• Korean Journal of Remote Sensing
• /
• v.23 no.2
• /
• pp.113-117
• /
• 2007

In this paper, we propose a new method for determining automatically the size of the matching window using histogram of the gradient in order to improve the performance of stereo matching using one-meter resolution satellite imagery. For each pixel, we generate Flatness Index Image by calculating the mean value of the vertical or horizontal intensity gradients of the 4-neighbors of every pixel in the entire image. The edge pixel has high flatness index value, while the non-edge pixel has low flatness index value. By using the histogram of the Flatness Index Image, we find a flatness threshold value to determine whether a pixel is edge pixel or non-edge pixel. If a pixel has higher flatness index value than the flatness threshold value, we classify the pixel into edge pixel, otherwise we classify the pixel into non-edge pixel. If the ratio of the number of non-edge pixels in initial matching window is low, then we consider the pixel to be in homogeneous region and enlarge the size of the matching window We repeat this process until the size of matching window reaches to a maximum size. In the experiment, we used IKONOS satellite stereo imagery and obtained more improved matching results than the matching method using fixed matching window size.

• The KIPS Transactions:PartD
• /
• v.12D no.6 s.102
• /
• pp.863-868
• /
• 2005

Hardware and software codesign framework called PeaCE(Ptolemy extension as a Codesign Environment) was developed. It allows to express both data flow and control flow which is described as fFSM which extends traditional finite state machine. While the fFSM model provides lots of syntactic constructs for describing control flow, it has a lack of their formality and then difficulties in verifying the specification. In order to define the formal semantics of the fFSM, in this paper, firstly the hierarchical structure in the model is flattened and then the step semantics is defined. As a result, some important bugs such as race condition, ambiguous transition, and circulartransition can be formally detected in the model.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.7
• /
• pp.665-670
• /
• 2002

The most common form of pre-emphasis is y(n)=s(n)-As(n-1), where A typically lies between 0.9 and 1.0 in voiced signal. Also, this value reflects the degree of pre-emphasis and equals R(1)/R(0) in conventional method. This paper proposes a new flattening method to compensate the weaked high frequency components that occur by vocal cord characteristic. We used interval information of LSP to estimate formant frequency, After obtaining the value of slope and inverse slope using linear interpolation among formant frequency, flattening process is followed. Experimental results show that the proposed method flattened the weaked high frequency components effectively. That is, we could improve the flattening characteristics by using interval information of LSP as flattening factor at the process that compensates weaked high frequency components.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.7 no.5
• /
• pp.907-912
• /
• 2006

Pre-emphasizing the speech compensates for falloff at high frequencies. The most common form of pre-emphasis is y(n)=s(n)-A${\cdot}$s(n-1), where A typically lies between 0.9 and 1.0 in voiced signal. And, this value reflects the degree of pre-emphasis and equals R(1)/R(0) in conventional method. This paper proposes a new flattening method to compensate the weaked high frequency components that occur by vocal cord characteristic. We used QMF(Quardrature Mirror Filter) to minimize the output signal distortion. After using the QMF to compensate high frequency components, flattening process is followed by R(1)/R(0) at each frame. Experimental results show that the proposed method flattened the weaked high frequency components effectively than auto correlation method. Therefore, the flattening algorithm will apply in speech signal processing like speech recognition, speech analysis and synthesis.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.105-106
• /
• 2004

음성인식, 합성 및 분석과 같은 음성신호처리 분야에 있어서 피치검출이나 포만트검출은 매우 중요하다. 주파수 영역의 스펙트럼 신호는 잡음이 부가되는 경우에도 고조파정보와 포만트 포락선 정보를 유지하기 때문에 음성신호처리분야에서 매우 유용하다고 할 수 있다. 고조파 정보나 포만트 포락선 정보는 피치검출과 포만트 주파수 검출에 직접 이용된다 하지만 두 성분을 분리하는 방법에 따라 피치검출이나 포만트 주파수 검출에 영향을 미칠 수 있으므로 기존의 방법보다 두 성분을 더 잘 분리할 수 있는 방법이 필요한 것이다. 본 논문에서는 스펙트럼 신호를 최대한 평탄화시킴으로써 포만트의 영향을 제거하고 고조파 성분을 분리해 내어 이를 피치검출에 사용한다. LSP를 이용하여 적응적 밴드에서 평탄화를 시도하고 이를 피치 검출에 이용하였다.

• Annual Conference on Human and Language Technology
• /
• 2002.10e
• /
• pp.118-123
• /
• 2002

많은 자연언어처리 분야에서 문장의 단위화는 기본적인 처리 단계로서 중요한 위치를 차지하고 있다. 한국어 단위화에 대한 기존 연구들은 규칙 기반 방법이나 기계 학습 기법을 이용한 것이 대부분이었다. 본 논문에서는 통계 기반 방식의 일환으로 순수 확률기반 모델을 이용한 단위화 방법을 제시한다. 확률 기반 모델은 처리하고자 하는 해당 언어에 대한 깊은 지식 없이도 적용 가능하다는 장점을 가지므로 다양한 언어의 단위화에 대한 기본 모델로서 이용될 수 있다. 또한 자료 부족 문제를 해결하기 위해 메모리 기반 학습 시에 사용하는 IG back-off 평탄화 방식을 시스템에 적용하였다. 본 논문의 모텔을 적용한 단위화 시스템을 이용하여 한국어와 영어에 대해 실험한 결과 비교적 작은 규모의 말뭉치를 학습하였음에도 불구하고 각각 90.0%, 90.0%의 정확도를 보였다.

• Journal of Navigation and Port Research
• /
• v.31 no.10
• /
• pp.833-837
• /
• 2007

In this study, we have fabricated a gain flattened erbium-doped optical fiber amplifier. Gain flattening filters were realized by the strain-induced long period fiber gratings, which are made of periodically arrayed metal wires. Using the filter of $550{\mu}m$ period, spontaneous emission amplified at C-band wavelength by a 980nm pumping laser was flattened within 1dB of gain ripple. The performance of the simultaneous multi channel amplification was measured using a fabry-perot laser diode. Amplification ratio was above 20dB. This amplifier can be applied to the long distance transmission system based on a wavelength division multiplexing for boosting an attenuated signal.

• Korean Chemical Engineering Research
• /
• v.53 no.4
• /
• pp.517-523
• /
• 2015

As the aluminum (Al) metallization process was replaced with copper (Cu), the damascene process was introduced, which required the planarization step to eliminate over-deposited Cu with Chemical Mechanical Polishing (CMP) process. In this study, the verification of the corrosion inhibitors, one of the Cu CMP slurry components, was conducted to find out the tendency regarding the carboxyl and amino functional group in neutral environment. Through the results of etch rate, removal rate, and chemical ability of corrosion inhibitors based on 1H-1,2,4-triazole as the base-corrosion inhibitor, while the amine functional group presents high Cu etching ability, carboxyl functional group shows lower Cu etching ability than base-corrosion inhibitor which means that it increases passivation effect by making strong passivation layer. It implies that the corrosion inhibitor with amine functional group was proper to apply for 1st Cu CMP slurry owing to the high etch rate and with carboxyl functional group was favorable for the 2nd Cu CMP slurry due to the high Cu removal rate/dissolution rate ratio.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.46 no.6
• /
• pp.132-145
• /
• 2009

We present the framework for JND based Illumination and Color Restoration Using Edge-preserving filter for restoring distorted images taken under the arbitrary lighting conditions. The proposed method is effective for appropriate illumination compensation, vivid color restoration, artifacts suppression, automatic parameter estimation, and low computation cost for HW implementation. We show the efficiency of the mean shift filter and sigma filter for illumination compensation with small spread parameter while considering the processing time and removing the artifacts such as HALO and noise amplification. The suggested CRF (color restoration filter) can restore the natural color and correct color distortion artifact more perceptually compared with current solutions. For the automatic processing, the image statistics analysis finds suitable parameter using JND and all constants are pre-defined. We also introduce the ROI-based parameter estimation dealing with small shadow area against spacious well-exposed background in an image for the touch-screen camera. The object evaluation is performed by CMC, CIEde2000, PSNR, SSIM, and 3D CIELAB gamut with state-of-the-art research and existing commercial solutions.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2008.06a
• /
• pp.539-539
• /
• 2008

본 논문에서는 투명전도박막의 균일한 표면특성을 확보하기 위해 광역평탄화 공정을 적용하여 투명전도 박막의 표면 거칠기를 연구하였으며 슬러리의 종류에 따른 박막의 연마특성을 연구하였다. 본 실험에서 사용된 ITO 박막은 RF Sputtering에 의해 제작되었고 하부 기판은 석영 Glass가 사용되었다. 광역평탄화를 위한 CMP 공정은 고분자 물질계열의 패드위에 슬러리입자를 공급하고 웨이퍼 캐리어에 하중을 가하며 웨이퍼의 표면을 연마하는 방법으로 가공물을 탄성패드에 누르면서 상대 운동시켜 가공물과 친화력이 우수한 부식액으로 화학적 제거를 함과 동시에 초미립자로 기계적 제거를 하는 것이다. ITO 박막의 평탄화를 위한 공정조건은 Polisher pressure 300 g/$cm^2$, 슬러리 유속 80 ml/min, 플레이튼속도 60 rpm으로 하였다. 위의 조건에 따라 공정을 진행 후 연마특성을 측정하였으며 이때 사용된 슬러리는 산화막에 사용되는 실리카슬러리와 금속연마용 슬러리인 EPL을 사용하였다. 연마율은 실리카 슬러리가 EPL슬러리에 비해 높음을 확인 하였다. CMP 공정에 의해 평탄화를 수행 할 경우 실리카슬러리와 EPL슬러리 모두 CMP전에 비해 돌출된 힐록들이 감소되었음을 알 수 있었다. 비균일도 특성은 모든 슬러리가 양호한 특성을 나타내었다. 평탄화된 박막의 표면과 거칠기 특성은 AFM(XE-200, PSIA Company) 을 이용하여 분석을 하였다.