• Proceedings of the KIEE Conference
• /
• 2006.10c
• /
• pp.297-299
• /
• 2006

Searching the region of interest on the Hough transform domain is done to determine the real road boundary on the high speed way. The mathematical morphology is employed to obtain the gradient image which is utilized in Hough transform. Many possible candidates of lines could appear on the ordinary road environment and simple selection of the strongest line segments likely to be fault boundary lines. To solve such problem, the search area for the candidates of the road boundary which is called the region of interest is limited on the Hough space. The effectiveness of the proposed algorithm has been shown with experimental results.

• Proceedings of the Korea Society for Simulation Conference
• /
• 2001.10a
• /
• pp.223-230
• /
• 2001

This paper proposes a new contour detection method and adaptive reconstruction scheme for the cylindrical organs, such as blood vessels or arteries. Furthermore, we present java-based navigation controller which has been built to examine the inside of cylindrical objects. Tn the preprocessing procedure, a few preprocessing image filters are applied in order to remove unwanted artifacts from the medical images and to estimate threshold values for the object of interest. We define a context-free grammar, which is proper fur properties of contours of cylindrical objects. In the next procedure, we extract contours using advanced radial gradient method and represent contours as context-free grammar derivation trees. We build polygons between two contours efficiently by traversing the derivations trees of the contours. We fly through the reconstructed virtual models using java-based navigation controller and VRML viewer.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.6 no.8
• /
• pp.1319-1324
• /
• 2002

In this paper, we present a new technique for extraction of moving objects in moving image sequence. The detection method of change regions is based on spatial gradient difference and temporal difference. Moving objects are extracted by motion information and prediction error of each region. In the simulation results, the proposed technique shows good performance.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.1C
• /
• pp.45-49
• /
• 2004

An image enhancement method using modified anisotropic diffusion filter is proposed in this paper. It employs sensor noise estimation and scale space methods based on the minimum reliable scale. Then the anisotropic diffusion filter is modified by the calculated critical value function and local gradient. Through simulation, it is verified that the proposed algorithm has the capability of little or no noise amplification in homogenous region as well as superior edge enhancement.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.26 no.12
• /
• pp.2503-2510
• /
• 2002

Since the first shape-form-shading technique was developed by Horn in the early 1970s, many different approaches have been continuously emerging in the past three decades. Some of them improve existing techniques, while others are completely new approaches. Using the image reflectance equation, they estimate the 3-D shape of an object utilizing adequate constraints. Each algorithm applies different constraints such as brightness, smoothness, and integrability to solve the shape-from-shading problem. Especially for symmetric objects, a symmetry constraint is proposed to improve the performance of existing shape-from-shading algorithm in this paper. The symmetry constraint is imposed to a conventional algorithm and then the improvement in the performance of 3-D shape reconstruction is proved by quantitatively comparing the depth and gradient errors.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.34S no.6
• /
• pp.46-57
• /
• 1997

In this paper, we propose an efficient algorithm for adaptation of tracking windwo, which improves tracking performance of a correlation-based video tracker by rejecting background effect originated from a time-varying target. Th eproposed adaptation algorithm ajdusts the size of a tracking window by using the ratio of spatial gradient power in target region to that in backgorund region, which is especially adequate for a correlation-based tracker. Experimental results for synthetic and real image sequences show that the proposed method adapts a tracking window well to a time-varying target and so greatly suppresses background effect, which makes improvement of trakcing performance.

• Journal of the Chungcheong Mathematical Society
• /
• v.29 no.1
• /
• pp.59-71
• /
• 2016

To obtain more accurate approximation of the true images in the deblurring problems, the weighted global generalized cross validation(GCV) function to the inverse problem with multiple right-hand sides is suggested as an efficient way to determine the regularization parameter. We analyze the experimental results for many test problems and was able to obtain the globally useful range of the weight when the preconditioned global conjugate gradient linear least squares(Gl-CGLS) method with the weighted global GCV function is applied.

• International Journal of Contents
• /
• v.8 no.4
• /
• pp.95-102
• /
• 2012

In this paper, we present a technique to extract the title areas from book cover images. A typical book cover image may contain text, pictures, diagrams as well as complex and irregular background. In addition, the high variability of character features such as thickness, font, position, background and tilt of the text also makes the text extraction task more complicated. Therefore, we propose a two steps efficient method that uses Histogram of Oriented Gradients and color information to find the title areas. Firstly, text localization is carried out to find the title candidates. Finally, refinement process is performed to find the sufficient components of title areas. To obtain the best result, we also use other constraints about the size, ratio between the length and width of the title. We achieve encouraging results of extracted title regions from book cover images which prove the advantages and efficiency of the proposed method.

• Journal of the Korean Society of Visualization
• /
• v.2 no.1
• /
• pp.32-38
• /
• 2004

A new PIV technique was developed for decreasing optical error which was created during simultaneous measurements of velocity fields at a wall-normal plane and wall-parallel plane by a plane PIV and a Stereo-PIV systems. For experimental study, two different Reynolds number based on momentum thickness, Re$_{ =514 and 934 were generated in a blowing type wind tunnel under the condition of zero pressure gradient. The two Nd:Yag laser systems and three CCD cameras were synchronized. to obtain instantaneous velocity fields at the same time. To avoid optical noise at the crossing line by the two laser light sheets, a new optical arrangement using polarization was applied. The obtained velocity fields show the existence of hairpin packet structure vividly and the idealized hairpin vortex signature is confirmed by experiment.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2012.11a
• /
• pp.29-30
• /
• 2012

본 논문은 고해상도의 깊이맵을 얻기 위해서 대응되는 색상 영상의 최대 변화도를 이용한 깊이맵 업샘플링 기술을 제안한다. 기존 알고리즘들이 인접한 화소의 깊이 값을 참조할 때 거리에 따른 가중치를 부여하는 것과 달리, 제안한 방법은 현재 화소와 참조 화소 사이의 최대 색차 변화도를 이용하여 가중치를 부여한다. 이런 접근 방법은 비슷한 색상의 물체가 서로 붙어 있거나 큰 크기의 객체가 존재할 경우에도 모두 올바른 가중치를 부여할 수 있다는 장점을 갖는다. 먼저, 색상 영상의 색차 성분에 대한 변화도 영상을 계산하고, 업샘플링하고자 하는 화소와 참조 화소 사이의 최단 경로 위에서 가장 큰 변화도를 취한다. 변화도가 클수록 다른 객체에 존재할 확률이 높기 때문에 변화도가 큰 참조 화소에는 작은 가중치를 부여하고, 이들의 가중합을 통해 최종 깊이 값을 계산한다. 제안한 방법을 이용하여 깊이맵을 업샘플링한 결과가 기존 알고리즘들에 비해 우수한 결과를 보였다.