-
Article type: Cover
Pages
Cover1-
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Article type: Index
Pages
Toc1-
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Akinori TAGUCHI, Terumasa AOKI, Hiroshi YASUDA
Article type: Article
Session ID: BCS2001-59
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
Many people would like to enjoy fashions and to buy fashionable chothes. But actually they cannot. With the variety of fashion, there are many choices. But we cannot get all information of fashion when we take it. In general, when we take clothes, we cannot try on all clothes in a shop. This is because most of us hate trying on many clothes, a service of shop assistant and wasteful time of changing other clothes. Also when we will order new clothes in order-made shop, we should image a production and ourselves who are trying it on. In this paper, we propose Digital-Mirror system. This system is a kind of mirror and it enables to create images of trying on clothes we want without trying it on actually. Using this system, people can enjoy fashions more than current shopping.
View full abstract
-
Shuichi AOKI, Terumasa AOKI, Hiroshi YASUDA
Article type: Article
Session ID: BCS2001-60
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
The importance of picture retrieval is increasing by reason of spread of multimedia content, these days. Picture retrieval with rough sketch that take automatically tagged metadata, is one of efficient methods, when we retrieve target picture of some person furthermore after retrieving with proper name. Picture retrieval with rough sketch, though, had such problems as its efficiency of retrieval dependent on quality of user's sketch. We propose we take symbolic rough sketch as query for picture retrieval. In this paper, on the assumption that we use some automatically tagged descriptors of MPEG-7, we propose transforming algorithm for matching query rough sketch appropriately. Additionally we give proof of its effectiveness.
View full abstract
-
Mizuki KANADA, Kentaro DOBASHI, Shinichi TAKAGI, Akihisa KODATE, Hidey ...
Article type: Article
Session ID: BCS2001-61
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
Today, the environment to personally edit the video is being put into place by wide spreading of digital cameras, price reducing of video editing software, and its high functionality. However, generally, editing video takes a lot of time. This leads to the need of video editing support systems. For this, this report will study a method to extract more robust indexing parameter not only by using camera parameters derived from moving vectors, but also with the information from intervals newly divided from shots using color information.
View full abstract
-
Shin'ichi TAKAGI, Akihisa KODATE, Ryouhei OOGUSHI, Youhei OKUNO, ...
Article type: Article
Session ID: BCS2001-62
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We have developed a video-video searching system uding what we had proposed and called "Videoprint, "which had been adopted to MPEG7 as Visual Color Layout Descriptor. This system searches for TV commercials, by extracting and comparing it with the Videoprint of the commercials. Although we use TV commercials as an example here, video that this system can search for is not limited to commercials, but could be arbitrary.
View full abstract
-
Masaru SUGANO, Yasuyuki NAKAJIMA, Hiromasa Yanagihara
Article type: Article
Session ID: BCS2001-63
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We have already proposed the automatic summary extraction method from the MPEG coded audio-visual data, by using audio-visual characteristics on compressed domain. Our previous method assumes that the visual features such as motion are known a priori. In this paper, we show the extended method for extracting the audio-visual summary without any knowledge of the content by adaptively evaluating the visual features. Additionally, in order to extract summary, our method utilizes some descriptors defined in the MPEG-7 standard and can generate some MPEG-7 descriptions; our method is very flexible and can be easily applicable to the MPEG-7 applications.
View full abstract
-
Haruhisa KATO, Hiromasa YANAGIHARA, Yasuhiro NAKAJIMA
Article type: Article
Session ID: BCS2001-64
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We proposed the fast caption detect algorithm from MPEG coded data. This paper proposes the caption detection for news indexing. The captions are distinguished in order to create the effective news index. The caption geometric information and caption appearance time are used to determine how the caption is effective for index.
View full abstract
-
Yasuhiro TAKISHIMA, Akio YONEYAMA, Satoshi MIYAJI, Hiromasa YANAGIHARA ...
Article type: Article
Session ID: BCS2001-65
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper proposes the comprehensive architecture of a novel content-authoring system specific for mobile communication services. Considering unique requirements to mobile communications such as receiver property, transmission environment, and content-creating conditions, the authoring-system should be equipped with a high-quality low-bit-rate video encoder, a functional AV editor, a supportive caption inserter, a fast format converter, and a user-friendly previewer. Software system based on the proposed architecture is also introduced.
View full abstract
-
Tomokazu Murakami, Masahiro Kageyama, Hisao Tanabe
Article type: Article
Session ID: BCS2001-66
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
For the integration of the TV broadcasting and the internet environment, we have proposed an information searching system of the TV contents related documents on WWW. By using this system, users can retrieve information about the TV contents by selecting the areas of what they see on the TV display. Now in this paper, we propose a new chatting system, developed from above-mentioned search system, by which users can communicate with each other about the TV contents they are seeing with pointing the objects that they want to talk about. We report the constructed PC based system, and moreover, we propose a new bilateral communication system by the integration of search and chat system.
View full abstract
-
Koichiro TSUGE, Yasuyuki NOGAMI, Yoshitaka MORIKAWA
Article type: Article
Session ID: BCS2001-67
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper, we propose image digital watermark method which has robustness against image compression and clipping. In this method, we first at first take discrete wavelet transform of original image, and get out multi-resolution approximation (MRA). Next, we spread spectrum of the MRA by Pseudo Noise (PN) sequence, and then take 1D-discrete Fourier transform (DFT) of each column. We vary the DFT coefficients according to each bit of watermark information. This method has the features which the third eavesdropper, who don't know the PN sequence, can not decipher the information data and alter them. In detection process, we take the DFT difference between PT-spread MRA of original image and that of watermark-embedded image and calculate shift of the DFT coefficients. Therefore, this method has the advantage that we can detect the watermark information even from the clipping image.
View full abstract
-
Hiroki OGO, Nobumoto YAMANE, Yoshitaka MORIKAWA
Article type: Article
Session ID: BCS2001-68
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper, an impulse noise removal method for images is proposed. This method is a switching filter using adaptive Wiener filter based on the Gaussian mixture distribution model in the interpolation process. The adaptive restoration filter is designed as the Wiener filter that minimize mean square error for the Gaussian stationary processes adopted to the local image signals. A simulation result shows that proposed method improves S/N ratio from 1.5 to 2.5 dB compared with conventional switching filter using median filter in its restoration process.
View full abstract
-
Article type: Appendix
Pages
App1-
Published: December 13, 2001
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS