• Journal of Korea Multimedia Society
• /
• v.11 no.5
• /
• pp.623-632
• /
• 2008

In this paper, we used the color transfer function and the opacity transfer function for the internal 3D object visualization of an image volume data. In transfer function, creating values of between boundaries generally is ambiguous. We concentrated to extract boundary features for segmenting the visual volume rendering objects. Consequently we extracted an image gradient feature in spatial domain and created a multi-dimensional transfer function according to the GPU efficient improvement. Finally using these functions we obtained a good research result as an implementing object boundary visualization of the graphic hardware based 3D texture mapping.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.05a
• /
• pp.319-322
• /
• 2007

본 논문에서는 영상 감시 등의 응용을 위해 영상에서 자동으로 사람 수를 측정하는 군중 수 측정 방법을 제안한다. 제안한 방법에서는 전처리 과정으로 전경영상과 경계영상을 검출하여 객체의 픽셀 크기 히스토그램과 경계 방향 히스토그램을 특징으로 이용하고 카메라 투영행렬을 통해 픽셀 크기와 경계 방향에 대한 특징 정규화를 수행한다. 실제 사람 수와 얻어진 특징 히스토그램 간의 선형성은 사상 함수의 구성에 적용되며, 훈련 데이터를 통해 얻어진 사상 함수는 사람 수 측정에 이용되었다. 제안한 방법의 성능은 건물 내에서 촬영된 영상에 대한 실험 결과로 나타났으며 이 방법이 영상 감시 분야에 다양하게 적용될 수 있음이 확인되었다.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.3
• /
• pp.182-187
• /
• 2003

In this thesis, we propose the algorithm to improve the performance of speaker verification that is mapping feature parameters by using RBF neural network. There is a big difference between wire vector region and wireless one which comes from the same speaker. For wire/wireless speakers model production, speaker verification system should distinguish the wire/wireless channel that based on speech recognition system. And the feature vector of untrained channel models is mapped to the feature vector(LPC Cepstrum) of trained channel model by using RBF neural network. As a simulation result, the proposed algorithm makes 0.6%∼10.5% performance improvement compared to conventional method such as cepstral mean subtraction.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.11a
• /
• pp.263-266
• /
• 2000

본 논문에서는 복잡한 카오스 신호를 발생시키는 HVPM(hyperchaotic volume preserving maps) 모델과 HVPM 모델의 구현회로를 제안한다. 랜덤한 카오스 신호를 발생시키기 위하여 3차원 이산시간(discrete-time) 연산과 비선형 사상(maps)으로 모듈러(modulus) 함수를 이용하여 하이퍼카오스 신호를 발생시킨다. 그리고 HVPM 모델은 여러 가지 시스템 파라미터들을 변화시키면 다양한 카오스 신호를 발생시킬 수 있으며, 출력되는 카오스 신호는 비주기성을 갖게 된다. 이러한 특징을 갖는 HVPM 모델의 회로 구현을 위하여 2단 N형의 함수를 CMOS와 선형 연산증폭기 및 비교기를 이용하여 보드상에서 구현하여, 다양한 하이퍼카오스 신호를 확인할 수 있었다.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.1
• /
• pp.187-194
• /
• 1996

In this paper, we propose regression neural networks based on regression trees. We map regression trees into three layered feedforward networks. We put multi feature split functions in the first layer so that the networks have a better chance to get optimal partitions of input space. We suggest two supervised learning algorithms for the network training and test both in single feature split and multifeature split functions. In experiments, the proposed regression neural networks is proved to have the better learning performance than those of the single feature split regression trees and the single feature split regression networks. Furthermore, we shows that the proposed learning schemes have an effect to prune an over-grown tree without degrading the learning performance.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.2
• /
• pp.359-365
• /
• 2020

Sasang Constitution data that were classified by the QSCCII (Questionnaire for the Sasang Constitution Classification II) and artery pulse waves of Chon, Guan, and Chuck data measured using an electronic manometer, were obtained from 732 subjects who visited an oriental hospital. The pulse width, peak height, and number of peaks were extracted from the pulse waves as feature variables. Validity and reliability analyses were performed to obtain the feature variables. The feature variables with high validity and reliability were selected as the discriminant variables. The pulse wave data were divided into training and predicting samples by applying a fivefold cross-validation technique. Discriminant analysis was performed for the training sample, and discriminant functions were obtained. The discriminant functions were applied to the predicting sample and the Sasang Constitution was predicted. The accuracy of prediction was estimated by comparing the predicted Sasang Constitution and that obtained by QSCCII. The accuracy of the predicted Sasang Constitution before (after) age and sex calibration was 73.6 % (70.4 %), 68.4 % (84.2 %), and 74.2 % (67.7 %) for Taeumin, Soumin, and Soyangin, respectively, and 72.5 % (73.8 %) in total.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.404-404
• /
• 2017

이상기후현상으로 인해 극치 수문 사상들이 빈번히 발생함에 따라 상대적으로 높은 재현기간에 해당하는 극치 수문 사상해석에 대한 관심이 높아지고 있다. 그러나 우리나라의 경우 이러한 극치 수문 사상을 추정하기 위한 표본의 수가 부족한 실정이다. 지역빈도해석은 지점의 표본 수가 적거나 수문자료의 수집이 불가능한 미계측지점인 경우, 해당 지점과 수문학적으로 동질하다고 여겨지는 주변 지점들의 자료를 확보하여 확률수문량을 추정함으로써 상대적으로 지점빈도해석 보다 roubst한 추정값을 얻을 수 있다는 장점을 가지고 있다. 따라서 최근 확률수문량 산정 기법으로 지역빈도해석 방법에 관한 관심이 높아지고 있다. 지역구분은 지역빈도해석이 지점빈도해석과 구분될 수 있는 큰 특징이고 지역구분 결과 따라 지역의 표본 크기가 결정되기 때문에 수문학적으로 동질한 지역을 나누는 방법은 매우 중요하다고 볼 수 있다. 인공신경망은 인간의 뇌가 학습하는 방식을 모사한 통계적 모델링 기법이다. 즉, 인간의 뇌가 일정한 반복 학습을 통해 어떠한 문제의 해법을 추론하거나 예측, 또는 패턴을 인식하는 일련의 과정을 알고리즘화 하여 목적함수의 해를 찾는 방식이다. 특히, 주어진 자료들로 부터 특징을 추출하고 그 특징을 학습하여 전체 자료의 분류나 군집화를 이루는데 널리 이용되고 있다. 본 연구에서는 낙동강유역을 대상으로 인공신경망을 이용한 군집분석을 수행하고 구분된 지역을 이용하여 지역빈도해석을 수행하였다.

• Journal of Korea Water Resources Association
• /
• v.49 no.3
• /
• pp.217-225
• /
• 2016

The drought is generally characterized by duration and severity, thus it is required to conduct the bivariate frequency analysis simultaneously considering the drought duration and severity. However, since a bivariate joint probability distribution function (JPDF) has a 3-dimensional space, it is difficult to interpret the results in practice. In order to suggest the technical solution, this study employed copula functions to estimate an JPDF, then developed conditional JPDFs on various drought durations and estimated the critical severity corresponding to non-exceedance probability. Based on the historical severe drought events, the hydrologic risks were investigated for various extreme droughts with 95% non-exceedance probability. For the drought events with 10-month duration, the most hazardous areas were decided to Gwangju, Inje, and Uljin, which have 1.3-2.0 times higher drought occurrence probabilities compared with the national average. In addition, it was observed that southern regions were much higher drought prone areas than northern and central areas.

• Journal of KIISE:Software and Applications
• /
• v.31 no.3
• /
• pp.361-368
• /
• 2004

The purpose of this study is to propose a new on-line nonlinear PCA(OL-NPCA) method for a nonlinear feature extraction from the incremental data. Kernel PCA(KPCA) is widely used for nonlinear feature extraction, however, it has been pointed out that KPCA has the following problems. First, applying KPCA to N patterns requires storing and finding the eigenvectors of a N${\times}$N kernel matrix, which is infeasible for a large number of data N. Second problem is that in order to update the eigenvectors with an another data, the whole eigenspace should be recomputed. OL-NPCA overcomes these problems by incremental eigenspace update method with a feature mapping function. According to the experimental results, which comes from applying OL-NPCA to a toy and a large data problem, OL-NPCA shows following advantages. First, OL-NPCA is more efficient in memory requirement than KPCA. Second advantage is that OL-NPCA is comparable in performance to KPCA. Furthermore, performance of OL-NPCA can be easily improved by re-learning the data.

• Journal of the Korean GEO-environmental Society
• /
• v.21 no.4
• /
• pp.5-17
• /
• 2020

In recent years, the frequency of heavy rainfall associated with high rainfall intensity has been continuously increasing due to the effects of climate change; and thus also causes an increase in watershed soil erosion. The existing estimation techniques, used for the prediction of soil erosion in Korea have limitations in predicting the: average soil erosion in watersheds, and the soil erosion associated with abnormal short-term rainfall events. Therefore, it is necessary to consider the characteristics of torrential rainfall, and utilize physics-based model to accurately determine the soil erosion characteristics of a watershed. In this study, the rainfall kinetic energy equation, in the form of power function, is proposed by applying the probability density function, to analyze the rainfall particle distribution. The distributed rainfall-erosion model, which utilizes the proposed rainfall kinetic energy equation, was utilized in this study to determine the soil erosion associated with various typhoon events that occurred at Cheoncheon watershed. As a result, the model efficiency parameters of the model for NSE and RMSE are 0.036 and 4.995 ppm, respectively. Therefore, the suggested soil erosion model, coupled with the proposed rainfall-energy estimation, shows accurate results in predicting soil erosion in a watershed due to short-term rainfall events.