• 한국콘텐츠학회논문지
• /
• 제12권6호
• /
• pp.9-20
• /
• 2012

컴퓨터비전에서 안정적으로 대응점을 획득하는 것은 매우 중요한 일이다. 그러나 이들은 스케일, 조명, 시점 등이 변하는 환경에서 정확한 대응점을 찾는 과정은 쉽지 않다. SIFT 알고리즘은 객체의 모서리나 꼭지점으로부터 추출한 특징벡터를 사용하므로 스케일링, 회전, 조명변화를 가지는 영상에서도 뛰어난 매칭을 수행한다. 그러나 SIFT는 엣지에 의해 특징점을 추출하므로 엣지가 존재하지 않는 영역에서는 원하는 대응점을 찾을 수 없다. 본 연구는 SIFT에 의한 대응 특징점 추출과 매칭 성능을 향상시키기 위한 마커 모양 및 배치 방법을 제안한다. 제안 방법에서 사용한 마커의 모양은 부착 방향에 따라 SIFT 알고리즘에 의해 한 방향으로 우세한 벡터를 검출할 수 있는 반원형(SemiCircle)으로 구성한다. 그리고 대응점 매칭의 성능을 향상시키기 위하여 마커의 방향 배치는 드 브루인 수열(De Bruijn Sequence)을 이용한다. 실험을 통해 제안한 방법이 기존의 방법보다 더 정확한 특징점 검출과 매칭에 효과적임을 증명하였다.

• 한국군사과학기술학회지
• /
• 제12권4호
• /
• pp.524-531
• /
• 2009

In this paper, we propose an automatic image registration method for multi-sensor image fusion such as visible and infrared images. The registration is achieved by finding corresponding feature points in both input images. In general, the global statistical correlation is not guaranteed between multi-sensor images, which bring out difficulties on the image registration for multi-sensor images. To cope with this problem, mutual information is adopted to measure correspondence of features and to select faithful points. An update algorithm for projective transform is also proposed. Experimental results show that the proposed method provides robust and accurate registration results.

• Korean Journal of Acupuncture
• /
• 제17권1호
• /
• pp.157-171
• /
• 2000

Front Points are told to be deeply related with viscera and bowels in oriental medicine. Since the Front Points are treated as response zone, it can be used for the diagnosis and treatment of disease in viscera and bowels. The location of Front Points are very similar to the several response zones in western medicine. Diagnostic aspect of Front Points to several response zones were discussed in here based on many thesis and reported laboratorial experiments. Front Points are located in the same latitude of viscera and bowels. So, diagnosis result in Front Points represent condition of each organs. Palpitation of the Front Points can make diagnosis of body surface and the organ beneath the body surface. Such use of Front Points for diagnosis can used as the treatment points also. Only three Front Points (LU-1, LI-14, GB-24) in lung, liver and gall bladder meridian are located on its meridian. The Front Points of Stomach (CV-12) cross its meridian or closely located to it. Unlike to those four Front Points that are matched to its meridian, other eight Front Points do not located or cross its meridian at all. It seems that the location of Front Points are decided by the location of organs and the conditions at a certain organ do not delivered by the meridians but delivered by main collaterals, tertiary collaterals and superficial collaterals instead. Among visceral response zones, Five Front Points (CV-3, CV-4, CV-12, LI-14, GB-25) are exactly matched to Head's response zone and other Front Points are closely matched to the Head's response zone. There are five Front Points (CV-12, CV-14, CV-17, LI-14, GB-24) that are matched with the location of pressing palpitation point and other Front Points are closely located to the pressing palpitation point. So far, it was clear that the Front Points do have important role as response points. Symptoms expressed to the Front Points were delivered conditions or symptoms occur in corresponding organ and the anatomical location of Front Points were also found near the corresponding organ. Diagnostic and therapeutic application of Front Points for Organ theory and in the Interpromotion-restraint of the five elements in oriental medicine can be made in future to increase its potential.

• 정보처리학회논문지:소프트웨어 및 데이터공학
• /
• 제7권1호
• /
• pp.9-18
• /
• 2018

최근 가상현실 기반의 콘텐츠들이 늘어나면서 이미지 Stitching 기술의 사용이 증가하고 있다. 이미지 Stitching이란 고해상도 이미지 및 넓은 시야(Wide Field of View)의 이미지를 생성하기 위해 다중의 영상을 정합하는 방법이다. 이런 이미지 Stitching은 하나의 카메라로부터 생성되는 영상의 한계를 넘어 다양한 분야에서 활용되고 있다. 이미지 Stitching은 다중의 영상을 정합하기 위해 특징 점 및 대응점을 검출하고 RANSAC 알고리즘을 이용하여 영상간의 변환관계(Homography)를 계산한다. 일반적으로 변환관계 계산을 위해 대응점들이 필요하다. 그러나 대응점들에는 변환관계에 대한 잘못된 가정이나 오류로 인해 발생할 수 있는 다양한 유형의 노이즈(Noise)가 포함되어 있다. 이러한 노이즈는 변환관계를 정확히 예측하는 방해 요인이 된다. 이처럼 일반적으로 사용되는 대응점 매칭(Matching) 방법들은 잘못된 대응점들을 매칭할 수 있는 경우가 발생하기 때문에 모델 파라미터의 예측을 방해하는 대응점(Outlier)로부터 정확한 변환관계를 구축하기 위해 RANSAC 알고리즘을 사용한다. 본 논문에서는 RANSAC 알고리즘에 사용되는 대응점 관계 정보를 이용하여 좀 더 정확한 대응점(Inlier)을 추출하고 정확한 변환관계를 계산하는 알고리즘을 제안한다. 대응점 관계 정보는 이미지 매칭에 사용되는 대응점 간의 거리 비율을 사용하며, 본 논문은 기존 RANSAC 알고리즘과 같은 성능을 유지하면서 처리 시간을 단축시키는데 있다.

• Structural Engineering and Mechanics
• /
• 제68권5호
• /
• pp.633-638
• /
• 2018

One of the important problems in the vehicle design is vehicle handling and stability. Effective parameters which should be considered in the vehicle handling and stability are roll angle, camber angle and scrub radius. In this paper, a planar vehicle model is considered that two right and left suspensions are double wishbone suspension system. For a better analysis of the suspension geometry, a kinestatic model of vehicle is considered which instantaneous kinematic and statics relations are analyzed simultaneously. In this model, suspension geometry is considered completely. In order to optimum design of double wishbones suspension system, a multi-objective genetic algorithm is applied. Three important parameters of suspension including roll angle, camber angle and scrub radius are taken into account as objective functions. Coordinates of suspension hard points are design variables of optimization which optimum values of them, corresponding to each optimum point, are obtained in the optimization process. Pareto solutions for three objective functions are derived. There are important optimum points in these Pareto solutions which each point represents an optimum status in the model. In other words, corresponding to any optimal point, a specific geometric position is determined for the suspension hard points. Each of the obtained points in the Pareto optimization can be selected for a special design purpose by designer to create an optimum condition in the vehicle handling and stability.

• 대한치과의사협회지
• /
• 제32권10호통권305호
• /
• pp.741-747
• /
• 1994

• 한국측량학회:학술대회논문집
• /
• 한국측량학회 2004년도 춘계학술발표회논문집
• /
• pp.519-520
• /
• 2004

Using the baseline vectors obtained by GPS observations on 32 triangulation points over Gyunggido area, Bessel coordinates of the corresponding points are calculated according to trilateration adjustment. The result shows the possibility of construction of a new cadastral survey network with the degree of correctness of 5cm, when newly computing the coordinates of the survey datum point of the area with values from GPS baseline vectors.

• 충청수학회지
• /
• 제28권1호
• /
• pp.13-28
• /
• 2015

In this paper, we use two mappings satisfying certain expansive conditions to construct convergent sequences in complex valued metric spaces, and then we prove that the limits of the convergent sequences are the points of coincidence or common fixed points for the two mappings. The main theorems in this paper are the generalizations and improvements of the corresponding results in real metric spaces, cone metric spaces and topological vector space-valued cone metric spaces.

• 한국통신학회논문지
• /
• 제22권5호
• /
• pp.928-938
• /
• 1997

We propose an algorithm for extracting singular points of fingerprint image using directional information. First, we extract the candidates of singular points using Poincare index in two(lower and higher) resolutional directional images. Then we remove the false singular points using smoothing technique from lower resolutional directional image. And finally we select the singular points in higher resolution corresponding to those in lower resolution. The possible missing points in lower resolution are found by computing Poincare index algong the proposed small curve. And the reliable points are selected from analysis around them. We also propose a method for segmentation of fingerprint as preprocessing step to enhance the computational speed and the performance of system.

• 로봇학회논문지
• /
• 제9권4호
• /
• pp.242-249
• /
• 2014

By a SLAM (simultaneous localization and mapping) method, we get a map of an environment for autonomous navigation of a robot. In this case, we want to know how accurate the map is. Or we want to know which map is more accurate when different maps can be obtained by different SLAM methods. So, several methods for map comparison have been studied, but they have their own drawbacks. In this paper, we propose a new method which compares the accuracy or error of maps relatively and quantitatively. This method sets many corresponding points on both reference map and SLAM map, and computes the translational and rotational values of all corresponding points using least-squares solution. Analyzing the standard deviations of all translational and rotational values, we can know the error of two maps. This method can consider both local and global errors while other methods can deal with one of them, and this is verified by a series of simulations and real world experiments.