• 한국지능시스템학회논문지
• /
• 제20권5호
• /
• pp.701-709
• /
• 2010

본 논문에서는 영상간의 거리에 반비례하고 상관성에 비례하는 조합형 유사성 척도에 의한 효율적인 서명인식 방법을 제안하였다. 여기서 거리는 영상의 공간적 속성을 반영하기 위함이고, 상관성은 통계적 속성을 반영하기 위함이다. 이렇게 하면 서명의 위치, 크기, 회전과 같은 기하학적 변화와 모양변화에 강건한 인식이 가능하다. 상관성의 척도로 이진영상의 히스토그램에 기반을 둔 4 방향의 위치를 고려한 정규상호상관계수를 이용함으로써 서명사이의 유사성을 좀 더 빠르고 정확하게 반영하였다. 제안된 방법을 20개의 288$\times$288 픽셀 트럭영상과 105개의 256$\times$256 픽셀의 서명영상을 대상으로 각각 실험한 결과, 영상의 속성을 잘 반영한 우수한 인식성능이 있음을 확인하였다. 특히 정규상호상관계수와 순서값의 거리를 조합한 척도가 city-block이나 Euclidean 거리를 각각 조합한 척도보다 우수한 인식성능이 있음도 알 수 있었다.

• 한국멀티미디어학회논문지
• /
• 제14권6호
• /
• pp.742-751
• /
• 2011

본 논문에서는 형상 특징자인 열 커널 인증 (Heat Kernel Signature, HKS)를 기반으로 강인한 3D 모델 해싱을 제안한다. 키와 매개변수에 의존한 형상 특징자 기반 3D 모델 해싱을 제안한다. 제안한 방법에서는 Mesh Laplace 연산자의 고유치와 고유벡터에 의하여 각 꼭지점에 대한 전역 및 국부 타임 HKS 계수를 구한 다음, 이 계수들을 정방형 2D 셀로 군집화한다. 그리고 각 셀에 할당된 HKS 계수 쌍의 거리 가중치 기반으로 정의된 특징계수와 랜덤 계수 키와의 조합에 의하여 중간 해쉬 계수를 생성한 다음, 이진화 과정에 의하여 최종 이진 해쉬를 생성한다. 본 실험에서는 3D 범용 툴을 이용한 다양한 기하하적 공격과 위상학적 공격을 통하여 강인성을 평가하였고, 모델과 키 조합에 대한 해쉬의 유일성을 평가하였다. 또한 인증 범위를 만족히는 공격 세기를 측정함으로써 모델 공간성을 평가하였다. 실험결과로부터 제안한 3D 모델 해싱이 기존 해싱에 비하여 강인성 모델 공간성 및 유일성이 우수함을 확인하였다.

• 한국군사과학기술학회지
• /
• 제17권4호
• /
• pp.463-470
• /
• 2014

Stealth technology of combat aircraft is most significant capability in recent air battlefield. As the detector of IR missiles is being developed, IR stealth capability which is evaluated by IR signature level become more important than it was in previous generation. Among IR signature of aircraft from various sources, aerodynamic heating dominates in long-wavelength IR spectrum of $8{\sim}12{\mu}m$. Skin temperature change by aerodynamic heating which is derived by effects of Mach number and structure. The 4th and 5th generation aircraft are selected for calculation of the skin temperature, and its height and velocity in numerical conditions are 10,000 m and Ma 0.9~1.9 respectively. Aircraft skin temperature is calculated by computing convection of fluid and conduction, convection and radiation of surface. As the aircraft accelerates to higher Mach number, maximum skin temperature increases more rapidly than average temperature and temperature distribution changes in more sharp, interactive ways. The 4th generation aircraft whose shape is more complex than that of the 5th generation aircraft have complicated temperature distribution. On the other hand, the 5th generation aircraft whose shape is relatively simple shows plain temperature distribution and lower skin temperature in terms of both average and maximum value.

• 전기학회논문지
• /
• 제64권6호
• /
• pp.900-911
• /
• 2015

In this paper, 3D face recognition model is designed by using Polynomial based RBFNN(Radial Basis Function Neural Network) and PNN(Polynomial Neural Network). Also recognition rate is performed by this model. In existing 2D face recognition model, the degradation of recognition rate may occur in external environments such as face features using a brightness of the video. So 3D face recognition is performed by using 3D scanner for improving disadvantage of 2D face recognition. In the preprocessing part, obtained 3D face images for the variation of each pose are changed as front image by using pose compensation. The depth data of face image shape is extracted by using Multiple point signature. And whole area of face depth information is obtained by using the tip of a nose as a reference point. Parameter optimization is carried out with the aid of both ABC(Artificial Bee Colony) and PSO(Particle Swarm Optimization) for effective training and recognition. Experimental data for face recognition is built up by the face images of students and researchers in IC&CI Lab of Suwon University. By using the images of 3D face extracted in IC&CI Lab. the performance of 3D face recognition is evaluated and compared according to two types of models as well as point signature method based on two kinds of depth data information.

• 한국지능시스템학회논문지
• /
• 제22권6호
• /
• pp.748-753
• /
• 2012

본 논문에서는 최적 pRBFNNs 패턴분류기 기반 3차원 스캐너를 이용한 얼굴인식 알고리즘을 설계한다. 일반적으로 2차원 영상을 이용한 얼굴인식 시스템은 사진의 명암도를 이용하여 얼굴의 특징을 추출하게 된다. 그렇기 때문에 빛이나 조명, 또는 얼굴 포즈와 같은 환경 변화들은 시스템의 성능을 저하시킨다. 따라서 본 논문에서 제안된 얼굴인식 알고리즘은 2차원 얼굴인식 시스템의 한계를 극복하기 위하여 3차원 스캐너를 사용하여 설계한다. 먼저 3차원 스캐너를 이용하여 얼굴 형상을 스캔하고 스캔된 얼굴 형상은 포즈 보상 과정을 통하여 정면으로 변환된다. 그 후에 Point Signature 기법을 사용하여 얼굴의 깊이 정보를 추출하고 마지막으로 고차원 패턴인식 문제에 대한 해결을 위하여 최적화된 pRBFNNs (Polynomial-based Radial Basis Function Neural Networks) 모델을 사용하여 인식성능을 확인한다.

• Journal of Information Processing Systems
• /
• 제14권1호
• /
• pp.176-190
• /
• 2018

This paper presents a novel convolutional neural network based multi-feature fusion learning method for non-rigid 3D model retrieval, which can investigate the useful discriminative information of the heat kernel signature (HKS) descriptor and the wave kernel signature (WKS) descriptor. At first, we compute the 2D shape distributions of the two kinds of descriptors to represent the 3D model and use them as the input to the networks. Then we construct two convolutional neural networks for the HKS distribution and the WKS distribution separately, and use the multi-feature fusion layer to connect them. The fusion layer not only can exploit more discriminative characteristics of the two descriptors, but also can complement the correlated information between the two kinds of descriptors. Furthermore, to further improve the performance of the description ability, the cross-connected layer is built to combine the low-level features with high-level features. Extensive experiments have validated the effectiveness of the designed multi-feature fusion learning method.

• Structural Engineering and Mechanics
• /
• 제70권4호
• /
• pp.457-466
• /
• 2019

This study develops a damage detection method based on neural networks. The performance of the method is numerically and experimentally verified using a three-story shear building model. The framework is mainly composed of two hierarchical stages to identify damage location and extent using artificial neural network (ANN). The normalized damage signature index, that is a normalized ratio of the changes in the natural frequency and mode shape caused by the damage, is used to identify the damage location. The modal parameters extracted from the numerically developed structure for multiple damage scenarios are used to train the ANN. The positive alarm from the first stage of damage detection activates the second stage of ANN to assess the damage extent. The difference in mode shape vectors between the intact and damaged structures is used to determine the extent of the related damage. The entire procedure is verified using laboratory experiments. The damage is artificially modeled by replacing the column element with a narrow section, and a stochastic subspace identification method is used to identify the modal parameters. The results verify that the proposed method can accurately detect the damage location and extent.

• 한국전산유체공학회지
• /
• 제3권1호
• /
• pp.100-107
• /
• 1998

A numerical analysis of thermal stress over temperature variations near the crystal-melt interface is carried out for a floating-zone growth of Cadmium Telluride (CdTe). Thermocapillary convection determines crystal-melt interfacial shape and signature of temperature in the crystal. Large temperature gradients near the crystal-melt interface yield excessive thermal stresses in a crystal, which affect the dislocations of the crystal. Based on the assumption that the crystal is elastic and isotropic, thermal stresses in a crystal are computed and the effects of operating conditions are investigated. The results show that the extreme thermal stresses are concentrated near the interface of a crystal and the radial and the tangential stresses are the dominant ones. Concentrated heating profile increases the stresses within the crystal, otherwise, the pulling rate decreases the stresses.

• 대한기계학회논문집B
• /
• 제23권5호
• /
• pp.621-627
• /
• 1999

This paper presents Raman scattering of liquid water to obtain the characteristics with variation of temperature. Very clear Stokes-Raman signals were observed, which shows H-O vibration stretching and H-O-H vibration bending. The obtained spectrum were processed by FFT filter to extract the noise and base. The spectral shape of the H-O stretching provided a various sensitive signature which allowed temperature to be determined by a curve-fitting technique. Those are Maximum Intensity, Maximum Wave Length, FWHM(Full Width at Half Maximum), PMCR(Polymer Monomer Concentration Ratio) and TSIR(Temperature Sensitive Intensity Ratio). TSIR method shows the highest accuracy of $0.1^{\circ}C$ in mean error and $0.32^{\circ}C$ In maximum error.

• 대한조선학회논문집
• /
• 제43권6호
• /
• pp.693-699
• /
• 2006

This paper proposes a design optimization technique for ship RCS signature reductions using Taguchi method. The proposed technique comprises of i)evaluating initial RCS signatures, ii)defining critical areas which should be modified as design parameters, and threat factors which can't be controlled artificially as noise parameters, and finally iv)finding optimum parameters via analyzing signal to noise ratios for designated characteristics. We applied the technique to a model ship and found that it is suitable for radar stealth designs. In addition, the proposed technique is applicable to submarine designs against sonar threats.