• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2004년도 Proceedings of ISRS 2004
• /
• pp.518-521
• /
• 2004

Change detection is one of the common research topics in remote sensing. In general, global change detection methods using image difference method, etc, are used in low resolution images and local change detection methods using floating windows, etc, are used in high resolution images. But, these methods have disadvantages in practical use. If changed area images are automatically produced, these images will be used in public area such as regional planning, regional development managements. In this research, we developed new change detection method applicable KOMPSAT EOC images. This method automatically produces subset images in changed area.

• ETRI Journal
• /
• 제32권5호
• /
• pp.795-800
• /
• 2010

Embedding a large vocabulary speech recognition system in mobile devices requires a reduced acoustic model obtained by eliminating redundant model parameters. In conventional optimization methods based on the minimum description length (MDL) criterion, a binary Gaussian tree is built at each state of a hidden Markov model by iteratively finding and merging similar mixture components. An optimal subset of the tree nodes is then selected to generate a downsized acoustic model. To obtain a better binary Gaussian tree by improving the process of finding the most similar Gaussian components, this paper proposes a new distance measure that exploits the difference in likelihood values for cases before and after two components are combined. The mixture weight of Gaussian components is also introduced in the component merging step. Experimental results show that the proposed method outperforms MDL-based optimization using either a Kullback-Leibler (KL) divergence or weighted KL divergence measure. The proposed method could also reduce the acoustic model size by 50% with less than a 1.5% increase in error rate compared to a baseline system.

• Structural Engineering and Mechanics
• /
• 제77권1호
• /
• pp.19-35
• /
• 2021

Existing methods to estimate the probability of seismic pounding occurrence of adjacent buildings do not account for nonlinear behavior or only apply to simple lumped mass systems. The present study proposes an efficient method based on subset simulation for fragility and risk assessment of seismic pounding occurrence between nonlinear adjacent buildings neglecting pounding effects with application to finite element models. The proposed method is first applied to adjacent buildings modeled as elastoplastic systems with substantially different dynamic properties for different structural parameters. Seismic pounding fragility and risk of adjacent frame structures with different floor levels is then assessed, paying special attention to modeling the non-linear material behavior in finite element models. Difference in natural periods and impact location are identified to affect the pounding fragility simultaneously. The reliability levels of the minimum code-specified separation distances are also determined. In addition, the incremental dynamic analysis method is extended to assess seismic pounding fragility of the adjacent frame structures, resulting in higher fragility estimates for separation distances larger than the minimum code-specified ones in comparison with the proposed method.

• 한국안전학회지
• /
• 제35권3호
• /
• pp.49-57
• /
• 2020

In comparison with the existing static reliability analysis methods, the dynamic reliability analysis(DyRA) method is more suitable for estimating the failure probability of a structure subjected to earthquake excitations because it can take into account the frequency characteristics and damping capacity of the structure. However, the DyRA is known to have an issue of numerical stability due to the uncertainty in random sampling of the earthquake excitations. In order to solve this numerical stability issue in the DyRA approach, this study proposed two earthquake-scale factors. The first factor is defined as the ratio of the first earthquake excitation over the maximum value of the remaining excitations, and the second factor is defined as the condition number of the matrix consisting of the earthquake excitations. Then, we have performed parametric studies of two factors on numerical stability of the DyRA method. In illustrative example, it was clearly confirmed that the two factors can be used to verify the numerical stability of the proposed DyRA method. However, there exists a difference between the two factors. The first factor showed some overlapping region between the stable results and the unstable results so that it requires some additional reliability analysis to guarantee the stability of the DyRA method. On the contrary, the second factor clearly distinguished the stable and unstable results of the DyRA method without any overlapping region. Therefore, the second factor can be said to be better than the first factor as the criterion to determine whether or not the proposed DyRA method guarantees its numerical stability. In addition, the accuracy of the numerical analysis results of the proposed DyRA has been verified in comparison with those of the existing first-order reliability method(FORM), Monte Carlo simulation(MCS) method and subset simulation method(SSM). The comparative results confirmed that the proposed DyRA method can provide accurate and reliable estimation of the structural failure probability while maintaining the superior numerical efficiency over the existing methods.

• 한국정보통신학회논문지
• /
• 제4권3호
• /
• pp.517-526
• /
• 2000

본 논문에서는 직접대역확산통신시스템에서 코딩 이득을 향상시키고자 다중 심벌 검파를 수행하는 트렐리스 부호화 변조를 적용하였다. $MDPSK(\pi/4 shift QPSK)$를 트렐리스 부호화된 직접대역확산시스템 적용하고 정보가 인접한 채널 신호의 위상차에 전송된다는 것을 이용하여 1차 위상차 뿐만 아니라 다중위상차를 추출한다. BER특성을 향상시키기 위해, 이러한 다중 위상차를 이용하여 $MDPSK(\pi/4 shift QPSK)$에서 다중 심벌 검파를 수행하는 트렐리스 부호화된 직접대역확산시스템의 비터비 디코더 알고리듬을 설계하여 향상된 코딩 이득을 얻고자 한다. 이러한 시스템을 직접대역확산통신시스템에 적용하였을 때 얻을 수 있는 코딩 이득은 시뮬레이션 결과, AWGN채널에서 TCM의 콘볼류션부호화기의 상태수 4, 8, 16에 따라 3-4dB 정도의 성능향상이 있으며, 레일레이 페이딩 채널에서는 4-5dB정도의 성능 향상이 있음을 알 수 있다. 일반적으로 상태수가 증가할수록 더 큰 코딩 이득을 얻을 수 있다.

• 대한원격탐사학회지
• /
• 제29권6호
• /
• pp.611-619
• /
• 2013

Although the need of crop production has increased in Mongolia, crop cultivation is very limited because of the harsh climatic and topographic conditions. Crop lands are sparsely distributed with relatively small sizes and, therefore, it is difficult to survey the exact area of crop lands. The study aimed to find an easy and effective way of accurate classification to map crop lands in Mongolia using satellite images. To classify the crop lands over the study area in northern Mongolia, four classifications were carried out by using 1) Thematic Mapper (TM) image August 23, 2) TM image of July 6, 3) combined 12 bands of TM images of July and August, and 4) both TM images of July and August by layered classification. Wheat and potato are the major crop types and they show relatively high variation in crop conditions between July and August. On the other hands, other land cover types (forest, riparian vegetation, grassland, water and bare soil) do not show such difference between July and August. The results of four classifications clearly show that the use of multi-temporal images is essential to accurately classify the crop lands. The layered classification method, in which each class is separated by a subset of TM images, shows the highest classification accuracy (93.7%) of the crop lands. The classification accuracies are lower when we use only a single TM image of either July or August. Because of the different planting practice of potato and the growth condition of wheat, the spectral characteristics of potato and wheat cannot be fully separated from other cover types with TM image of either July or August. Further refinements on the spatial characteristics of existing crop lands may enhance the crop mapping method in Mongolia.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 학술대회 논문집 정보 및 제어부문
• /
• pp.346-348
• /
• 2004

In this rarer, we introduce a new Fuzzy Polynomial Neural Networks (FPNNs)-like structure whose neuron is based on the Fuzzy Set-based Fuzzy Inference System (FS-FIS) and is different from that of FPNNs based on the Fuzzy relation-based Fuzzy Inference System (FR-FIS) and discuss the ability of the new FPNNs-like structurenamed Fuzzy Set-based Polynomial Neural Networks (FSPNN). The premise parts of their fuzzy rules are not identical, while the consequent parts of the both Networks (such as FPNN and FSPNN) are identical. This difference results from the angle of a viewpoint of partition of input space of system. In other word, from a point of view of FS-FIS, the input variables are mutually independent under input space of system, while from a viewpoint of FR-FIS they are related each other. In considering the structures of FPNN-like networks such as FPNN and FSPNN, they are almost similar. Therefore they have the same shortcomings as well as the same virtues on structural side. The proposed design procedure for networks' architecture involves the selection of appropriate nodes with specific local characteristics such as the number of input variables, the order of the polynomial that is constant, linear, quadratic, or modified quadratic functions being viewed as the consequent part of fuzzy rules, and a collection of the specific subset of input variables. On the parameter optimization phase, we adopt Information Granulation (IG) based on HCM clustering algorithm and a standard least square method-based learning. Through the consecutive process of such structural and parametric optimization, an optimized and flexible fuzzy neural network is generated in a dynamic fashion. To evaluate the performance of the genetically optimized FSPNN (gFSPNN), the model is experimented with using gas furnace process dataset.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2004년도 추계학술대회 학술발표 논문집 제14권 제2호
• /
• pp.433-436
• /
• 2004

In this paper, we introduce a new Fuzzy Polynomial Neural Networks (FPNNS)-like structure whose neuron is based on the Fuzzy Set-based Fuzzy Inference System (FS-FIS) and is different from that of FPNNS based on the Fuzzy relation-based Fuzzy Inference System (FR-FIS) and discuss the ability of the new FPNNS-like structure named Fuzzy Set-based Polynomial Neural Networks (FSPNN). The premise parts of their fuzzy rules are not identical, while the consequent parts of the both Networks (such as FPNN and FSPNN) are identical. This difference results from the angle of a viewpoint of partition of input space of system. In other word, from a point of view of FS-FIS, the input variables are mutually independent under input space of system, while from a viewpoint of FR-FIS they are related each other. The proposed design procedure for networks architecture involves the selection of appropriate nodes with specific local characteristics such as the number of input variables, the order of the polynomial that is constant, linear, quadratic, or modified quadratic functions being viewed as the consequent part of fuzzy rules, and a collection of the specific subset of input variables. On the parameter optimization phase, we adopt Information Granulation (IC) based on HCM clustering algorithm and a standard least square method-based learning. Through the consecutive process of such structural and parametric optimization, an optimized and flexible fuzzy neural network is generated in a dynamic fashion. To evaluate the performance of the genetically optimized FSPNN (gFSPNN), the model is experimented with using the time series dataset of gas furnace process.

• 산업경영시스템학회지
• /
• 제32권2호
• /
• pp.29-37
• /
• 2009

National R&D programs play an important role in the development of a country in this age of the knowledge economy. Since many numbers of R&D programs compete for limited resources such as national R&D budget, the R&D program evaluation problem is a challenging decision-making problem faced by decision makers that deal with R&D management. In this sense, DEA(Data Envelopment Analysis) has been regarded as one of the most widely accepted methods to measure the relative efficiency of productivity of R&D programs. DEA is a methodology to measure and to evaluate the relative efficiency of a homogeneous set of decision-making units(DMUs) in a process which uses multiple inputs to produce multiple outputs. However, the sample of the R&D programs could consist of two or more naturally occurring subsets, thus exhibiting clear signs of heterogeneity such as different objectives. In such situations, the fairness of DEA is limited, for the nature of the relative efficiency of a DMU is likely to be influenced by its membership in a particular subset of the sample. In this study, we propose a methodology AI-DEA(attainment index DEA) allowing for reflecting decision maker's subjective judgement on difference among different subsets of R&D programs which have heterogeneous objectives. This methodology combines AHP and Delphi in order to decide the attainmnet index of each DMU for each outputs, and apply them to DEA model. We illustrate the proposed approach with a pilot evaluation of 13 programs involving 6 different subsets of Korean National R&D programs and compares the results of the original DEA model and AI-DEA model.

• 대한두개안면성형외과학회지
• /
• 제22권6호
• /
• pp.303-309
• /
• 2021

Background: Transferring the hypoglossal nerve to the facial nerve using an end-to-end method is very effective for improving facial motor function. However, this technique may result in hemitongue atrophy. The ansa cervicalis, which arises from the cervical plexus, is also used for facial reanimation. We retrospectively reviewed cases where facial reanimation was performed using the ansa cervicalis to overcome the shortcomings of existing techniques of hypoglossal nerve transfer. Methods: The records of 15 patients who underwent hypoglossal nerve transfer were retrospectively reviewed. Three methods were used: facial reanimation with hypoglossal nerve transfer (group 1), facial nerve reanimation using the ansa cervicalis (group 2), and sural nerve interposition grafting between the hypoglossal nerve and facial nerve (group 3). In group 1, the ansa cervicalis was coapted to neurotize the distal stump of the hypoglossal nerve in a subset of patients. Clinical outcomes were evaluated using the House-Brackmann (H-B) grading system and Emotrics software. Results: All patients in group 1 (n= 4) achieved H-B grade IV facial function and showed improvements in the oral commissure angle at rest (preoperative vs. postoperative difference, 6.48° ± 0.77°) and while smiling (13.88° ± 2.00°). In groups 2 and 3, the oral commissure angle slightly improved at rest (group 2: 0.95° ± 0.53°, group 3: 1.35° ± 1.02°) and while smiling (group 2: 2.06° ± 0.67°, group 3: 1.23° ± 0.56°). In group 1, reduced tongue morbidity was found in patients who underwent ansa cervicalis transfer. Conclusion: Facial reanimation with hypoglossal nerve transfer, in combination with hypoglossal nerve neurotization using the ansa cervicalis for complete facial palsy patients, might enable favorable facial reanimation outcomes and reduce tongue morbidity. Facial reanimation using the ansa cervicalis or sural nerve for incomplete facial palsy patients did not lead to remarkable improvements, but it warrants further investigation.