• Journal of Internet Computing and Services
• /
• v.6 no.5
• /
• pp.85-95
• /
• 2005

This study proposes a method of feature point extraction using scale-space filtering and a feature point tracking algorithm based on a texturedness similarity comparison, With well-defined operators one can select a scale parameter for feature point extraction; this affects the selection and localization of the feature points and also the performance of the tracking algorithm. This study suggests a feature extraction method using scale-space filtering, With a change in the camera's point of view or movement of an object in sequential images, the window of a feature point will have an affine transform. Traditionally, it is difficult to measure the similarity between correspondence points, and tracking errors often occur. This study also suggests a tracking algorithm that expands Shi-Tomasi-Kanade's tracking algorithm with texturedness similarity.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.10d
• /
• pp.286-289
• /
• 2007

기존 모바일 이미지 서비스와 차별화하여 고객이 원하는 기호에 맞춘 이미지를 휴대폰 배경으로 전달하는 서비스이다. 본 논문에서는 휴대폰으로 이미지를 전송해주는 모바일 서버의 구현, SK-VM 기반의 휴대폰에서 구동되는 어플리케이션 구현, MVC패턴을 적용한 JSP 기반의 웹서버의 구축을 기술하였고, 휴대폰 이미지를 관리하는 Tomcat 서버와 고객 정보 데이터베이스간의 필터링을 방법을 기술하였다.

• Journal of Broadcast Engineering
• /
• v.26 no.3
• /
• pp.283-295
• /
• 2021

This paper proposes a method for generating dense depth map using information of color images and depth map generated based on lidar based on self-organizing map. The proposed depth map upsampling method consists of an initial depth prediction step for an area that has not been acquired from LiDAR and an initial depth filtering step. In the initial depth prediction step, stereo matching is performed on two color images to predict an initial depth value. In the depth map filtering step, in order to reduce the error of the predicted initial depth value, a self-organizing map technique is performed on the predicted depth pixel by using the measured depth pixel around the predicted depth pixel. In the process of self-organization map, a weight is determined according to a difference between a distance between a predicted depth pixel and an measured depth pixel and a color value corresponding to each pixel. In this paper, we compared the proposed method with the bilateral filter and k-nearest neighbor widely used as a depth map upsampling method for performance comparison. Compared to the bilateral filter and the k-nearest neighbor, the proposed method reduced by about 6.4% and 8.6% in terms of MAE, and about 10.8% and 14.3% in terms of RMSE.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2004.05b
• /
• pp.721-723
• /
• 2004

As the Internet use has been spreading worldwide, illegal and harmful contents have been increasing on the Internet, which has become a very serious social problem. To prevent children form exposing themselves to such illegal and harmful contents on the Internet, harmful information protection systems have been developed. We examine component technologies of harmful information protection systems including text and image-based filtering solutions as well as url-based filtering solution. Also we examine the related trends and strategies which effectively prevent access to the harmful contents.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.4
• /
• pp.766-772
• /
• 2007

H.264/AVC is a new international standard for the compression of video images, in which a deblocking filter has been adopted to remove blocking artifacts. This paper proposes an efficient architecture of deblocking filter in H.264/AVC. By making good use of data dependence between neighboring $4{\times}4$ blocks, the memory size is reduced and the throughput of the deblocking filter processing is increased. Compared to the conventional deblocking filters, the proposed architecture enhances the performance of deblocking filter processing from 1.75 to 4.23 times. Hence the proposed architecture is able to perform real-time deblocking of high-resolution($2048{\times}1024$) video applications.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.41 no.11
• /
• pp.1507-1514
• /
• 2016

Full-image guided filter reflects all pixels of image in filtering by using weight propagation and two-pass model, whereas the existing guide filter is processed based on the kernel window. Therefore the computational complexity can be improved while maintaining characteristics of guide filter, such as edge-preserving, smoothing, and so on. In this paper, we propose an efficient VLSI architecture for the full-image guided filter by analyzing the data dependency, the data frequency and the PSNR analysis of the image in order to achieve enough speed for various applications such as stereo vision, real-time systems, etc. In addition, the proposed efficient scheduling enables the realtime process by minimizing the idle period in weight computation. The proposed VLSI architecture shows 214MHz of maximum operating frequency (image size: 384*288, 965 fps) and 76K of gates (internal memory excluded).

• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.5
• /
• pp.541-550
• /
• 2003

It is important for the strategy of product sales to investigate the consumer's sensitivity and preference degree in the environment that the process of material development has been changed focusing on the consumer renter. In the present study, we propose the Fashion Design Recommender System (FDRS) of textile design applying collaborative filtering personalization technique as one of methods in the material development centered on consumer's sensibility and preferences. In collaborative filtering personalization technique based on textile, Pearson Correlation Coefficient is used to calculate similarity weights between users. We build the database founded on the sensibility adjective to develop textile designs by extracting the representative sensibility adjective from users' sensibility and preferences about textile designs. FDRS recommends textile designs to a consumer who has a similar propensity about textile. Ultimately, this paper sugeests empirical applications to verify the adequacy and the validity on this system with the development of Fashion Design Recommender System (FDRS)

• Journal of Korea Multimedia Society
• /
• v.7 no.1
• /
• pp.44-53
• /
• 2004

This paper aims at presenting high quality at low resolution apply by a new antialiasing method for voxelized implicit surfaces. Implicit surfaces create a unique type of 3D-modeling. Some use of implicit surfaces are scientific and medical visualization, animation, medical simulation and interactive modeling. One of previous antialiasing methods for implicit surfaces presented by raytracing or texture mapping is making use of a stochastic sampling. But this method requires more calculation time and costs which is caused by complicated and difficult implicit functions. In the meanwhile, voxelized implicit surfaces generally use high resolution for good quality images but it costs to generate. In order to this problem, this paper suggests a shifted double Z-buffer which is very simple, more efficient and easy. Tn addition, there are applied box-filter and tent-filter to the double Z-buffer antialiasing method for better images. For results this method generate high quality image and it is easy to apply to various filters and is able to extend to multi Z-buffer.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.45-48
• /
• 2001

립리딩은 잡음 환경 하에서 음성 인식 성능을 향상을 위해 영상정보를 이용한 바이모달(bimodal)음성인식으로 연구되었다[1][2]. 그 일환으로 이미 영상정보를 이용한 립리딩은 구현되었다. 그러나 현재까지의 시스템들은 환경의 변화에 강인하지 못하다. 본 논문에서는 이미지 기반 립리딩 방법을 적용하여 입술 영역을 보다 안정적으로 찾아 성능을 향상 시켰다. 그러나 이 방법은 많은 데이터량을 처리해야 하므로 전처리 과정이 필요하다. 전처리로 입력영상을 그레이 레벨로 변환하는 방법과, 입술을 반으로 접는 방법, 그리고 주성분 분석(PCA: Principal Component Analysis)을 사용하였다. 또한 인식성능 향상을 위해 음성에서 잡음 제거나 분석$\cdot$합성에 효과적인 성능을 보이는 RASTA(Relative Spectral)필터를 적용하여 시간 영역에서의 변화가 적은 성분이나 급변하는 성분, 그 밖의 잡음 등을 제거하였다. 그 결과 $72.7\%$의 높은 인식 성능을 보였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.421-424
• /
• 2011

최근에 스마트폰이 발달하고 대부분의 모바일 기기에 카메라가 달리면서 카메라를 이용한 애플리케이션 또한 늘어나고 있는데 기존의 PC상에서 로고 인식등을 위해 사용되는 SURF를 이용한 이미지 매칭에는 유클리드 거리 계산을 사용하고 있다. 그러나 이 방법으로는 PC보다는 사양이 낮은 모바일 기기에 적용하기에는 기존에 사용하고 있는 방법이 인식할 이미지마다 모든 특징점을 비교하는 방법을 사용하기 때문에 연산량이 높은 편이다. 본 논문에서는 미리 인식할 이미지를 뉴럴넷에 학습시킨 뒤, 뉴럴넷을 필터링으로 사용하여 일부의 특징점만을 비교해 연산량을 줄여서 속도를 향상시키는 방법을 제안하였으며 이를 이용하여 대략 30%가량의 성능 향상이 나타난 것을 알 수 있었다.