• The Journal of the Korea Contents Association
• /
• v.16 no.5
• /
• pp.104-112
• /
• 2016

The Adaboost chooses a good set of features in rounds. On each round, it chooses the optimal feature and its threshold value by minimizing the weighted error of classification. The involved process of classification performs a hard decision. In this paper, we expand the process of classification to a soft fuzzy decision. We believe this expansion could allow some flexibility to the Adaboost algorithm as well as a good performance especially when the size of a training data set is not large enough. The typical Adaboost algorithm assigns a same weight to each training datum on the first round of a training process. We propose a new algorithm to assign different initial weights based on some statistical properties of involved features. In experimental results, we assess that the proposed method shows higher performance than the traditional one.

• Journal of Digital Contents Society
• /
• v.9 no.4
• /
• pp.687-695
• /
• 2008

One of the most significant characteristics of remote real time lecture, the emerging still image standard, is the OOI (Object of Interest) coding. The functionality of OOI is important in applications where certain parts of the image are more important than others. In such cases, these objects need to be encoded at higher quality than the background. JPEG2000 provides a number of OOI coding mechanisms. Many researchers have actively studied the preferred processing from the OOI coding methods to the new methods complementing them. But, there do not exist OOI coding methods suitable for all applications. Therefore, this study shows a criterion to select according to the application requirements for applying the best OOI coding method in JPEG2000 applications, and also shows the experimental results deciding the best parameters in the selected methods.

• Journal of Practical Engineering Education
• /
• v.14 no.2
• /
• pp.439-445
• /
• 2022

Due to the development of online media platforms and the COVID-19 incident, the mass production and consumption of digital video content are rapidly increasing. In order to select digital video content, users grasp it in a short time through thumbnails and teaser videos, and select and watch digital video content that suits them. It is very inconvenient to check all digital video contents produced around the world one by one and manually edit teaser videos for users to choose from. In this paper, keyframes are extracted based on YCbCr color models to automatically generate teaser videos, and keyframes extracted through clustering are optimized. Finally, we present a method of producing a teaser video to help users check digital video content by connecting the finally extracted keyframes.

• Journal of Broadcast Engineering
• /
• v.15 no.6
• /
• pp.791-802
• /
• 2010

Recent prevalence of progressive scan display such as LCD TV demands the quality improvement of existing deinterlacing techniques that convert interlaced scan images including HDTV broadcasting to progressive scan images. In this paper, we proposea motion adaptive deinterlacing technique which can be used for spatial methods, temporal methods, and the spatial-temporal methods can be used for the deinterlacing techniques selectively based on the threshold values calculated by the statistics of motion in the video contents. We also propose an improved spatial deinterlacing technique that adaptively use M-ELA and DOI based on the slant of edges that are obtained by Sobel operation. The improved picture quality of the proposed algorithm is confirmed by objective and subjective quality tests on many test image sequences.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.4
• /
• pp.661-674
• /
• 2021

Edwardsiellosis is difficult to treat in cultured olive flounder, Paralichthys olivaceus. It is present in the fish for a long period during all growth stages, and it often leads to mass mortalites. In this paper, the clinical patterns of Edwardsiellosis were analyzed by dividing the data into the whole-water temperature, low-water temperature, low-high water temperature, high-water temperature, and high-low water temperature groups based on various clinical signs of diseased cultured olive flounder using a decision tree technique. In the clinical sign patterns in the decision trees analyzed in the experiment, clinical signs in the liver, such as liver nodules, liver hemorrhages, and liver degeneration, were selected as the criteria for determining Edwardsiellosis. The selected clinical signs were known as the major clinical signs of Edwardsiellosis, and through consultation with fishery disease experts, the analysis confirmed that the clinical signs of Edwardsiellosis were successfully found in this study.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.8
• /
• pp.375-382
• /
• 2006

This paper is about how to identify emotional information and how to detect a specific emotion from speech signals. For emotion identification and detection task. we use long-term acoustic feature parameters and select the optimal Parameters using the feature selection technique based on F-score. We transform the conventional SVM into probabilistic output SVM for our emotion identification and detection system. In this paper we propose three approximation methods for log-likelihoods in a hypothesis test and compare the performance of those three methods. Experimental results using the SUSAS database showed the effectiveness of both feature selection and Probabilistic output SVM in the emotion identification task. The proposed methods could detect anger emotion with 91.3% correctness.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2006.11a
• /
• pp.313-316
• /
• 2006

디지털 핑거프린팅(Digital Fingerprinting)은 기 밀 정보를 디지털 콘텐츠에 삽입하는 측면에서는 디지털 워터마킹과 동일 하다고 볼 수 있으나 저작권자나 판매자의 정보가 아닌 콘텐츠를 구매한 사용자의 정보를 삽입함으로써 콘텐츠 불법 배포자를 추적할 수 있도록 한다는 점에서 워터마킹과 차별화된 기술이다. 이러한 핑거프린팅 기술은 소유권에 대한 인증뿐만 아니라 개인 식별 기능까지 제공해야 하므로 기존의 워터마킹이 갖추어야 할 요구사항인 비가시성, 견고성, 유일성과 더불어 공모허용, 비대칭성, 익명성, 조건부 추적성 등이 부가적으로 필요하다. 본 논문에서는 행렬의 한 열을 선택 후 쉬프팅 기법을 사용 하서 사용자 정보로 조합하여 핑거프린트를 생성하였다. 이렇게 생성된 핑거프린트 정보를 2레벨 웨이블릿 변환 영역 중 LH2, HL2, HH2 부대역에 삽입하였다. 쉬프팅 정보와 도메인 개념을 사용하여 보다 많은 사용자에게 핑거프린트 정보를 삽입할 수 있으며, 공모공격과 JPEG 압축에서도 최소한 1명 이상의 공모자를 검출할 수 있는 핑거프린팅의 기본 조건을 만족하였다.

• The Journal of the Korea Contents Association
• /
• v.14 no.1
• /
• pp.24-29
• /
• 2014

This paper proposes a kinect-based human motion recognition model for the 3D contents control after tracking the human body gesture through the camera in the infrared kinect project. The proposed human motion model in this paper computes the distance variation of the body movement from shoulder to right and left hand, wrist, arm, and elbow. The human motion model is classified into the movement directions such as the left movement, right movement, up, down, enlargement, downsizing. and selection. The proposed kinect-based human motion recognition model is very natural and low cost compared to other contact type gesture recognition technologies and device based gesture technologies with the expensive hardware system.

• The Journal of the Korea Contents Association
• /
• v.17 no.3
• /
• pp.329-336
• /
• 2017

In video compression, Motion Estimation(ME) limits the performance of image quality and generated bit rates. However, it requires much complexity in the encoder part. Multi-view video uses many cameras at different positions. Multi-view video coding needs huge computational complexity in proportion to the number of the cameras. To reduce computational complexity and maintain the image quality, an effective motion estimation method is proposed in this paper. The proposed method exploiting the characteristics of motion vector distribution and the motion of video. The proposed is a kind of a hierarchical search strategy. This strategy consists of multi-grid rhombus pattern, diagonal pattern, rectangle pattern, and refinement pattern. Experiment results show that the complexity reduction of the proposed method over TZ search method and PBS (Pel Block Search) on JMVC (Joint Multiview Video Coding) can be up to 40~75% and 98% respectively while maintaining similar video image quality and generated bit rates.

• Journal of Digital Contents Society
• /
• v.18 no.1
• /
• pp.141-147
• /
• 2017

Scalable extension of High Efficiency Video Coding (SHVC) standard uses the up-sampled residual data from the base layer to make a residual data in the enhancement layer. This paper describes an efficient algorithm for improving coding gain by using the filtered residual signal of base layer in the Scalable extension of High Efficiency Video Coding (SHVC). The proposed adaptive filter selection mechanism uses the smoothing and sharpening filters to enhance the quality of inter-layer prediction. Based on two filters and the existing up-sampling filter, a rate-distortion (RD)-cost fuction-based competitive scheme is proposed to get better quality of video. Experimental results showed that average BD-rate gains of 1.5%, 2.1%, and 1.7% for Y, U and V components, respectively, were achieved, compared with SHVC reference software 5.0, which is based on HEVC reference model (HM) 13.