• 한국멀티미디어학회논문지
• /
• 제11권9호
• /
• pp.1302-1309
• /
• 2008

K-means나 퍼지 군집화와 같은 전통적인 군집화 기법들이 원형(prototype)을 기반으로 하고 볼록한 형태의 집단들에 적합한 반면, 스펙트럼 군집화(spectral clustering)는 국부적인 유사성을 기반으로 전역적인 집단을 찾아내는 기법으로 오목한 형태의 집단들에도 적용할 수 있어 커널을 기반으로 하는 SVM과 더불어 각광을 받고 있다. 하지만 SVM이 그러하듯이 스펙트럼 군집화에서도 커널의 폭은 성능에 지대한 영향을 끼치는 요인으로, 이를 결정하기 위한 다양한 방법이 시도되었지만 여전히 휴리스틱에 의존하는 실정이다. 이 논문에서는 유사도 행렬이 보다 명백한 블록 대각 형태를 가지도록 하기 위해 국부적인 커널의 폭을 거리 히스토그램을 바탕으로 적응적으로 결정하는 방법을 제시한다. 제안한 방법은 스펙트럼 군집화에 사용되는 유사도 행렬(affinity matrix)이 블록 형태의 대각 행렬을 이룰 때 이상적인 결과를 낸다는 사실에 기반하고 있으며, 이를 위해서 전통적인 유클리디안 거리와 무작위 행보 거리(random walk distance)를 함께 사용한다. 제안한 방법은 기존의 방법들에서 사용하는 유사도 행렬에 비해 명확한 블록 대각 행렬을 나타내고 있음을 실험 결과를 통해 확인할 수 있다.

• Wind and Structures
• /
• 제9권4호
• /
• pp.255-270
• /
• 2006

Several analytical procedures available in literature, for the evaluation of wind induced fatigue damage of structures, either assume the wide band random stress variations as narrow band random process or use correction factors along with narrow band assumption. This paper compares the correction factors obtained using the Rainflow Cycle (RFC) counting of the measured stress time histories on a lamp mast and a lattice tower, with those evaluated using different frequency domain methods available in literature. A Bi-modal spectral method has been formulated by idealising the single spectral moment method into two modes of background and resonant components, as considered in the gust response factor, for the evaluation of fatigue of slender structures subjected to "along-wind vibrations". A closed form approximation for the effective frequency of the background component has been developed. The simplicity and the accuracy of the new method have been illustrated through a case study by simulating stress time histories at the base of an urban light pole for different mean wind speeds. The correction factors obtained by the Bi-modal spectral method have been compared with those obtained from the simulated stress time histories using RFC counting method. The developed Bi-modal method is observed to be a simple and easy to use alternative to detailed time and frequency domain fatigue analyses without considerable computational and experimental efforts.

• ETRI Journal
• /
• 제35권2호
• /
• pp.311-320
• /
• 2013

In most spectral clustering approaches, the Gaussian kernel-based similarity measure is used to construct the affinity matrix. However, such a similarity measure does not work well on a dataset with a nonlinear and elongated structure. In this paper, we present a new similarity measure to deal with the nonlinearity issue. The maximum flow between data points is computed as the new similarity, which can satisfy the requirement for similarity in the clustering method. Additionally, the new similarity carries the global and local relations between data. We apply it to spectral clustering and compare the proposed similarity measure with other state-of-the-art methods on both synthetic and real-world data. The experiment results show the superiority of the new similarity: 1) The max-flow-based similarity measure can significantly improve the performance of spectral clustering; 2) It is robust and not sensitive to the parameters.

• Journal of the Optical Society of Korea
• /
• 제20권6호
• /
• pp.752-761
• /
• 2016

Hyperspectral images are often contaminated with stripe noise, which severely degrades the imaging quality and the precision of the subsequent processing. In this paper, a variational model is proposed by employing spectral-spatial adaptive unidirectional variation and a sparse representation. Unlike traditional methods, we exploit the spectral correction and remove stripes in different bands and different regions adaptively, instead of selecting parameters band by band. The regularization strength adapts to the spectrally varying stripe intensities and the spatially varying texture information. Spectral correlation is exploited via dictionary learning in the sparse representation framework to prevent spectral distortion. Moreover, the minimization problem, which contains two unsmooth and inseparable $l_1$-norm terms, is optimized by the split Bregman approach. Experimental results, on datasets from several imaging systems, demonstrate that the proposed method can remove stripe noise effectively and adaptively, as well as preserve original detail information.

• 한국방재안전학회논문집
• /
• 제10권2호
• /
• pp.21-28
• /
• 2017

시공간적으로 불규칙하게 작용하는 변동 풍속 난류의 자료는 풍공학적으로 돌풍계수 평균풍속 변동 풍하중등의 계산에서 요구되지만, 내풍 및 사용성에 따른 동적응답의 평가에서는 변동 풍속의 파워 스펙트럴 밀도함수가 요구된다. 본 논문에서는 1987-2016.12.1일까지의 일순간최대풍속 자료를 확률과정으로 가정했고, 이 실측된 자료와 확률이론을 근거로 평균류방향 파워 스펙트럴 밀도 함수에 대한 기초적 자료를 얻고자 대표지점(6개 지점)을 선정했다. 선정된 각 지점에 대한 일순간최대풍속자료는 기상청으로부터 획득했다. 해석결과 본 논문에서 평가된 스펙트럼 모델은 저진동수 영역에서는 Solari, 고진동수 영역에서는 von Karman의 모델과 근접한 현상을 나타냈다.

• 대한원격탐사학회지
• /
• 제35권4호
• /
• pp.561-571
• /
• 2019

In this paper, we analyze the effect of the representative spectral indices, normalized difference vegetation index (NDVI), normalized difference water index (NDWI) and normalized difference built-up index (NDBI) on classification accuracies of Landsat-8 OLI image.After creating these spectral index images, we propose five methods to select the spectral index images as classification features together with Landsat-8 OLI bands from 1 to 7. From the experiments we observed that when the spectral index image of NDVI or NDWI is used as one of the classification features together with the Landsat-8 OLI bands from 1 to 7, we can obtain higher overall accuracy and kappa coefficient than the method using only Landsat-8 OLI 7 bands. In contrast, the classification method, which selected only NDBI as classification feature together with Landsat-8 OLI 7 bands did not show the improvement in classification accuracies.

• Current Optics and Photonics
• /
• 제5권3호
• /
• pp.322-328
• /
• 2021

In this paper, a pretreatment method for a matrix in convex optimization is proposed to optimize the spectral reconstruction process of a disordered dispersion spectrometer. Unlike the reconstruction process of traditional spectrometers using Fourier transforms, the reconstruction process of disordered dispersion spectrometers involves solving a large-scale matrix equation. However, since the matrices in the matrix equation are obtained through measurement, they contain uncertainties due to out of band signals, background noise, rounding errors, temperature variations and so on. It is difficult to solve such a matrix equation by using ordinary nonstationary iterative methods, owing to instability problems. Although the smoothing Tikhonov regularization approach has the ability to approximatively solve the matrix equation and reconstruct most simple spectral shapes, it still suffers the limitations of reconstructing complex and irregular spectral shapes that are commonly used to distinguish different elements of detected targets with mixed substances by characteristic spectral peaks. Therefore, we propose a special pretreatment method for a matrix in convex optimization, which has been proved to be useful for reducing the condition number of matrices in the equation. In comparison with the reconstructed spectra gotten by the previous ordinary iterative method, the spectra obtained by the pretreatment method show obvious accuracy.

• Journal of information and communication convergence engineering
• /
• 제21권3호
• /
• pp.208-215
• /
• 2023

Deep learning techniques provide powerful solutions to several pattern-recognition problems, including Raman spectral classification. However, these networks require large amounts of labeled data to perform well. Labeled data, which are typically obtained in a laboratory, can potentially be alleviated by data augmentation. This study investigated various data augmentation techniques and applied multiple deep learning methods to Raman spectral classification. Raman spectra yield fingerprint-like information about chemical compositions, but are prone to noise when the particles of the material are small. Five augmentation models were investigated to build robust deep learning classifiers: weighted sums of spectral signals, imitated chemical backgrounds, extended multiplicative signal augmentation, and generated Gaussian and Poisson-distributed noise. We compared the performance of nine state-of-the-art convolutional neural networks with all the augmentation techniques. The LeNet5 models with background noise augmentation yielded the highest accuracy when tested on real-world Raman spectral classification at 88.33% accuracy. A class activation map of the model was generated to provide a qualitative observation of the results.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제21권1호
• /
• pp.1-16
• /
• 2017

The Allen-Cahn equation is solved numerically by operator splitting Fourier spectral methods. The basic idea of the operator splitting method is to decompose the original problem into sub-equations and compose the approximate solution of the original equation using the solutions of the subproblems. The purpose of this paper is to characterize higher order operator splitting schemes and propose several higher order methods. Unlike the first and the second order methods, each of the heat and the free-energy evolution operators has at least one backward evaluation in higher order methods. We investigate the effect of negative time steps on a general form of third order schemes and suggest three third order methods for better stability and accuracy. Two fourth order methods are also presented. The traveling wave solution and a spinodal decomposition problem are used to demonstrate numerical properties and the order of convergence of the proposed methods.

• ETRI Journal
• /
• 제23권1호
• /
• pp.16-22
• /
• 2001

Efficient quantization methods of the line spectrum pairs (LSP) which have good performances, low complexity and memory are proposed. The adaptive quantization range method utilizing the ordering property of LSP parameters is used in a scalar quantizer and a vector-scalar hybrid quantizer. As the maximum quantization range of each LSP parameter is varied adaptively on the quantized value of the previous order's LSP parameter, efficient quantization methods can be obtained. The proposed scalar quantization algorithm needs 31 bits/frame, which is 3 bits less per frame than in the conventional scalar quantization method with interframe prediction to maintain the transparent quality of speech. The improved vector-scalar quantizer achieves an average spectral distortion of 1 dB using 26 bits/frame. The performances of proposed quantization methods are also evaluated in the transmission errors.