• Proceedings of the Korean Information Science Society Conference
• /
• 2002.10d
• /
• pp.193-195
• /
• 2002

XML 기반의 MPEG-7은 사용자가 시스템을 통해서 원하는 정보를 얻을 수 있도록 멀티미디어컨텐츠의 다양한 정보를 기술하고 있다. 그러나 많은 샷으로 이루어진 MPEG-7 정보 저장 공간에서 질의가 들어왔을 때 원하는 정보를 찾기 위하여 모든 문서를 검색하는 것은 상당한 시간이 요구된다. 본 논문에서 이를 해결하기 위하여 동영상 MPEG-7 문서에 대한 효율적인 인덱싱 알고리즘을 제시한다. 본 논문에서 제안한 인덱싱 알고리즘의 적용으로 MPEG-7 검색에 있어서 효율적임을 할 수 있었다.

• Proceedings of the IEEK Conference
• /
• 2001.06d
• /
• pp.101-104
• /
• 2001

This paper proposes a unique shot boundary detection algorithm for the video indexing and/or browsing. Conventional methods based on the frame differences and the histogram differences are improved. Instead of using absolute frame differences, block by block based relative frame differences are employed. Frame adaptive thresholding values are also employed for the better detection. for the cases that the frame differences are not enough to detect the shot boundary, histogram differences are selectively applied. Experimental results show that the proposed algorithm reduces both the “false positive” errors and the “false negative” errors especially for the videos of dynamic local and/or global motions

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.2
• /
• pp.441-445
• /
• 2011

In this paper, we propose an algorithm for generating panoramic videos using fixed multiple cameras. We estimate a background image from each camera. Then we calculate perspective relationships between images using extracted feature points. To eliminate stitching errors due to different image depths, we process background images and foreground images separately in the overlap regions between adjacent cameras by projecting regions of foreground images selectively. The proposed algorithm can be used to enhance the efficiency and convenience of wide-area surveillance systems.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10b
• /
• pp.607-609
• /
• 2003

동영상 자동 분석은 비디오 데이터의 내용 기반 색인과 검색을 위한 첫 단계이다. 본 논문에서는 정형적인 구조를 가진 뉴스와는 달리 비정형적인 특성을 가진 축구 동영상에서 사용자의 관심이 되는 하이라이트 장면의 영상 특징을 이용하여 그 구조를 분석하여 하이라이트 장면을 검출하는 방법을 제안한다. 이전 연구를 토대로 그라운드영역과 골대 유무에 따라 하이라이트 후보 장면을 찾는 과정에서 경기마다 달라지는 임계값에 영향을 받지 않는 알고리즘을 제안하였다. 실험결과 제안된 방법이 여러 종류의 축구 경기 하이라이트 분석에 있어서 그 성능이 우수함을 확인할 수 있었다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.06a
• /
• pp.90-91
• /
• 2018

We propose a novel high dynamic range (HDR) video synthesis algorithm using alternatively exposed low dynamic range (LDR) videos. We first develop a superpixel-based illumination invariant correspondence estimation algorithm. Then, we propose a reliability weight to further improve the quality of the synthesized HDR frame. Experimental results show that the proposed algorithm provides high-quality HDR frames compared to conventional algorithms.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.366-369
• /
• 2021

본 논문에서는 동영상에서 일련의 과정을 거쳐 얻었던 학습데이터를 보다 간편하고 빠른 속도로 획득하는 방법을 제안한다. 음성과 영상 스트림을 처리하는 ffmpeg을 이용해 영상을 프레임화하고, 딥 러닝 기반의 YOLO 알고리즘을 사용하여 객체를 검출한다.

• Journal of Digital Contents Society
• /
• v.12 no.4
• /
• pp.429-436
• /
• 2011

Various documents, images, videos and other materials on the web has been increasing rapidly. Efficient search of those things has become an important topic. From keyword-based search, internet search has been transformed to semantic search which finds the implications and the relations between data elements. Many annotation processing systems manipulating the metadata for semantic search have been proposed. However, annotation data generated by different methods and forms are difficult to process integrated search between those systems. In this study, in order to resolve this problem, we categorized levels of many annotation documents, and we proposed the method to measure the similarity between the annotation documents. Similarity measure between annotation documents can be used for searching similar or related documents, images, and videos regardless of the forms of the source data.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2012.05a
• /
• pp.901-904
• /
• 2012

In this paper, we proposed the context-aware system which can estimate the information on the objects and transmit video information by utilizing multi-sensors. The proposed system is to reduce the excessive video information from a system capturing videos outdoor. This system uses the human-detect sensor attached on the multi-sensor board and four ultrasonic sensor to measure the object's size and movement speed, to recognize the human body's information, and finally to send videos. In order to assess the performance of the context-aware system based on the multi sensor, a comparison has been made between video system and human-detect sensor. As a result, The body human-detect sensor had more reliable images and transmitted information more effectively than when the images were sent by server without sensors attached.

• Fashion & Textile Research Journal
• /
• v.14 no.5
• /
• pp.701-712
• /
• 2012

This study examines the various expressions and immanent value of fashion and beauty style based on Craig Owens's Allegory theory. I analyzed four application elements of Borrow, Site Specificity, Accumulation of Strategy, and Hybridization in? Ladygaga's Music Videos. The results are as follows. 'Borrow' presents a kitsch style and playful Pop-art style, transformation of gender from Mini Mouse body suit, telephone headpiece, and can hair. 'Site specificity' presents the temporarity of fashion material through rebirth, aging, natural extinction from the chain over the black jump suit, crime scene tape, and skull-face makeup with masculine style. 'Accumulation of strategy' presents a futuristic chic fashion style from a layered style, retrospective fashion, repetition and duplication in Music Videos. It shows the physical beauty of an Asian warrior style in Poker Face. 'Hybridization' present Cyborg feminism and 'Will of Power' from iPOD LCD glasses and Pyro-Bra. The Pyro-Bra represents how the female body can be used as a weapon in the outfits of Lady Gaga. In addition, Immanent value is as follows. Textual interaction with high art is accomplished through a combination of contemporary social and cultural significance to understand the cultural code and to extend the value. Combined with high culture, popular music genre is accomplished through musical appreciation by a woman in fashion styling and sensual pleasures of the body as a tool to express a sublime advantage. Fashion style is accomplished by overcoming a self-transcendent body image representation. The way in which mutual coupling dismantling, destruction, and uncertainty is to re-launch the static, with a pluralistic context of Textuality.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.48 no.6
• /
• pp.77-87
• /
• 2011

In this paper, we propose a disparity correction technique to reduce the inherent visual discomfort while watching stereoscopic videos. The visual discomfort must be solved for commercial 3D display systems to provide natural stereoscopic videos to human eyes. The proposed disparity correction technique consists of horizontal and vertical disparity corrections. The horizontal disparity correction is implemented by controlling the depth budget of stereoscopic video using the geometric relations of a stereoscopic camera system. In addition, the vertical disparity correction is implemented by using a feature-based stereo matching algorithm. Conventional vertical disparity corrections have been done by only using camera calibration parameters, which still cause systematic errors in vertical disparities. In this paper, we minimize the vertical disparity as small as possible by using a feature-based correction algorithm. Through the comparisons of conventional feature-based correction algorithms, we analyze the performance of the proposed technique.