Reports of the Technical Conference of the Institute of Image Electronics Engineers of Japan
Reports of the 258th Technical Conference of the Institute of Image Electronics Engineers of Japan
Showing 1-12 articles out of 12 articles from the selected issue
Date: October 28, 2011 Location: TOYAMA Prefectural University
Reports of the 258th Technical Conference of the Institute of Image Electronics Engineers of Japan
  • Yuya WATANABE, Hidekazu TSUJIAI
    Session ID: 11-02-0
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    A laser range sensor combines the range finding and rotation by laser. Three-dimensional measurement can be carried out by the system which added the equipment which can carry out yawing to this sensor. We verify whether the structure of a liquid can be caught by this system.
    Download PDF (291K)
  • Yanli LIN, Kouki ITOH, Tsukasa KIKUCHI
    Session ID: 11-02-02
    Published: 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    This paper proposes a simulation method that can enable the user to control the movements of China ink being dispersed in water. In the proposed method, the fluid calculations for China ink dispersing in water were made into a model with a grid base using a 2-dimensional incompressible fluid equation, then it was expanded into a 3-dimensional fluid field using the Kolmogorov spectrum. Furthermore, when an animator wishes to express desired shapes using the diffusion phenomenon of China ink, he/she prepares beforehand the shapes for which he/she wants to control the diffusion phenomenon (a process called “targeting”) and defines them in the interior of the 3-dimensional fluid field. In the 3-dimensional fluid field where the target shape is defined, a particle equivalent in size to one voxel is moved into the target shape according to the fluid body field, and a decision is made about collision between the particle and the target shape. In other words, the 3-dimensional fluid field is renewed according to the movement path of the particle. This enables the formation of new 3-dimensional fluid fields inside the target shape and allows the animator to express the diffusion phenomenon of China ink inside the desired shape.
    Download PDF (901K)
  • Takatoshi SOGA, Kunio OHNO
    Session ID: 11-02-03
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    This paper describes a trial methodology to analyze the business process of a small company by means of UML and to apply the result to an ERP system. Though ERP systems have been introduced to big business companies since 1990s, small businesses could not have introduced them because of their heavy specification and high price. Recently an open source ERP system for small businesses has been developed, and some venture companies which intend to make global businesses. However, ERP system requires various viewpoint of the businesses, compared with existing financial software packages, so modeling of business process will be necessary to introduce ERPs. Then business process modeling by UML is tried, and mapped to the open source ERP system.
    Download PDF (304K)
  • Ikuro UENO, Fumitaka ONO
    Session ID: 11-02-04
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    We have proposed a fast and simple binary arithmetic coder, in which renormalization operations can be done without storing both of the bottom address of the valid interval and renormalized interval, by executing flush operation at the same time. In the flush process, the MPS/LPS interval is carefully divided into the combination of current and succeeding symbols if necessary, to prevent the loss of the coding efficiency. In this report, we will further improve the coding efficiency, allowing the omission of flushing, and shows that it can be implemented by a state transition table. We will also examine the performance of the system having larger register size that improves the efficiency for sources having small LPS probabilities.
    Download PDF (336K)
  • Kenta TOBISAWA, Terumasa AOKI
    Session ID: 11-02-05
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    We are usually watching a movie while predicting the next scene. In this paper, we observe this phenomenon in detail and propose a new model of these characteristics. After that we prove this hypothesis is correct by experiment. The aim of our model is to clarify the characteristics of human prediction and to estimate human gaze motion automatically by detecting the object which is the most important in current scene, the object which will be the most important in the next scene, and the event information between them.
    Download PDF (181K)
  • Ichiro UMATA
    Session ID: 11-02-06
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    This paper reports an attempt to characterize interaction strategies based on gaze and non-verbal behavior incollaboration. Collaborators’ actions are led by their interaction strategies and characterizing them helps identify their actions.Collaboration analyses based on gaze and motion capture data are discussed.
    Download PDF (3040K)
  • Daisuke MATSUMOTO, Terumasa AOKI
    Session ID: 11-02-07
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    In this paper, we propose an adult animation detection method which uses "the appearance cycle of the similar frames" as new image features. In order to describe these features, we calculate the local average pixel values in each frames and record time variation of the value. We show that our proposal is a very powerful method for adult animation detection and independent of skin color information which are popular features, by experimental results.
    Download PDF (281K)
  • Masahiko FUJIWARA, Terumasa AOKI
    Session ID: 11-02-08
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    "Whether we can eat this or not" is one of the most important judgments for animals to maintain their lives. Human beings must also judge "whether this is delicious or not" in the same way. In this paper, we will propose a new hypothesis that human beings possess a "reflection characteristic database for various kinds of foods". Using this database, we will compare a current object with data in our brain and judge "whether we can eat it or not". As a result of eye-tracking experiment, we will prove this hypothesis is correct by experiment.
    Download PDF (473K)
  • Shouhei MATSUMOTO, Hirotaka TANAKA, Hiroshi OCHI
    Session ID: 11-02-09
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    The error diffusion method creates high-quality dithered pictures. It can be expanded to multilevel dither processing. However, the multilevel error diffusion method causes picture quality deterioration around middle quantization level. Some studies are reported to solve this problem. However, the previous methods have complicated algorithm and may not be most suitable method. This paper proposes a error diffusion method which resolve the problem by the restricting quantization errors to small value. It also shows that high quality picture of multilevel error diffusion method can be obtained.
    Download PDF (306K)
  • Toshiya SUZUKI
    Session ID: 11-02-10
    Published: 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    Conventional outline fonts are designed to store a self-standing routine for a specified glyph, because they are designed for small character sets without duplicated glyphic components, like Latin, Greek and Cyrillic scripts. The ratio of the numbers of the glyphs and glyphic components are almost comparable. But in the case of CJK Ideographs, the number of the glyphic components is remarkably smaller than the size of the character set. Thus, modern outline fonts provide the composite glyph mechanism which composes a glyph by referring the subroutines for the glyphic components. To reduce the number of the glyphic component subroutines, some fonts use TrueType hinting instruction to stretch or compress the components. Such fonts have no flexibilities for the applications of the TrueType hinting. In this report, the several methods how to detect such fonts are discussed and the benchmark of the proposed methods in the document rendering process is reported.
    Download PDF (742K)
  • Kenya HONMA, Hirokazu MADOKORO, Kazuhito SATO
    Session ID: 11-02-11
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    This paper presents a method for classification and recognition of behavior patterns based on interests from human trajectories at an event site. Our method creates models of Hidden Markov Models (HMMs) for each human trajectory quantized using one-dimensional Self- Organizing Maps (SOMs). Subsequently, we apply two-dimensional SOMs for unsupervised classification of behavior patterns from features of the distance between models. Furthermore, we use Unified distance Matrix (U-Matrix) for visualizing category boundaries based on the Euclidean distance between weights of SOMs. Our method can extract typical behavior patterns and peculiar behavior patterns of interests obtained using a questionnaire and visualize relationships between these patterns. We evaluated our method based on Cross Validation (CV) methods using trajectories of typical behavior patterns. The recognition accuracy reached to 83.3% for estimating interests of three types. We consider that our method is useful to estimate interests from behavior patterns at an event site.
    Download PDF (2083K)
  • Seiya ODAGIRI, Kazuhito SATO, Hirokazu MADOKORO
    Session ID: 11-02-12
    Published: October 28, 2011
    Released: December 20, 2016
    CONFERENCE PROCEEDINGS RESTRICTED ACCESS
    This paper presents a method for optimizing and sharing parameters of Level Set Methods (LSMs). We used Genetic Algorithm (GA) of evolutional learning for optimizing parameters of LSM as a deformable shape model. However, our former method required the information of Ground Truth (GT) for each target image. Therefore, the method was impossible to apply clinical images without GT. In this paper, we focus on a trade-off problem of sharing and separating parameters used for clinical images without GT. We applied our method to head Magnetic Resonance (MR) images of 52 patients from 30s through 70s classified into two groups under the instruction by a radiologist. For the result, the mean extraction accuracy is improved to 84.7% because we can share parameters of LSM in each group. However, iterations of contours are sensitive to the information of anatomical structures and luminance properties in each target image. We consider that the adjustment of iterations is required after the optimization.
    Download PDF (774K)
feedback
Top