The purpose of the paper is as in the following. As the broadcasting environment is changing and the nonlinear viewing patterns are getting generalized, the needs for digital convergence type 본 연구의 배경 및 목적은 다음과 같다. 최근 비선형 시청행위 확산에 따라 고정형TV, 웹TV 및 N스크린 서비스 등을 기반으로 하는 디지털 융합형 VOD에대한 소비자들의 요구가 증가하고 있다. 그러나 현재 VOD 시스템은 VOD 콘텐츠 생성부분을 제외한 모든부분이 시스템적으로 자동화되어 있다. 따라서 본 논 문에서는 VOD 콘텐츠 생성부분을 자동화하는 방법에 관해 연구하였다. 본 연구의 연구방법 및 내용, 그 결과는 다음과 같다. 본 연구에서는 우선 각 방송사의 프로그램 구성을 분석하였고, 분석 결과 방송프로그램 전환 시 페이드인, 페이드 아웃에 의한 흑레벨이 존재하는 것을 발견하였으나, 일부 방송 프로그램에는 존재하지 않아 일반화된 프로그램 특징점으로 사용할 수 없었다. 따라서 본 논문에서는 각 프로그램의 시작지점과 종료지점의 특징점인 그래픽 패턴을 추출 지정하고, 실시간으로 베이스밴드 영상에서 지정된 시작, 종료지점의 특징점 그래픽 패턴과 매칭함으로써 VOD 콘텐츠의 시작지점과 종료지점을 인지할 수 있게 하였다. 또한 이를 VOD 인코더와 연동함으로써 VOD 콘텐츠 생성을 자동화 하였고 이를 IPTV에 직접 적용한 결과, 시작과 종료지점이 100% 검출되었고, 실시간으로 파일을 인코딩함으로써 지연시간 없이 VOD 콘텐츠를 자동 생성할 수 있게 되었다. VOD service of consumers are increasing. But today’s VOD system is all automated, except the part for posting contents. This study is focused on the methods for automation of the content posting. The research method, content, and result of the paper is as in the following. This paper analyzed the program formation of all broadcasting companies in Korea. It is clear that there are black levels found between the fade in and fade out of the program transitions. As it does not exist in certain programs, however, it is impossible to use a black level as a generalized key point of a program. We carefully applied a specific graphic pattern as the key method in order to determine a starting and ending point of a program. As a result, it is possible to fully recognize the starting and ending point in VOD content by matching them in real time with the key points for a baseband video. It is also possible to automate VOD content posting by interconnecting a proposed scheme with the VOD encoder. This approach was implemented and thoroughly tested on IPTV services proving complete recognition of a program and encoding the file in real time, resulting in a non-delayed and automated posting of the VOD content.

The purpose of the paper is as in the following. As the broadcasting environment is changing and the nonlinear viewing patterns are getting generalized, the needs for digital convergence type VOD service of consumers are increasing. But today’s VOD system is all automated, except the part for posting contents. This study is focused on the methods for automation of the content posting. The research method, content, and result of the paper is as in the following. This paper analyzed the program formation of all broadcasting companies in Korea. It is clear that there are black levels found between the fade in and fade out of the program transitions. As it does not exist in certain programs, however, it is impossible to use a black level as a generalized key point of a program. We carefully applied a specific graphic pattern as the key method in order to determine a starting and ending point of a program. As a result, it is possible to fully recognize the starting and ending point in VOD content by matching them in real time with the key points for a baseband video. It is also possible to automate VOD content posting by interconnecting a proposed scheme with the VOD encoder. This approach was implemented and thoroughly tested on IPTV services proving complete recognition of a program and encoding the file in real time, resulting in a non-delayed and automated posting of the VOD content.