• Journal of the Korea Society of Computer and Information
• /
• v.16 no.1
• /
• pp.53-59
• /
• 2011

In many face recognition applications such as security systems, it is assumed that upright faces are given to the system. In order for the system to be used in more general environments, the system should be able to deal with the rotated faces properly. It is a generally used approach to rotate the face detection window and apply face detector repeatedly to detect a rotated face in the given image. But such an approach requires a lot of computation time. In this paper, a method of extracting the axis of symmetry for a given set of points is proposed. The axis of symmetry for the edge points in the face detection window is extracted in a way that is fast and accurate, and the face detector is applied only for that direction. It is shown that the mean and standard deviation of the symmetry detection error is $0^{\circ}$ and $3^{\circ}$ respectively, for the database used.

• Journal of Korea Technical Association of The Pulp and Paper Industry
• /
• v.40 no.2
• /
• pp.37-42
• /
• 2008

Defects of corrugated board were limiting factors to the corrugated industry's growth. On-line detecting systems of defects are beneficial to increase profits of the corrugated board companies by reducing base paper consumption, increasing process reliability, and increasing productivity. By replacing naked eye inspection of defects to a defects detecting system, continuous inspection without fatigue may guarantee final products quality. The system was developed, which was consisted of line scan CCD camera, lens, illuminating parts, high speed image processor, software, various input parts, and output parts. First installation location of the system was at before fluting process after base paper unwinding, and surface of liner board was inspected by the system. Second installation location was after fluting process and combining process for liner board. Production loss includes ragged appearance of edge, irregular flute shape, wrong combination of flute, score cracking, defects in base paper, and flute cracking. The production loss was analyzed before and after the system installation at a commercial domestic mill. The production loss ratio was defined as a production loss weight per total production weight. The production loss ratio before the installation was decreased a lot from 1.28%(during 3 months before installation) to 0.76%(during 12 months after the first installation) and to 0.25%(during 6 months after the second installation).

• Proceedings of the Korean Vacuum Society Conference
• /
• 1999.07a
• /
• pp.71-71
• /
• 1999

현재 소자 제작에 응용되는 수소화된 비정질 실리콘은 PECVD 방법으로 제작하는 것이 보편적인 방법이다. 그러나 비정질 실리콘 박막 트랜지스터는 band gap edge 근처에서 국재준위가 많아 mobility가 작으며 상온에서 조차 불안정하여 신뢰성이 높지 않고, 도핑된 비정질 실리콘의 높은 비저항 등의 단점으로 인하여 고속 회로에 응용이 불가능하다. 반면 다결정질 실리콘 박막 트랜지스터는 a-Si:H TFT 에 비해 재현성이 우수하고 high resolution, high resolution, high contrast LCD에 응용할 수 있다. 하지만, 다결정 실리콘의 grain boundary로 인해 단결정에 비해 많은 defect 들이 존재하여 전도성을 감소시킨다. 따라서 Mobility를 증가시키기 위해서 grain size를 증가시키고 grain boundary 내에 존재하는 trap center를 감소시켜야 한다. 따라서 본 실험에서는 PECVD 장비로 초기 기판을 plasma 처리하여 다결정 실리콘 박막을 제작하여, 기판 처리에 대한 다결정 실리콘 박막의 성장의 특성을 조사하였다. 실험 방법으로는 PECVD 시스템을 이용하여 SiH4 gas와 H2 gas를 선택적으로 증착시키는 LBL 방법을 사용하여 $\mu$c-Si:H 박막을 제작하였다. 비정질 층을 gas plasma treatment 하여 다결정질 실리콘의 증착 initial stage 관찰을 주목적으로 관찰하였다. 다결정 실리콘 박막의 구조적 성질을 조사하기 위하여 Raman, AFM, SEM, XRD를 이용하여 grain 크기와 결정화도에 대해 측정하여 결정성장 mechanism을 관측하였다. LBL 방법으로 증착시킨 박막의 Raman 분석을 통해서 박막 증착 초기에 비정질이 증착된 후에 결정질로 상태가 변화됨을 관측할 수 있었고, SEM image를 통해서 증착 회수를 증가시키면서 grain size가 작아졌다 다시 커지는 현상을 볼 수 있었다. 이 비정질 층의 transition layer를 gas plasma 처리를 통해서 다결정 핵 형성에 영향을 관측하여 적정한 gas plasma를 통해서 다결정질 실리콘 박막 증착 공정을 단축시킬 수 있는 가능성을 짐작할 수 있었고, 또한 표면의 roughnes와 morphology를 AFM을 통하여 관측함으로써 다결정 박막의 핵 형성에 알맞은 증착 표면 특성을 분석 할 수 있었다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.38 no.4
• /
• pp.411-421
• /
• 2001

The generalized symmetry transform evaluates symmetry without segmentation and extracts regions of interest in an image by combining locality and reflectional symmetry The demand that the symmetry transform be local is reflected by the distance weight function. When calculating large regions-of-interest, we should select a large standard deviation of distance weight function. But such a large standard deviation makes the execution time increase in the second power of r, which is a radius of search area. In this paper we propose modified scan line based GST with selectively directional attention to improve time complexity The symmetry map of our proposed GST is found to be very similar to that of the existing GST. However the computation time of the proposed GST increases linearly with respect to r because our proposed GST evaluates symmetry between a pair of edge pixels along the scan lines. The GST computation decreases considerably when the proposed GST is peformed with selectively directional attention in case of large r. Several experiments in this paper demonstrate the time efficiency and the usefulness of our proposed GST.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.52 no.9
• /
• pp.89-95
• /
• 2015

In this paper, we propose a real-time two-dimensional bar code detection system even at long range using a vision technique. We first perform short-range detection, and then long-range detection if the short-range detection is not successful. First, edge map generation, image binarization, and connect component labeling (CCL) are performed in order to select a region of interest (ROI). After interpolating the selected ROI using bilinear interpolation, a location symbol pattern is detected as the same as for short-range detection. Finally, the symbol pattern is arranged by applying inverse perspective transformation to localize bar codes. Experimental results demonstrate that the proposed system successfully detects bar codes at two or three times longer distance than existing ones even at indoor environment.

• The Research Journal of the Costume Culture
• /
• v.25 no.5
• /
• pp.611-628
• /
• 2017

As a plan for establishing Korea's cultural identity and its competitive edge in the world market and for enhancing Korea's cultural status, creative and unique high value-added cultural products need to be developed utilizing our inherent cultural assets. Accordingly, this study focused on the development of the design of fashion cultural products that utilize the convergence of Hangeul our peculiar font style and Korea's cultural heritage, which is registered as part of UNESCO's World Heritage. A design method was devised that converges archetypal images of cultural property with the unique Hangeul font in a way that targets Korea's symbolic architectures. The symbolic architecture includes Korea's world-heritage pagoda architecture, such as Seokgatap pagoda and Dabotap pagoda at Bulguksa temple. It also included the architecture of royal palace, such as Injeongjeon hall at Changdeokgung palace. Finally, it also included the architecture of the fortress wall, such as Paldalmun gate in Hwaseong fortress. Thus, by developing cultural assets made from a convergence between architecture and the Hangeul font as a consumer-product image that has universality, the possibility of cultural products was pursued by applying color planning after an analysis that involved extracting the compositional colors of the flags of the world. This research and approach will lead to opportunities for further progress for Korea's cultural products in the global market as a results of additional recognition for their value, excellence, and universal appeal.

• Journal of Korea Multimedia Society
• /
• v.14 no.8
• /
• pp.981-991
• /
• 2011

MIP(Maximum Intensity Projection) is a volume rendering technique which is essential for the medical imaging system. MIP rendering based on the ray casting method produces high quality images but takes a long time. Our aim is improvement of the rendering speed using GPGPU(General-purpose computing on Graphic Process Unit) technique. In this paper, we present the ray casting algorithm based on CUDA(an acronym for Compute Unified Device Architecture) which is a programming language for GPGPU and we suggest new acceleration methods for CUDA. In detail, we propose the block based space leaping which skips unnecessary regions of volume data for CUDA, the bisection method which is a fast method to find a block edge, and the initial value estimation method which improves the probability of space leaping. Due to the proposed methods, we noticeably improve the rendering speed without image quality degradation.

• Journal of KIISE:Software and Applications
• /
• v.34 no.10
• /
• pp.929-936
• /
• 2007

The novel vision-based scheme for real-time extracting traffic parameters is proposed in this paper. Detecting and tracking of vehicle is processed at local region installed by operator. Local region is divided to segmented regions by edge and frame difference, and the segmented regions are classified into vehicle, road, shadow and headlight by statistical and geometrical features. Vehicle is detected by the result of the classification. Traffic parameters such as velocity, length, occupancy and distance are estimated by tracking using template matching at local region. Because background image are not used, it is possible to utilize under various conditions such as weather, time slots and locations. It is performed well with 90.16% detection rate in various databases. If direction, angle and iris are fitted to operating conditions, we are looking forward to using as the core of traffic monitoring systems.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.4C
• /
• pp.371-379
• /
• 2009

In this paper, we proposed new de-interlacing algorithm for video data using intensity gradient filter and median filter with texture detection in the image block. We first introduce the texture detection. According to texture detection, the current region is determined into smooth region or texture region. In case that the smooth region interpolated by median filter. In addition, in case of the texture region, we calculate missing pixel value using intensity gradient filter. Therefore, we analyze the local region feature using the texture detection and classify each missing pixel into two categories. And then, based on the classification result, a different de-interlacing algorithm is activated in order to obtain the best performance. Experimental results show that the proposed algorithm performs well with a variety of moving sequences compared with conventional intra-field method in the literature.

• Journal of KIISE
• /
• v.45 no.1
• /
• pp.30-35
• /
• 2018

GuessWhat?! is a game in which two machine players, composed of questioner and answerer, ask and answer yes-no-N/A questions about the object hidden for the answerer in the image, and the questioner chooses the correct object. GuessWhat?! has received much attention in the field of deep learning and artificial intelligence as a testbed for cutting-edge research on the interplay of computer vision and dialogue systems. In this study, we discuss the objective function and characteristics of the GuessWhat?! game. In addition, we propose a simple solver for GuessWhat?! using a simple rule-based algorithm. Although a human needs four or five questions on average to solve this problem, the proposed method outperforms state-of-the-art deep learning methods using only two questions, and exceeds human performance using five questions.