The Journal of The Institute of Image Information and Television Engineers
Online ISSN : 1881-6908
Print ISSN : 1342-6907
ISSN-L : 1342-6907
Search
Advanced search
OR
Browse
Search
Volume 62, Issue 2
Displaying 1-23 of 23 articles from this issue
Show all abstracts
    • |<
    • <
    • 1
    • >
    • >|
  • A Visual Similarity Metric for View-based Web Page Retrieval
    Toru Furusawa, Yasuyuki Watai, Toshihiko Yamasaki, Kiyoharu Aizawa
    2008 Volume 62 Issue 2 Pages 209-215
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.209
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    We are developing a system for retrieving web browsing history based on visual information such as design and layout. This system's performance would be improved by refining the sketch retrieval results. For this purpose, we have developed “color-layout signature,” a visual similarity metric for web pages. We showed that use of the developed metric improves retrieval performance from 27%(conventional method) to 42%.
    View full abstract
    Download PDF (8249K)
  • A Motion Parsing and Composition Technique for Editing Time-Varying Mesh
    Jianfeng Xu, Toshihiko Yamasaki, Kiyoharu Aizawa
    2008 Volume 62 Issue 2 Pages 216-221
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.216
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    A motion parsing and composition technique for editing time-varying mesh (TVM) is described. TVM is a sequence of 3D mesh models, reproducing real-world objects, that is generated from multiple-view images using a shape-from-silhouette algorithm. Because TVM generation is very costly, efficient and effective reuse of TVM is an important issue. A motion database of generated TVM was established by motion analysis using histogram-based feature vectors so that the TVM can be easily accessed and reused. The motion synthesis to compose a new TVM sequence is also proposed. This technique will enable the efficient and effective reuse of TVM.
    View full abstract
    Download PDF (6773K)
  • Educational Applications for Learning Rhythmic Gymnastics Rules Using Motion Data
    Asako Soga, Yuka Myojin
    2008 Volume 62 Issue 2 Pages 222-226
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.222
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    We have developed educational applications to support the learning of rhythmic gymnastics rules using motion-captured animations. We developed a 3D rule book using Web3D technologies and an AR-RG system using augmented reality technologies. The 3D rule book plays 3DCG animations corresponding to the basic motions of rhythmic gymnastics. The AR-RG system helps users understand complicated rules requiring a combination of a basic motion and an apparatus action. These applications were evaluated by active athletes as useful learning tools.
    View full abstract
    Download PDF (5861K)
  • A Video Skimming Method for Detecting “Laughter” Scenes in Consumer Generated Videos
    Go Irie, Kota Hidaka, Naoya Miyashita, Takashi Satou, Yukinobu Taniguc ...
    2008 Volume 62 Issue 2 Pages 227-233
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.227
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    A method is described for detecting a “laughter” scene in a Consumer Generated Video (CGV). The growing number of CGVs is growing, so a video skimming method that allows viewers to quickly find “enjoyable” scenes would be useful. An experiment showed that viewers tend to find CGVs that include a “laughter” scene “enjoyable”. On that basis, we developed a method for detecting “laughter” scenes using prosodic information such as pitch, energy, power spectrum densities and their differential components to estimate the probability of laughter in the soundtrack. In this method, a generalized state space model, which consists of acoustic models and a state-transition model, calculates the probability of laughter to detect a “laughter” scene. An experiment showed that the precision rate of the method was about 83%, which suggests that the method enables effective video skimming.
    View full abstract
    Download PDF (8433K)
  • An Interactive Tool for Editing Shadows in Cartoon Animation
    Hidehito Nakajima, Eiji Sugisaki, Shigeo Morishima
    2008 Volume 62 Issue 2 Pages 234-239
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.234
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    Shadows in cartoon animation are generally used for dramatizing scenes. In hand-drawn animation, these shadows reflect the animators' intention and style rather than physical phenomena. On the other hand, shadows in 3 DCG animation are photorealistically rendered, and animators can not fully reflect their intention. This is because, in 3DCG animation, shadows are automatically generated once the light source is defined. Therefore, we develop an interactive tool for editing shadows that combines the advantages of hand-drawn animation and 3DCG technology. the advantage of our tool is that shadow attributes are inherited once animators edit the shape and location of shadows. Animators are only required mouse operations for editing shadows. Consequently, our tool enables animators to create shadows automatically and easily to reflect their intention and style.
    View full abstract
    Download PDF (1209K)
  • A Method for Generating Lecture Videos Relevant to Viewers from High Resolution Images
    Yudai Shinoki, Hironobu Fujiyoshi
    2008 Volume 62 Issue 2 Pages 240-246
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.240
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    A method for producing a dynamic lecture video from high resolution video recorded by an HDV camcorder is described. In this method, the high resolution video is cropped to focus on the region of interest, such as the instructor. Because this method uses virtual camerawork modeled on broadcast camera techniques, the generated video is very similar to professional broadcast video. Moreover, by focusing the viewer's attention on the chalkboard, the generated video is useful for students. We evaluated our method with subjective experiments and confirmed its effectiveness.
    View full abstract
    Download PDF (7265K)
  • Hitch-Haiku: Japanese Haiku Poem Creation Support System by Computer
    Naoko Tosa, Hideto Obara, Michihiko Minoh, Seigo Matsuoka
    2008 Volume 62 Issue 2 Pages 247-255
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.247
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    The authors maintain that there are unique and common forms of cultures that humans set aside such as the behavior and grammar within each culture. We suggest a computer model of this and a method of interactive expression and experiencing cultural understanding using IT called "cultural computing". This paper describes Hitch Haiku, a system based on cultural computing that interactively aids users in generating haiku, the world's shortest poems with imagery-maximizing mechanisms. First, "kire-ji" and particles are added to the word (s) input by a user to make a five or seven-character phrase, then phrases including terminology related to the user's input are located in a phrase database holding examples of haiku from the literary calendar, ensuring the cultural validity of the haiku. These phrases are then "hitched" together to generate a haiku. Although the haiku generated by this system have periodically resulted in flawed haiku, the ability to generate haiku that support the expansion of users' minds has been confirmed through assessment experimentation.
    View full abstract
    Download PDF (1880K)
  • Human Region Segmentation Based on Bayesian Estimation Using Chroma and Thermal Video
    Kohji Sawada, Keita Takahashi, Takeshi Naemura
    2008 Volume 62 Issue 2 Pages 256-261
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.256
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    We have been developing a thermo-key system that segments human regions from video in real-time using thermal information. We developed a new segmentation method for this system based on Bayesian estimation, which uses both chroma and thermal information integratively. Segmentation algorithms were previously based on a simple threshold processing of thermal images, which often results in wrong segmentation of objects/environments whose temperatures are close to body temperature. Trained with only a few sets of chroma and thermal images, our algorithm can extract human regions from video with much higher accuracy compared to previous algorithms.
    View full abstract
    Download PDF (5508K)
  • Picture Quality Assessment Test for Digitally Coded Motion Pictures in Multi-Formats
    Eisuke Nakasu, Kikufumi Kanda, Atsuro Ichigaya, Masaaki Kurozumi, Yuki ...
    2008 Volume 62 Issue 2 Pages 262-270
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.262
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    We present a subjective quality assessment scheme for digitally coded motion pictures in multi-formats. Several scanning formats are available in digital broadcasting, and broadcasters can select an adequate delivery format according to the overall picture quality and transmission bit-rate. Thus, the overall quality of digitally coded pictures should be evaluated in terms of both coding degradation and image resolution. This assessment-test scheme enables a direct comparison of picture quality of different scanning formats based on the double stimulus impairment scale (DSIS) method recommended in Rec. ITU-R BT.500. We show results of assessment tests performed on 1080i, 720p, and 480p videos encoded in the MPEG-2 video standard and give some consideration to picture quality in digital broadcasting.
    View full abstract
    Download PDF (2684K)
  • Fast Bilateral Filtering of Vector-Valued Multiple Images with Decomposition of Weights
    Kohei Inoue, Kiichi Urahama
    2008 Volume 62 Issue 2 Pages 271-274
    Published: February 01, 2008
    Released on J-STAGE: May 01, 2010
    DOIhttps://doi.org/10.3169/itej.62.271
    JOURNAL FREE ACCESS
    Show abstractHide abstract
    We present a technique for fast bilateral filtering of vector-valued multiple images including color, flow field, moving pictures, and volumetric. Weights in the bilateral filter were approximately decomposed into orthogonal factors. This decomposition enabled us to execute high-dimensional integration sequentially in 1-dimensional spaces. Acceleration rates in our method grew with the dimensionality. We experimentally verified that the approximation error caused by the decomposition of the weights in the bilateral filter stayed sufficiently small for color and volumetric images.
    View full abstract
    Download PDF (3580K)
    • |<
    • <
    • 1
    • >
    • >|
feedback
 Top