusted respectively and the followed numerical results of morphological changes are analyzed. There were erosion dominant patterns as the wave direction is perpendicular to Ilsan beach, the higher significant wave height, and the lower height of the submerged breakwater. Furthermore, the rate of uncertainty impacts among mean wave direction, significant wave height and the height of the submerged breakwater are compared. In the study area, the uncertainty influence by the wave input scenario was the largest, followed by the height of the submerged breakwater and the mean wave direction.

• 천문학회지
• /
• 제44권5호
• /
• pp.151-160
• /
• 2011

We present a study searching for globular cluster systems (GCSs) of two face-on low surface bright- ness galaxies (LSBGs), UGC 5981 and UGC 6614. Based on HSTWFPC2 images of F555Wand F814W filters, we detect 12 and 18 GC candidates for UGC 5981 and UGC 6614, respectively. Although these two LSBGs have very different bulge properties, they have very similar specific frequencies ($S_N$) of 0.1. However, $S_N$ ~ 0.1 is quite small even for their late morphological types, albeit within errors. This suggests that LSBGs have had star formation histories lacking dominant initial starburst events while accumulating their stellar masses through sporadic star formation activities.

• 한국음향학회:학술대회논문집
• /
• 한국음향학회 1994년도 제11회 음성통신 및 신호처리 워크샵 논문집 (SCAS 11권 1호)
• /
• pp.364-367
• /
• 1994

In this paper, we carried out some experiments on the Korean vowel recognition using the fractal dimension of the speech signals. We chose the Mincowski-Bouligand dimensioni as the fractal dimension, and computed it using the morphological covering method. For our experiments, we used both the fractal dimension and the LPC cepstrum which is conventionally known to be one of the best parameters for speech recognition, and examined the usefulness of the fractal dimension. From the vowel recognition experiments under various consonant contexts, we achieved the vowel recognition error rats of 5.6% and 3.2% for the case with only LPC cepstrum and that with both LPC cepstrum and the fractal dimension, respectively. The results indicate that the incorporation of the fractal dimension with LPC cepstrum gies more than 40% reduction in recognition errors, and indicates that the fractal dimension is a useful feature parameter for speech recognition.

• 인지과학
• /
• 제3권1호
• /
• pp.25-39
• /
• 1991

한국어 어미 중 '아/어'로 시작되는 어미는 용언의 어간과 결합할 때 '아/어'가 축약되거나 '아/어'의 'ㅇ' 이 탈락되기도 한다.특히'ㅎ'불규칙 용언과 결합할 때는 용언의 어간모음과 화합되어 '-ㅐ-','-ㅒ-'로 바뀌는 경우가 있고,'-하다-'로 끝나는 용언의 어간과 결합하면 '-여-'혹은'-ㅐ-'로 바뀌게 된다.본 논문에서는 한국어 형태소 분석기에서 '아/어'로 시작되는 어미의 변이체를 처리하기 위하여 문법형태소 사전을 어떻게 구성하여야하고, 문법형태소 사전이 주기억 장치에 적제될 때 2진트리를 어떻게 구성할 것인에 대한 방법론을 제시함으로써 '아/어' 로 시작되는 어미의 변이체를 형태소 분석할때 발생하기 쉬운 오류를 방지할수 있도록 한다.

• 한국정보과학회:학술대회논문집
• /
• 한국정보과학회 2011년도 한국컴퓨터종합학술대회논문집 Vol.38 No.1(A)
• /
• pp.237-240
• /
• 2011

띄어쓰기 오류는 한국어로 작성된 글에서 나타나는 가장 흔한 오류 중 하나로 문장의 의미적 모호성과 중의성을 가져온다. 규칙 기반 혹은 통계적접근 방법으로 띄어쓰기 오류를 교정하는다양한 방법이 제시되었으나, 기존의 방법들은 띄어쓰기를 형태소 분석의 전단계로 여기거나 띄어쓰기를 교정하기 위해서 형태소 분석을이용하는 등 각각을 독립된 과정으로 다루어, 한 과정에서 발생하는 오류가 다른 과정으로 전파되도록 하는 문제를 안고 있다. 본 논문에서는 띄어 쓰기와 최적 형태소 분석을 하나의 통합된 문제로 다루어각과정에서 발생할 수 있는 오류가 다른 과정에 영향을 주지 않도록 하고 상호 오류를 보완하여 좀더 정확한 띄어쓰기 오류 교정 및 형태소 분석을 가능하게 하는 확률적 접근 방법을 제시한다.

• 영어어문교육
• /
• 제17권1호
• /
• pp.81-99
• /
• 2011

The role of corrective feedback (CF) has long been discussed in the field of second language acquisition. It has been claimed that CF enables learners to notice the problems in their second language (L2) production. However, it should not be assumed that learners always adequately interpret teachers' responses to their problematic utterance as correction. Especially when feedback is provided in an implicit way, the possibility that CF goes unnoticed should not be excluded. In this regard, the study aims to investigate how learners perceive teachers' corrective feedback in English classrooms in Korea. The study focuses particularly on examining the relationship between type of feedback and target linguistic content with learner interpretation of teacher corrective intention. Nine classrooms were observed and videotaped. Forty-five students and nine teachers participated in stimulated recall interviews. Their comments were analyzed to document the learners' perception and the teachers' intention of feedback. It was found that learner perception of teacher corrective intention was at its greatest when feedback was provided explicitly and was focused on morphological errors.

• Journal of Electrical Engineering and information Science
• /
• 제2권6호
• /
• pp.7-13
• /
• 1997

Morphological Analysis of Korean has known to be a very complicated problem. Especially, the degree of part-of-speech(POS) ambiguity is much higher than English. Many researchers have tried to use a hidden Markov model(HMM) to solve the POS tagging problem and showed arround 95% correctness ratio. However, the lack of lexical information involves a hidden Markov model for POS tagging in lots of difficulties in improving the performance. To alleviate the burden, this paper proposes a method for combining multiword units, which are types of lexical information, into a hidden Markov model for POS tagging. This paper also proposes a method for extracting multiword units from POS tagged corpus. In this paper, a multiword unit is defined as a unit which consists of more than one word. We found that these multiword units are the major source of POS tagging errors. Our experiment shows that the error reduction rate of the proposed method is about 13%.