• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권7호
• /
• pp.2957-2980
• /
• 2020

Ensemble clustering commonly integrates multiple basic partitions to obtain a more accurate clustering result than a single partition. Specifically, it exists an inevitable problem that the incomplete transformation from the original space to the integrated space. In this paper, a novel ensemble clustering algorithm using a newly reconstructed mapping coefficient (ECRMC) is proposed. In the algorithm, a newly reconstructed mapping coefficient between objects and micro-clusters is designed based on the principle of increasing information entropy to enhance effective information. This can reduce the information loss in the transformation from micro-clusters to the original space. Then the correlation of the micro-clusters is creatively calculated by the Spearman coefficient. Therefore, the revised co-association graph between objects can be built more accurately because the supplementary information can well ensure the completeness of the whole conversion process. Experiment results demonstrate that the ECRMC clustering algorithm has high performance, effectiveness, and feasibility.

• ETRI Journal
• /
• 제35권1호
• /
• pp.100-108
• /
• 2013

In this paper, a feature extraction (FE) method is proposed that is comparable to the traditional FE methods used in automatic speech recognition systems. Unlike the conventional spectral-based FE methods, the proposed method evaluates the similarities between an embedded speech signal and a set of predefined speech attractor models in the reconstructed phase space (RPS) domain. In the first step, a set of Gaussian mixture models is trained to represent the speech attractors in the RPS. Next, for a new input speech frame, a posterior-probability-based feature vector is evaluated, which represents the similarity between the embedded frame and the learned speech attractors. We conduct experiments for a speech recognition task utilizing a toolkit based on hidden Markov models, over FARSDAT, a well-known Persian speech corpus. Through the proposed FE method, we gain 3.11% absolute phoneme error rate improvement in comparison to the baseline system, which exploits the mel-frequency cepstral coefficient FE method.

• 한국컴퓨터그래픽스학회논문지
• /
• 제24권3호
• /
• pp.61-71
• /
• 2018

RGB-D 카메라 촬영 영상에 대한 카메라 포즈 추정을 통하여 복원한 3차원 전역 공간의 점 집합으로부터 삼각형 메쉬를 생성할 때, 일반적으로 메쉬의 크기가 커질수록 3차원 모델의 품질 또한 향상된다. 하지만 어떤 한계를 넘어서 삼각형 메쉬의 해상도를 높일 경우, 메모리 요구량의 과도한 증가나 실시간 렌더링 성능저하 문제뿐만 아니라 RGB-D 센서의 정밀도 한계로 인한 접 집합 데이터의 노이즈에 민감해지는 문제가 발생한다. 본 논문에서는 실시간 응용에 적합한 3차원 모델 생성을 위하여 비교적 적은 크기의 삼각형 메쉬에 대하여 3차원 점 집합의 촬영 색상으로부터 고화질의 텍스쳐를 생성하는 기법을 제안한다. 특히 카메라 포즈 추정을 통하여 생성한 3차원 점 집합 공간과 2차원 텍스쳐 공간 간의 매핑 관계를 활용한 간단한 방법을 통하여 RGB-D 카메라 촬영 영상으로부터 복원한 3차원 모델에 대하여 효과적으로 텍스쳐를 생성할 수 있음을 보인다.

• 전자공학회논문지B
• /
• 제33B권11호
• /
• pp.44-52
• /
• 1996

This paper proposes a KLT-CVQ scheme using PCNN to improbe the quality of the reconstructed images at a given bit rate. By using the PCNN and classified vector quantization, we exploit the high energy compaction and compelte decorrelation capbilities of the KLT, and the pdf (probability density function) shape and space-filling advantages of the vQ to improve the performance of the proposed hybrid coding technique. In order to preserve the preceptual fetures such as the edge components in the reconstructed images, we classified the input image blocks according to the texture energy measures of the local statistics and vector-coded them adaptively, and thereby reduces the possible edge degradation in the reconstructed images. The results of the computer simulations show that the performance of the proposed KLT-CVQ is higher than that of the KLT-CSQ or the DCT-CVQ in the quality of the reconstructed images at a given bit rate.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권10호
• /
• pp.5056-5078
• /
• 2018

In this work we propose a compression technique that makes use of linear and perceptually uniform CIE $La^*b^*$ color space in the JPEG image compression framework to improve its performance at lower bitrates. To generate quantization matrices suitable for the linear and perceptually uniform CIE $La^*b^*$ color space, a novel linear Contrast Sensitivity Function (CSF) is used. The compression performance in terms of Compression Ratio (CR) and Peak Signal to Noise Ratio (PSNR), is further improved by utilizing image dependent, variable and non-uniform image sub-blocks generated using a proposed histogram-based merging technique. Experimental results indicate that the proposed linear CSF based quantization technique yields, on an average, 8% increase in CR for the same reconstructed image quality in terms of PSNR as compared to the conventional YCbCr color space. The proposed scheme also outperforms JPEG in terms of CR by an average of 45.01% for the same reconstructed image quality.

• 한국실내디자인학회논문집
• /
• 제26권3호
• /
• pp.91-100
• /
• 2017

The rapid growth of Korean-style buffet in recent years has increased the size of corporate investment. Under this circumstance, businesses make various marketing efforts while highlighting the features and advantages of their brands. Against this backdrop, this study aims to understand which space branding has been applied to the Korean-style buffet through case studies and to propose a method of application of space branding to increase sales in an effective manner in the future. First, the research is theoretical examination and case studies with focus on the characteristics of expressions of Korean-style buffet space branding. Basically, upon completion of examining the concept and the characteristics of expressions of the Korean-style buffet and analyzing the concept of space branding, the components of space branding have been reconstructed based on preceding studies on space branding and then have been applied to each brand space. Also, the hands-on experience of the characteristics of expression of the Korean-style buffet with space branding incorporated in it and prepared a checklist via visual inspections. And then, the field surveys based on these examinations and took approach of drawing a conclusion based on the results analysis conducted by using the SPSS statistical program. Through preceding studies, the three components of space branding, that is, sensory element, emotional element, and cognitive element have been reconstructed before proceeding with this study, which has obtained five major findings as a conclusion. First, the sensory element should be given elements differentiated enough to attract consumers' attention along with a sustainable effort to have brand image imprinted in their mind. Second, in terms of emotional element, the study has found that the brand experience oriented toward interest and participation results in higher utility frequency. Third, the study has found that the cognitive element should seek consistency in communicating with consumers with focus on face-to-face contact on the display in space. Fourth, it has been found that arranging independent spaces is necessary to attract consumers' participation. Finally, the study has identified in which location area of buffet the sensory, emotional, and cognitive elements have placed a significant weight.

• 대한전기학회논문지:전력기술부문A
• /
• 제48권5호
• /
• pp.658-663
• /
• 1999

2D sliced CT images hardly express the human disease in a space. This space expression can be reconstructed into 3D image by piling up the CT sliced image in succession. In medical image, in order to get the reconstructed 3D images, expensive system or much calculation time is needed. But by changing the method of reconstruction procedure and limit the range, the reconstruction time could be reduced. In this study, to reduce the processing time and memory, we suggested a method of interpolation and ray casting processing at the same time in a limited range. Such a limited range processing have advantages that we could reduce the unnecessary interpolation and ray casting. Through a experiment, it is founded that the reconstruction time and the memory was much reduced.

• 산업기술연구
• /
• 제24권B호
• /
• pp.29-36
• /
• 2004

The wind and wind-induced dynamic wind pressures fluctuate irregularly according to time and space. In this study, the proper orthogonal decomposition(POD) technique is applied to wind pressures acting on a tall tower model, and the following results are found: the along-wind and across-wind forces can be reconstructed by only four dominant POD modes, and the reconstructed errors are 4.71% and 22%, respectively for across-wind and along-wind directions. The physical meanings for dominant modes are also presented in the paper. The POD analysis can compress complex wind pressure data only by a few dominant modes and interpret spatio-temporal characteristics of wind pressure by novel way while existing statistical methods do not have such benefits.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2002년도 Proceedings of International Symposium on Remote Sensing
• /
• pp.400-406
• /
• 2002

Recently, the enormous increase in the volume of remotely sensed data is being acquired by an ever-growing number of earth observation satellites. The combining of diversely sourced imagery together is an important requirement in many applications such as data fusion, city modeling and object recognition. Aerial triangulation is a procedure to reconstruct object space from imagery. However, since the different kinds of imagery have their own sensor model, characteristics, and resolution, the previous approach in aerial triangulation (or georeferencing) is performed on a sensor model separately. This study evaluated the advantages of aerial triangulation of large number of images from multi-sensors simultaneously. The incorporated multi-sensors are frame, push broom, and whisky broom cameras. The limits and problems of push-broom or whisky broom sensor models can be compensated by combined triangulation with frame imagery and vise versa. The reconstructed object space from multi-sensor triangulation is more accurate than that from a single model. Experiments conducted in this study show the more accurately reconstructed object space from multi-sensor triangulation.

• Journal of Power Electronics
• /
• 제12권6호
• /
• pp.895-903
• /
• 2012

This paper presents a fast switching direct torque control (FS-DTC) using only a single DC-link current sensor. In FS-DTC, six new active voltage complex space vectors (CSVs) are synthesized by the conventional active voltage space vectors (SVs). The corresponding sectors are rotated in the anticlockwise direction by 30 degrees. A selection table is defined to select the CSVs. Based on the "Different Phase Mode", the output sequence of the selected CSV is optimized. Accordingly, a reconstruction method is proposed to acquire the phase currents. The core of the FS-DTC is that all of the three phase currents can be reliably reconstructed during every two sampling periods, which is the result of the fast switching between different phases. The errors between the reconstructed and actual currents are strictly limited in one sampling period. The FS-DTC has the advantages of the standard DTC scheme such as simple structure, quick torque response and robustness. As can be seen in the analysis, the FS-DTC can be thought of as an equivalent standard DTC scheme with 86.6% of the maximum speed, 173.2% of the torque ripple, and 115% of the response time of the torque. Based on a dSPACE DS1103 controller, the FS-DTC is implemented in an induction machine drive system. The results verify the effectiveness of the FS-DTC.