• Applied Microscopy
• /
• 제38권3호
• /
• pp.275-278
• /
• 2008

Electron tomography (ET) is an electron microscopic technique for obtaining a 3-D image from any electron microscopy specimen and its application in biomedical science has been increased thanks to development of electron microscopy and related technologies during the last decade. There are few researches on ET in Korea during this period. Although the importance of ET has been recognized recently by many researchers, initial approach to electron tomographic research is not easy for beginners. The 4th International Congress on Electron Tomography (4 ICET) was held on Nov $5{\sim}8$, 2006, at San Diego. The program dealt instrumentation, reconstruction algorithm, visualization/quantitative analysis and electron tomographic presentation of biological specimen and nano particles. 1 have summarized oral presentations and analyzed the posters presented on the meeting. Cryo-electron microscopic system was popular system for ET and followed conventional transmission electron microscopic system. Cultured cell line and tissue were most popular specimens analyzed and microorganisms including bacteria and virus also constituted important specimens. This analysis provides a current state of art in ET research and a guide line for practical application of ET and further research strategies.

• Journal of Mechanical Science and Technology
• /
• 제18권12호
• /
• pp.2265-2272
• /
• 2004

The aquaculture tank is used for breeding fish in sea water which was pumped up to land. The flow characteristics in the aquaculture were investigated with varying the tank geometry and flow rate. The numerical analysis has been employed for calculating the velocity and temperature distributions in a water tank of rectangular type. The finite volume method and SIMPLE algorithm with 3-dimensional standard $\kappa$-$\varepsilon$ turbulence model were used for the numerical analysis. For comparison with experimental results, the PIV system was applied to visualize the flow patterns quantitatively. The numerical results showed good agreements with the experimental results. The mean velocity and temperature versus aquarium depth were represented for various circulating flow rates. Especially, the aquaculture environment is recommended that the aquarium depth has to be d=0.5 m, and this case is the condition of higher velocity and temperature in winter season.

• 대한원격탐사학회지
• /
• 제28권6호
• /
• pp.623-632
• /
• 2012

In this paper, we analyzed the effect of wavelet decomposition levels in feature extraction for anomaly detection from hyperspectral imagery. After wavelet analysis, anomaly detection was experimentally performed using the RX detector algorithm to analyze the detecting capabilities. From the experiment for anomaly detection using CASI imagery, the characteristics of extracted features and the changes of their patterns showed that radiance curves were simplified as wavelet transform progresses and H bands did not show significant differences between target anomaly and background in the previous levels. The results of anomaly detection and their ROC curves showed the best performance when using the appropriate sub-band decided from the visual interpretation of wavelet analysis which was L band at the decomposition level where the overall shape of profile was preserved. The results of this study would be used as fundamental information or guidelines when applying wavelet transform to feature extraction and selection from hyperspectral imagery. However, further researches for various anomaly targets and the quantitative selection of optimal decomposition levels are needed for generalization.

• 한국산업정보학회논문지
• /
• 제12권4호
• /
• pp.164-168
• /
• 2007

본 논문에서 임베디드시스템에 적용 가능한 눈 검출 알고리즘을 제안하였다. 특히, 감시카메라나 자동현금인출장치 또는 운전자의 졸음운전방지를 위한 눈 검출에서는, 주로 정면얼굴에서의 눈 검출이 이루어지므로 본 논문에서는 이러한 조건을 목표로 눈 검출 알고리즘을 제안하였다. 눈영역을 검출하기 위해, 특성백터를 먼저 추출하고 그 다음, 주성분 분석법 및 진폭투시법에 의해 전체 특성백터를 구성한다. 이렇게 구성된 특성백터들의 공분산을 구한 후, 판별단계에서 베이즈 방법에 의해 구해진 확률분포함수를 이용하여 정면얼굴의 눈 영역 부분을 검출한다. 또한 본 논문에서 제안한 판별 알고리즘을 이용하여 입력영상의 눈영역을 찾기 위한 실험식도 제안하였다. 모의 실험결과 정면얼굴에서의 눈검출율은 매우 높았으며 계산을 위한 특성백터 또한 적음으로써 실시간 특성을 요하는 임베디드시스템에 적용 가능함을 알 수 있었다.

• 전기학회논문지
• /
• 제63권6호
• /
• pp.813-819
• /
• 2014

The purpose of this study was to extract accurate parameters of facial movement features using 3-D motion capture system in speech recognition technology through lip-reading. Instead of using the features obtained through traditional camera image, the 3-D motion system was used to obtain quantitative data for actual facial movements, and to analyze 11 variables that exhibit particular patterns such as nose, lip, jaw and cheek movements in monosyllable vocalizations. Fourteen subjects, all in 20s of age, were asked to vocalize 11 types of Korean vowel monosyllables for three times with 36 reflective markers on their faces. The obtained facial movement data were then calculated into 11 parameters and presented as patterns for each monosyllable vocalization. The parameter patterns were performed through learning and recognizing process for each monosyllable with speech recognition algorithms with Hidden Markov Model (HMM) and Viterbi algorithm. The accuracy rate of 11 monosyllables recognition was 97.2%, which suggests the possibility of voice recognition of Korean language through quantitative facial movement analysis.

• 디지털콘텐츠학회 논문지
• /
• 제10권2호
• /
• pp.319-326
• /
• 2009

본 논문에서는 별도의 센서를 부착하지 않고 영상만을 이용하여 실시간으로 손 형상을 인식하는 알고리즘에 대해 기술한다. 손은 형상이 매우 복잡하기 때문에 2차원 형상의 불변량에 해당하는 에지의 방향성 히스토그램을 이용하여 인식을 행한다. 이 방법은 복잡한 배경에서 피부색을 지닌 손 영역이 정확히 추출되며 손 형상을 인식하는데 있어서 수행속도가 빠르고 조명변화에 덜 민감하기 때문에 실시간 손 형상 인식에 적합하다. 본 논문에서는 손의 형상에서 방향이 틀어지는 경우에도 인식을 가능하게 하기위해 주성분 분석법을 사용하여 인식오차를 줄이는 방법을 기술한다. 이 방법을 사용함으로써 손 영상이 3차원적으로 회전에 의해 변하는 경우도 인식가능하게 되었다. 본 논문에서 제안하는 방법은 가정용 가전제품이나 게임을 제어하는 실시간 휴먼 인터페이스 제작에 사용 할 수 있다.

• 한국자동차공학회논문집
• /
• 제12권1호
• /
• pp.39-47
• /
• 2004

The objective of this study is to analyse the spray characteristics according to the injection duration under ambient pressure condition and to investigate the relationship between vorticity and entropy for controlling diffusion process that is the most important thing during the intake stroke injection process. Therefore, the spray velocity was obtained by using the PIV method that has been an useful optical diagnostics technology, and vorticity calculated from spray velocity component with vorticity algorithm. In addition, the homogeneous diffusion rate of spray was quantified by using the entropy analysis based on the Boltzmann's statistical thermodynamics. From these method, we found that as injection duration increases, spray velocity increases and the location of vortex is moved to the downstream of spray. In the same condition, as the entropy decrease, mean vorticity increases. This means that the concentration of spray droplets caused by the increase of injection duration is more effective than the increase of momentum dissipation.

• 정보과학회 논문지
• /
• 제42권7호
• /
• pp.868-877
• /
• 2015

영상에서 바코드 영역을 검출하는 다양한 방식들이 연구되어 왔다. 기존 방식들은 주파수 성분 특징을 이용하거나, Hough transform (HT)을 이용하여 바코드 영역을 검출한다. 하지만 이 방식들은 바코드의 기울기와 노이즈에 영향을 받는다. 또한 여러 개의 바코드가 있는 경우 정확히 검출하지 못한다. 본 논문에서는 바코드의 기울기와 노이즈에 강인하고, 복수 개의 바코드를 검출할 수 있는 방식을 제안한다. 우리는 전처리 단계로 Probabilistic Hough transform (PHT)를 이용하여 바코드 기울기, 노이즈, 그리고 개수에 상관없이 바코드가 존재할 가능성이 높은 영역을 추출한 후, 주파수 성분 분석을 통해 바코드를 찾아낸다. 구현된 시스템의 성능분석을 통해 다양한 환경에서 바코드 추출이 가능함을 확인했다.

• Biomedical Engineering Letters
• /
• 제8권4호
• /
• pp.373-382
• /
• 2018

Since epileptic seizure is unpredictable and paroxysmal, an automatic system for seizure detecting could be of great significance and assistance to patients and medical staff. In this paper, a novel method is proposed for multichannel patient-specific seizure detection applying the earth mover's distance (EMD) in scalp EEG. Firstly, the wavelet decomposition is executed to the original EEGs with five scales, the scale 3, 4 and 5 are selected and transformed into histograms and afterwards the distances between histograms in pairs are computed applying the earth mover's distance as effective features. Then, the EMD features are sent to the classifier based on the Bayesian linear discriminant analysis (BLDA) for classification, and an efficient postprocessing procedure is applied to improve the detection system precision, finally. To evaluate the performance of the proposed method, the CHB-MIT scalp EEG database with 958 h EEG recordings from 23 epileptic patients is used and a relatively satisfactory detection rate is achieved with the average sensitivity of 95.65% and false detection rate of 0.68/h. The good performance of this algorithm indicates the potential application for seizure monitoring in clinical practice.

• 방송공학회논문지
• /
• 제26권1호
• /
• pp.79-87
• /
• 2021

본 논문에서는 소량의 레이블 데이터로 딥러닝 기반 분류 모델을 훈련할 때 적용되는 준 지도 학습 기법 (semi-supervised learning: SSL)에 대해서 분석한다. 기존의 준 지도 학습 기법은 크게 일관성 정규화 (consistency regularization), 엔트로피 기반 (entropybased), 의사 레이블링 (pseudo labeling)으로 구분할 수 있다. 우선, 각 준 지도 학습 기법의 알고리즘에 대해서 서술한다. 실험에서는 준 지도학습 기법을 레이블 데이터의 수를 변화시키면서 훈련 후 분류 정확도를 평가한다. 최종적으로 실험 결과를 바탕으로 기존 준 지도 학습 기법의 한계에 대해서 서술하고, 분류 성능을 향상하기 위한 연구 방향을 제시한다.