• 융합보안논문지
• /
• 제22권2호
• /
• pp.105-114
• /
• 2022

본 논문은 퍼스널 모빌리티(Personal Mobility, PM)를 이용하는 청각 장애인에게 소리가 발생하는 도래각(Direction of Arrival, DOA)을 시각화하는 지능형 제어 시스템을 제시하며 도로에서 발생하는 경보음, 크락션 등 소리로 인한 위험한 상황들을 인지하고 예방하고자 한다. 소리 위치 추정 방법은 GCC-PHAT(Generalized Cross-Correlation Phase Transform) 기반 도착 지연 시간(Time Difference of Arrival, TDOA)을 특징으로 갖는 머신러닝 분류 모델을 사용한다. 도로 상황을 재현한 실험 환경에서 각각 풍속 0, 5.8, 14.2, 26.4km/h의 조건에 따라 학습 데이터를 추출한 후 학습한 4가지 분류 모델들을 Grid search cross validation으로 비교하며 성능이 가장 우수한 MLP(Multi-Layer Perceptron) 모델을 알고리즘으로 적용하였다. 최종적으로 바람이 발생하였을 때 제안된 알고리즘이 평균 90.7%의 정확도를 나타내었으며, 이는 기존의 일반적인 소리 위치 추정기법보다 평균 7.6-11.5% 정도의 성능 향상을 보이는 것이다.

• 한국음향학회지
• /
• 제42권4호
• /
• pp.364-369
• /
• 2023

두 개 수신기에 들어오는 신호 간의 시간 지연 추정 기술은 수중 음향 뿐만 아니라 실내 음향 및 로보틱스에 이르기까지 다양한 분야에서 응용되고 있는 기술이다. 시간 지연 추정 기술에는 수신기 사이 상호 상관으로부터 시간 지연량을 추정하는 방법이 한 기술 부류이고, 수신기 사이의 시간 지연을 파라메트릭 모델링을 하여 그 파라미터를 시스템 인식의 방법으로 추정하는 기술 부류가 있다. 두 부류 중 후자의 경우 시스템의 파라미터 중에서 지연과 직접 관련 있는 파라미터는 전체 중 극히 일부라는 특성이 있다. 이 특성을 이용하여 Lasso 정규화 같은 방법으로 추정 정확도를 높이기도 한다. 그러나 Lasso 정규화의 경우 필요한 정보가 소실되는 경우가 발생한다. 본 논문에서는 이를 보완하기 위해서 Lasso 정규화에 Ridge 정규화를 덧붙인 Elastic Net을 사용한 방법을 제안한다. 제안한 방법을 기존의 일반 상호 상관(Generalized Cross Correlation, GCC) 방법 및 Lasso 정규화를 사용한 방법과 비교하여, 백색 가우시안 신호원 및 유색 신호원에서도 추정 오차가 매우 적음을 보인다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권4호
• /
• pp.1638-1654
• /
• 2018

The heterogeneous network (HetNet) has been one of the key technologies in Long Term Evolution-Advanced (LTE-A) with growing capacity and coverage demands. However, the introduction of femtocells has brought serious co-layer interference and cross-layer interference, which has been a major factor affecting system throughput. It is generally acknowledged that the resource allocation has significant impact on suppressing interference and improving the system performance. In this paper, we propose a hybrid-clustering algorithm based on the $Mat{\acute{e}}rn$ hard-core process (MHP) to restrain two kinds of co-channel interference in the HetNet. As the impracticality of the hexagonal grid model and the homogeneous Poisson point process model whose points distribute completely randomly to establish the system model. The HetNet model based on the MHP is adopted to satisfy the negative correlation distribution of base stations in this paper. Base on the system model, the spectrum sharing problem with restricted spectrum resources is further analyzed. On the basis of location information and the interference relation of base stations, a hybrid clustering method, which takes into accounts the fairness of two types of base stations is firstly proposed. Then, auction mechanism is discussed to achieve the spectrum sharing inside each cluster, avoiding the spectrum resource waste. Through combining the clustering theory and auction mechanism, the proposed novel algorithm can be applied to restrain the cross-layer interference and co-layer interference of HetNet, which has a high density of base stations. Simulation results show that spectral efficiency and system throughput increase to a certain degree.

• 대한원격탐사학회지
• /
• 제25권6호
• /
• pp.475-486
• /
• 2009

Harmful algal blooms (HAB) caused by the dominant species Cochlodinium polykrikoides (C. polykrikoides) appear in the South Sea of Korea and are particularly present in summer and fall seasons. Environmental factors such as water temperature, weather conditions (air temperature, cloud cover, sunshine, precipitation and wind) influence on the initiation and subsequent development of HAB. The purpose of this research was to study spatial and temporal variations of HAB in the Yeosu area using environmental (oceanic and meteorological) and satellite data. Chlorophyll-a concentrations were calculated using Sea-viewing Wide Field-of-view Sensor (SeaWiFS) images by an Ocean Chlorophyll 4 (OC4) algorithm, and HAB were estimated using the Red tide index Chlorophyll Algorithm (RCA). We also used the surface velocity of sequential satellite images applying the Maximum Cross Correlation method to detect chlorophyll-a movement. The results showed that the water temperature during HAB occurrences in August 2002-2008 was $19.4-30.2^{\circ}C$. In terms of the frequency of the mean of cell density of C. polykrikoides, the cell density of the HAB found at low (<300 cells/ml), medium (300-1000 cells/ml), and high (>1000 cells/ml) levels were 27.01%, 37.44%, and 35.55%, respectively. Meteorological data for 2002-2008 showed that the mean air temperature, precipitation, wind speed and direction, and sunshine duration were $22.39^{\circ}C$, 6.54 mm/day, 3.98 m/s (southwesterly), and 1-11.7 h, respectively. Our results suggest that HAB events in the Yeosu area can be triggered and extended by heavy precipitation and massive movement of HAB from the East China Sea. Satellite images data from July to October 2002-2006 showed that the OC4 algorithm generally estimated high chlorophyll-a concentration ($2-20\;mg/m^3$) throughout the coastal area, whereas the RCA estimated concentrations at $2-10\;mg/m^3$. The surface velocity of chlorophyll-a movement from sequential satellite images revealed the same patterns in the direction of the Tsushima Warm Current.

• Smart Structures and Systems
• /
• 제22권5호
• /
• pp.587-609
• /
• 2018

There is greater significance in identifying the incipient damages in structures at the time of their initiation as timely rectification of these minor incipient cracks can save huge maintenance cost. However, the change in the global dynamic characteristics of a structure due to these subtle damages are insignificant enough to detect using the majority of the current damage diagnostic techniques. Keeping this in view, we propose a hybrid damage diagnostic technique for detection of minor incipient damages in the structures. In the proposed automated hybrid algorithm, the raw dynamic signatures obtained from the structure are decomposed to uni-modal signals and the dynamic signature are reconstructed by identifying and combining only the uni-modal signals altered by the minor incipient damage. We use these reconstructed signals for damage diagnostics using ARMAX model. Numerical simulation studies are carried out to investigate and evaluate the proposed hybrid damage diagnostic algorithm and their capability in identifying minor/incipient damage with noisy measurements. Finally, experimental studies on a beam are also presented to compliment the numerical simulations in order to demonstrate the practical application of the proposed algorithm.

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2006년도 춘계학술발표회 논문집
• /
• pp.181-185
• /
• 2006

Image co-registration is the process of overlaying two images of the same scene, one of which represents a reference image, while the other is geometrically transformed to the one. In order to improve efficiency and effectiveness of the co-registration approach, the author proposed a pre-qualified area matching algorithm which is composed of feature extraction with canny operator and area matching algorithm with cross correlation coefficient. For refining matching points, outlier detection using studentized residual was used and iteratively removes outliers at the level of three standard deviation. Throughout the pre-qualification and the refining processes, the computation time was significantly improved and the registration accuracy is enhanced. A prototype of the proposed algorithm was implemented and the performance test of 3 Landsat images of Korea showed: (1) average RMSE error of the approach was 0.436 Pixel (2) the average number of matching points was over 38,475 (3) the average processing time was 489 seconds per image with a regular workstation equipped with a 3 GHz Intel Pentium 4 CPU and 1 Gbytes Ram. The proposed approach achieved robustness, full automation, and time efficiency.

• 제어로봇시스템학회논문지
• /
• 제17권2호
• /
• pp.149-156
• /
• 2011

This paper reveals a distance estimation algorithm based on multi-code ultrasonic and wireless sensor network. For measuring the distances among the sensor nodes, each ultrasonic transmitter transmits multi-code ultrasonic signal simultaneously. Receivers use cross correlation method to separate the coded signals. The information of measured distances is broadcasted to each sensor node by wireless sensor network. The wireless sensor network measures the distance among the sensor nodes using the received signal strength of the broadcasting. The multi-code ultrasonic have a limitation of measurable distance. And the received signal strength is affected from an environment. This paper measures a distance using ultrasonic and a received signal strength in short range. These measured data are applied to the least square estimation algorithm. By the expansion of the fitting curve, a distance measurement in long range using the received signal strength is compensated. The coupled system reduce the error to an acceptable level.

• 한국정보과학회논문지:소프트웨어및응용
• /
• 제28권11호
• /
• pp.855-863
• /
• 2001

본 논문에서는 실감 있는 3차원 얼굴모델 생성에 필요한 얼굴 텍스쳐를 생성하기 위해 실린더형 원형모델의 가변적 블록크기를 기반으로 하는 영상정렬 알고리즘을 제안한다. 이 알고리즘은 실린더형 3차원 모델에 대한 2차원 영상을 영상의 상관관계를 이용하는 블록 정합 알고리즘을 사용해서 정렬하는 방법으로서 블록 정합을 수행 할 때 동일한 블록 크기를 사용하는 기존 방법과는 달리 모델의 곡률을 고려해 가변적인 크기의 블록을 사용한다. 그리고 영상 모자익 기법을 사용해서 정렬된 영상을 하나의 텍스쳐 영상으로 생성한다. 이 때 중첩된 영역에 따라 선형적인 가중치를 부여하면서 크로스디졸브 기법을 이용해 영상을 스티칭한다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2002년도 춘계학술발표논문집 (상)
• /
• pp.647-650
• /
• 2002

The human faces do not have distinct features unlike other general objects. In general the features of eyes, nose and mouth which are first recognized when human being see the face are defined. These features have different characteristics depending on different human face. In this paper, We propose a face recognition algorithm using the hidden Markov model(HMM). In the preprocessing stage, we find edges of a face using the locally adaptive threshold scheme and extract features based on generic knowledge of a face, then construct a database with extracted features. In training stage, we generate HMM parameters for each person by using the forward-backward algorithm. In the recognition stage, we apply probability values calculated by the HMM to input data. Then the input face is recognized by the euclidean distance of face feature vector and the cross-correlation between the input image and the database image. Computer simulation shows that the proposed HMM algorithm gives higher recognition rate compared with conventional face recognition algorithms.

• 전기학회논문지
• /
• 제60권4호
• /
• pp.877-879
• /
• 2011

Bowel sounds (BS) are produced by the movement of the intestinal contents in the lumen of the gastro-intestinal tract during peristalsis and thus, it can be used clinically as useful indicators of bowel motility. We devised an estimation algorithm of bowel motility based on the regression modeling of the jitter and shimmer of BS signals measured by auscultation. Ten healthy males ($23.5\pm2.1$ years) were examined. Consequently, the correlation coefficient, coefficient of determination and standard error between the colon transit times (CTT) measured by a conventional radiograph and the values estimated by our algorithm were 0.98, 0.96 and 2.86, respectively. Also, through k-fold cross validation, the average value of the absolute differences between them was $5.0\pm2.5$ hours. This method could be used as a complementary tool for the non-invasive measurement of bowel motility.