• 한국정보과학회논문지:소프트웨어및응용
• /
• 제36권12호
• /
• pp.967-976
• /
• 2009

본 논문에서는 CT 및 CT 혈관 조영 영상에서 강체 정합 및 뼈 소거법을 이용한 하지 혈관 자동 추출 방법을 제안한다. 첫째, 뼈의 부분적인 움직임을 반영하기 위하여 해부학 정보를 바탕으로 하지를 자동 구역화하고, 둘째, CT와 CTA 영상간 움직임을 산정하기 위하여 거리지도 기반의 강체 정합을 수행한다. 셋째, CTA 영상에서 복잡한 구조를 갖는 뼈를 제거하고 뼈에 인접한 혈관이 깎이는 것을 방지하기 위하여 뼈 소거법과 혈관 마스킹 기법을 제안한다. 넷째, 정합오차 및 연골 등의 잡음을 줄이기 위하여 혈관 추적 기반의 후 처리 과정을 통하여 보정한다. 제안 방법의 평가를 위해 육안 평가와 정확성 평가 그리고 수행시간을 측정하였다. 육안 평가를 위해 차감 기법, 정합 후 차감 기법, 제안 방법을 적용한 결과를 볼륨렌더링과 최대 강도 투영영상을 사용하여 비교하였다. 정확성 평가를 위해 CTA 영상과 차감 기반 기법 및 제안 방법을 적용한 결과의 밝기값 분포도를 분석하였다. 실험 결과 뼈는 제거되고 가는 혈관 및 다른 조직의 손실 없이 혈관이 정확하게 추출되었음을 볼 수 있었고, 13명의 환자 데이터 전채에 대한 전체 수행시간은 약 40포 정도로 측정되었다.

• 한국음향학회지
• /
• 제15권5호
• /
• pp.38-43
• /
• 1996

본 논문에서는 잡음 및 주변 환경에 강인한 것으로 알려져 있는 특징 벡터들을 이용한 인식 성능을 비교하였다. 아울러 스펙트럼 차감법을 적용하여 높은 인식 성능을 얻도록 하였다. 본 논문에서는 환경 변화에 강인한 인식 성능을 얻기 위하여 SMC(Short time Modified Coherence) 분석, 루트(root) 켑스트럼 분석, LDA(Linear Discriminant Analysis), PLP(Perceptual Linear Prediction), RASTA(RelAtive SpecTrAl) 처리 등을 이용하여 인식 실험을 수행하였다. 실험을 위하여 반연속 HMM을 이용한 단독음 인식 시스템을 구현하였고 전시장 및 컴퓨터실의 잡음을 첨가하여 0, 10 및 20dB의 SNR에 대한 인식 실험을 수행하였다. 실험 결과, LPCC(Linear Prediction Cepstral Coefficient)를 이용한 경우에 비하여 SMC나 루트처리를 이용한 멜 켑스트럼(루트_멜 켑스트럼)을 이용한 경우 10dB의 SNR에서 각각 9.86%, 12.68% 향상된 가장 좋은 인식률을 얻었다. 또한 멜 켑스트럼과 루트_멜 켑스트럼을 스펙트럼 차감법과 결합하여 잡음을 제거한 경우 10dB에서 각각 16.7%, 8.4% 향상된 94.91%, 94.28%의 인식률을 얻을 수 있었다.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.628-631
• /
• 2003

In this paper, we present real-time, accurate face region detection and tracking technique for an intelligent surveillance system. It is very important to obtain the high-resolution images, which enables accurate identification of an object-of-interest. Conventional surveillance or security systems, however, usually provide poor image quality because they use one or more fixed cameras and keep recording scenes without any cine. We implemented a real-time surveillance system that tracks a moving person using four pan-tilt-zoom (PTZ) cameras. While tracking, the region-of-interest (ROI) can be obtained by using a low-pass filter and background subtraction. Color information in the ROI is updated to extract features for optimal tracking and zooming. The experiment with real human faces showed highly acceptable results in the sense of both accuracy and computational efficiency.

• 소성∙가공
• /
• 제6권5호
• /
• pp.400-407
• /
• 1997

Geometric mapping technique has been used to find the shape of initial blank for sheet forming. The method was chosen because of its simplicity and numerical efficiency. Error in blank shape were measured along deformation path by FE analysis of forming. Blank shape was modified by volume additionaddition/Subtractiontraction method with taking with taking into account of deformation path. Modified blank shape shows an acceptable result, showing the current method can be an useful tool for predicting blank shape in the practical application. More test will be done to verify the validity of the method.

• 제어로봇시스템학회논문지
• /
• 제16권12호
• /
• pp.1137-1142
• /
• 2010

This paper studied the speech parameters less affected by the human emotion for the development of the robust speech recognition system. For this purpose, the effect of emotion on the speech recognition system and robust speech parameters of speech recognition system were studied using speech database containing various emotions. In this study, mel-cepstral coefficient, delta-cepstral coefficient, RASTA mel-cepstral coefficient and frequency warped mel-cepstral coefficient were used as feature parameters. And CMS (Cepstral Mean Subtraction) method were used as a signal bias removal technique. Experimental results showed that the HMM based speaker independent word recognizer using vocal tract length normalized mel-cepstral coefficient, its derivatives and CMS as a signal bias removal showed the best performance of 0.78% word error rate. This corresponds to about a 50% word error reduction as compare to the performance of baseline system using mel-cepstral coefficient, its derivatives and CMS.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 신호처리소사이어티 추계학술대회 논문집
• /
• pp.317-320
• /
• 2003

In this paper, we present real-time, accurate face region detection and tracking technique for an intelligent surveillance system. It is very important to obtain the high-resolution images, which enables accurate identification of an object-of-interest. Conventional surveillance or security systems, however, usually provide poor image quality because they use one or more fixed cameras and keep recording scenes without any clue. We implemented a real-time surveillance system that tracks a moving person using pan-tilt-zoom (PTZ) cameras. While tracking, the region-of-interest (ROI) can be obtained by using a low-pass filter and background subtraction. Color information in the ROI is updated to extract features for optimal tracking and zooming. The experiment with real human faces showed highly acceptable results in the sense of both accuracy and computational efficiency.

• 천문학회보
• /
• 제37권1호
• /
• pp.76.1-76.1
• /
• 2012

One of the main challenges for future 21-cm observations is to remove foregrounds which are several orders of magnitude more intense than the HI signal. We propose a new technique for removing foregrounds of the redshifted 21-cm observations. We consider multi-frequency interferometer observations. We assume that the 21-cm signals in different frequency channels are uncorrelated and the foreground signals change slowly as a function of frequency. When we add the visibilities of all channels, the foreground signals increase roughly by a factor of N because they are highly correlated. However, the 21-cm signals increase by a factor of sqrt{N} because the signals in different channels contribute randomly. This enables us to obtain an accurate shape of the foreground angular power spectrum. Then, we obtain the 21-cm power spectrum by subtracting the foreground power spectrum obtained this way. We describe how to obtain the average power spectrum of the 21-cm signal.

• 전기학회논문지
• /
• 제64권7호
• /
• pp.1047-1055
• /
• 2015

We proposes a method for improving the identification and tracking of the moving objects in intelligent video surveillance system. The proposed method consists of 3 parts: object detection, object recognition, and object tracking. First of all, we use a GMM(Gaussian Mixture Model) to eliminate the background, and extract the moving object. Next, we propose a labeling technique forrecognition of the moving object. and the method for identifying the recognized object by using the optical flow and EMD algorithm. Lastly, we proposes method to track the location of the identified moving object regions by using location information of moving objects and Kalman filter. Finally, we demonstrate the feasibility and applicability of the proposed algorithms through some experiments.

• Journal of Chest Surgery
• /
• 제25권1호
• /
• pp.105-111
• /
• 1992

Total atherosclerotic obstruction of the juxtarenal abdominal aorta is a relatively rarely encountered form of atherosclerotic vascsular disease, accounting for less than 5% of all arterial obstrutive disease. We have encountered two patients with such lesions, both of whom were admitted for intermittent claudication of the lower extremities and symptoms of vascular ischemia. Digital subtraction angiography[DSA] was performed on both patients, the results of which revealed total obstruction of the aorta just inferior to the renal arteries without involving the latter. Operative technique involved the use of the sup-raceliac aorta as the site of proximal anastomosis of aortofemoral bypass followed by a fem-orofemoral bypass graft with Smm sized Woven Dacron[Vascutek] through a subcutaneous tunnel within the retroperitoneal space. Both patients experienced restoration of blood flow distal to the obstruction postoperatively without any complications, and OPD follow-up one month postoperatively and postoperative DSA showed evidence of continued graft patency with persistent symptomatic improvement.

• 대한전자공학회논문지TC
• /
• 제40권6호
• /
• pp.213-224
• /
• 2003

제안된 알고리듬은 채널 잡음전력에 따라 적응적으로 문턱값 레벨을 결정하는 적응 문턱값 설정방식을 사용하여 채널특성에 관계없이 항상 최적의 문턱값을 설정한다. 또한, 이것은 그 추정성능이 다중경로채널에서의 전력지연 프로파일 변동에 대해 덜 민감하도록 만드는 특별하게 설계된 훈련심볼을 사용한다. 그 결과, 제안된 기법의 추정성능은 채널특성 변동에 대해 영향을 적게 받는다.