• Annual Conference on Human and Language Technology
• /
• 2004.10d
• /
• pp.151-156
• /
• 2004

여러 개의 절로 이루어진 긴 문장에서 절 단위를 인식해냄으로써 구문분석의 복잡도를 크게 줄일 수 있다. 본 논문에서는 SVM 모델을 이용하여 한국어 문장에서 절의 경계를 인식하는 방법을 제안하였다. 첫 번째 단계로 중심어가 후행하는 한국어 문장의 특성을 고려하여 절의 끝점을 먼저 찾고, 첫 번째 단계의 결과인 절의 끝점 정보와 절의 끝점 인식을 위한 정보보다 더 전역적인 정보를 이용해 절의 시작점을 인식하는 두 번째 단계로 나누어 진행하였다. 구문구조 부착 말뭉치를 이용하여 학습하고 실험한 결과, F-score 86.87%와 단어 단위의 정확도 96.63%의 성능을 나타내었다.

• Proceedings of the IEEK Conference
• /
• 2006.06a
• /
• pp.205-206
• /
• 2006

In this paper, we have proposed an ad-hoc routing protocol named "RRMR (Reliable and Resilient Multipath Routing)" that is expanded from AODV to acquire multiple path at one time. This protocol needs less overhead to configure data routing paths and the paths are maintained to transmit data stably. To get these advantages, RRMR protocol is constructed by expanding from multipath acquiring and maintaining method based on distance vector method and hop-by-hop routing of AODV.

• Journal of the Korea Computer Graphics Society
• /
• v.11 no.1
• /
• pp.40-49
• /
• 2005

Several optimization techniques have been proposed for volume ray casting, but these cannot achieve real-time frame rates. In addition, it is difficult to apply them to some applications that require perspective projection. Recently, hardware-based methods using 3D texture mapping are being used for real-time volume rendering. Although rendering speed approaches real time, the larger volumes require more swapping of volume bricks for the limited texture memory. Also, image quality deteriorates compared with that of conventional volume ray casting. In this paper, we propose a data structure for real-time volume ray casting named PERM (Precomputed dEnsity and gRadient Map). The PERM stores interpolated density and gradient vector for quantized cells. Since the information requiring time-consuming computations is stored in the PERM, our method can ensure interactive frame rates on a consumer PC platform. Our method normally produces high-quality images because it is based on conventional volume ray casting.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.12
• /
• pp.125-130
• /
• 2016

Automatic quality inspection by computer vision can be applied and give a solution to the pharmaceutical industry field. Pharmaceutical capsule can be easily affected by flaws like dents, cracks, holes, etc. In order to solve the quality inspection problem, it is required computationally efficient image processing technique like thresholding, boundary edge detection and segmentation and some automated systems are available but they are very expensive to use. In this paper, we have developed a dented capsule image processing technique using edge-based image segmentation, TLS(Total Least Squares) curve fitting technique and adopted low cost camera module for capsule image capturing. We have tested and evaluated the accuracy, training and testing time of the classification recognition algorithms like PCA(Principal Component Analysis), ICA(Independent Component Analysis) and SVM(Support Vector Machine) to show the performance. With the result, PCA, ICA has low accuracy, but SVM has good accuracy to use for classifying the dented capsule.

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.9
• /
• pp.19-25
• /
• 2016

This paper is to recognize and play tab chords from guitar musical sheets. The musical chord area of an input image is segmented by changing the image in saturation and applying the Grabcut algorithm. Based on a template matching, our approach detects tab starting sections on a segmented musical area. The virtual block method is introduced to search blanks over chord lines and extract tab fret segments, which doesn't cause the computation loss to remove tab lines. In the experimental tests, the prototype based classification outperforms Bayesian method and the nearest neighbor rule with the whole set of training data and its performance is similar to that of the support vector machine. The experimental result shows that the prediction rate is about 99.0% and the number of selected prototypes is below 3.0%.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2015.07a
• /
• pp.328-331
• /
• 2015

본 논문은 EBME(Extended Bilateral Motion Estimation) 알고리듬에서 움직임 벡터들의 방향과 크기를 고려한 알고리듬을 제안하였다. EBME는 높은 연산량을 요구하기 때문에 프레임 내의 x, y방향 각각의 평균 움직임 벡터크기를 이용하여 동적 프레임과 정적프레임을 판단하고, EBME 수행여부를 결정하여 연산량을 줄인다. 또한 동일한 움직임 벡터들의 방향과 크기를 비교하여 MVS(Motion Vector Smoothing)단계 수행여부를 판단함으로써 연산량을 줄인다. 제안하는 알고리듬을 적용한 실험 결과 기존의 EBME 알고리듬에 비해 수행시간은 단축되었으나 PSNR(Peak Signal to Noise Ratio)은 향상 되었다.

• Annual Conference of KIPS
• /
• 2004.11a
• /
• pp.855-858
• /
• 2004

본 논문에서는 원영상 영역내 포함된 우성의 에지에 대한 구체적 정보를 이용하기 위하여 Haar 웨이블릿을 이용한 에지영상 추출한다. 추출된 에지영상에 얼굴영역을 검출하기위해 이진화된 영상에 설정된 임계값을 통하여 얻은 이진영상으로부터 얼굴영역을 검출하기 위하여 얼굴의 일반적인 구조적 정보와 처리시간이 빠른 수평, 수직히스토그램 분석법을 이용하였다. 얼굴영역을 분리한 영상에 얼굴영역의 특징벡터를 구하기 위하여 26개의 특징벡터를 사용한 효율적인 고차 국소 자동 상관함수를 사용하였다. 계산된 특징벡터는 BP 신경망의 학습을 통하여 얼굴인식을 위한 데이터로 사용하여 제안된 알고리즘에 의한 인식률향상과 속도 향상을 입증한다.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.22 no.6
• /
• pp.547-550
• /
• 2017

This paper presents a loss analysis and compares three pulse width modulation (PWM) methods applied in a three-phase grid-connected bidirectional inverter for an energy storage system. The losses in switching devices and output low pass filters are theoretically analyzed by using PWM control techniques. Grid-connected bidirectional inverters are designed by using PWM techniques, and the designed inverters are simulated to verify the analysis results.

• Journal of Information Processing Systems
• /
• v.14 no.1
• /
• pp.218-227
• /
• 2018

Vocal effort detection is important for both robust speech recognition and speaker recognition. In this paper, the spectral information entropy feature which contains more salient information regarding the vocal effort level is firstly proposed. Then, the model fusion method based on complementary model is presented to recognize vocal effort level. Experiments are conducted on isolated words test set, and the results show the spectral information entropy has the best performance among the three kinds of features. Meanwhile, the recognition accuracy of all vocal effort levels reaches 81.6%. Thus, potential of the proposed method is demonstrated.

• Annual Conference of KIPS
• /
• 2015.10a
• /
• pp.1756-1759
• /
• 2015

웨어러블 디바이스의 등장과 여러종류의 센서탑재로 행위데이터를 수집하는것이 수월해졌다. 행위패턴 모델링에 앞서 사용자의 행위에따른 신호변화와 신호패턴을 파악하기위해 분석을 실시하였다. Moto360의 3축 가속도 센서를 이용 사용자에 행위에대한 센서신호값을 수집하여 행위에따른 신호값을 수집하였으며, 수집된 신호값과 신호값으로부터 SVM(Signal Vector Magnitude)값을 구해 사용자의 각 행위들에 대해 신호값과 SVM값의 특징을 분석하여 측정 신호값으로부터 행위를 인식할수 있도록 시도하였다.