ITE Technical Report
Online ISSN : 2424-1970
Print ISSN : 1342-6893
ISSN-L : 1342-6893
35.52
Showing 1-17 articles out of 17 articles from the selected issue
  • Type: Cover
    Pages Cover1-
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (12K)
  • Type: Index
    Pages Toc1-
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (47K)
  • Tomomi Yamaguchi, Ryosuke Oda, Gou Koutaki, Keiichi Uchimura, Takami S ...
    Type: Article
    Session ID: ME2011-131
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    We propose a novel face recognition system against the various expressions. Our proposal method needs only one neutral face image and produced various expression face images which are registered in the training database from one of them. The results show that four expressions using parameters obtained can be made and we performed an experiment to show the effectiveness of the proposal method.
    Download PDF (619K)
  • Daishi WATABE, Hideyasu SAI, Katsuhiro SAKAI, Osamu NAKAMURA
    Type: Article
    Session ID: ME2011-132
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    An algorithm that improves the robustness of an ear biometric system is proposed and targeted for use in surveillance systems based on ear biometrics. To deal with pose variation when ears are rotated in depth, Gabor jets of different poses are estimated and used as training data for a discriminant-analysis-based classifier. Experimental evaluations show the effectiveness of the proposed algorithm and the possibility of improving the robustness of a single-view-based ear surveillance-system.
    Download PDF (654K)
  • Kouhei MATSUZAKI, Terumasa AOKI
    Type: Article
    Session ID: ME2011-133
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In this paper, we propose a method of estimating illumination distribution from a single real image including a human face. It is necessary to acquire optics information of the real space in order to blend a virtual object and a real image without the sense of incongruity. Some methods to estimate light source environment by real image have already been suggested, but it is difficult to use them for a general image because there are a lot of strict limitations in them such as three-dimensional shapes in an image must be known beforehand. Therefore, in this paper, we propose a method to estimate light source by making the face of the person 3D model for general image without any strict limitations. We extract a face domain and a hair domain from a human image, and estimate light source environment that paid its attention to each surface reflection properties together. Finally, we evaluate the performance of our proposal for human images.
    Download PDF (593K)
  • Ruka KOBAYASHI, Mitsugu KAKUTA, Tadao HAYASHI
    Type: Article
    Session ID: ME2011-134
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Much more progress of the game in a sports is expected in the future. Moreover, practical tactics are important. Motion analysis and game analysis are conducted for the game sports. In game analysis, information extraction is performed from the image under game. And analysis based on the extracted information is performed by sports science. We have studied the method of analyzing for game sports analysis. In this paper, the tactics of the lacrosse which is one of the field sports is described. When analyzing a field sports, a player's position information is important. In lacrosse, the point which can build an offensive starting point with the space of goal back is one of the features of this sports. The analyzing method proposed by this research can be performed by the premise that a player's position information can be extracted from the image under game. In this paper, the result built theoretically about the attack side player's optimal positioning is shown. Player's movement with 2π/3 based on this theory can be practiced, therefore the image analysis software in which an algorithm like this research is mounted will be developed.
    Download PDF (521K)
  • Daisuke MATSUMOTO, Terumasa AOKI
    Type: Article
    Session ID: ME2011-135
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In this paper, we propose a new objectionable content detection method which uses "the rate of a pixel with the same luminosity value" as new image features and the filtering system which combined two or more features. In order to describe these features, we calculate the histogram of pixel values and judges whether an animation video is pornographic or not by performing the threshold processing. We show that our proposal is a very powerful for pornographic animation detection and a filtering system based on some image features can detect pornographic animation more sufficiently, by experimental results.
    Download PDF (675K)
  • Wataru KISHIMOTO, Kazuhito MURAKAMI
    Type: Article
    Session ID: ME2011-136
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    The prior researches have been reported that the region which has high saliency of hue/saturation/brightness can be judged subjectively noticeable in many case. In this paper, in order to clarify which component of hue/saturation/brightness is principal, the correlations between the gaze distribution and saliency of hue/saturation/brightness were examined. As a result, it was found that the gaze distribution tends to be highly correlated with the saturation saliency images. Based on this results, this paper proposes a method to attract human attention by changing its design.
    Download PDF (934K)
  • Takashi SAKAMOTO, Kazuya KODAMA, Takayuki HAMAMOTO
    Type: Article
    Session ID: ME2011-137
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    We previously proposed a method of reconstructing Light-Field from multi-focus images by dimension reduction and a 2-D filtering. On the other hand, it is easy to synthesize multi-focus images from dense Light-Field. Therefore, we can achieve conversion between 4-D Light-Field and 3-D multi-focus images without significant degradation. Based on the conversion, we study Light-Field compression via synthesized multi-focus images. By setting blurring filters used for synthesizing multi-focus images from Light-Field appropriately, about 35dB is obtained at 0.03bpp. In this paper, we show experimental results by using real images and degradation of reconstructed images is analyzed for improvement of compression quality.
    Download PDF (861K)
  • Yuya Iida, Kan Okubo, Norio Tagawa
    Type: Article
    Session ID: ME2011-138
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    The small vibration of the eye ball, which occurs when we fix our gaze on an object, is called "fixational eye movement." It has been reported that this vibration may work not only as a fundamental function to preserve photosensitivity but also as a clue to image analysis, for example contrast enhancement and edge detection. This mechanism can be interpreted as an instance of stochastic resonance, which is inspired by biology, more specifically by neuron dynamics. Moreover, researches for a depth recovery method using camera motions based on an analogy of fixational eye movement are in progress. In this study, using camera motions especially corresponding to the smallest type of fixational eye movement called "tremor." We have constructed the algorithms which are defined as a differential form, i.e. spatio-temporal derivatives of successive two images are analyzed. However, in these methods, observed noise of derivatives causes serious recovering error. Therefore, we newly examine a method in which a lot of images captured with the same camera motions are integrated and the observed local image blurring is analyzed for extracting depth information, and confirm its effectiveness.
    Download PDF (597K)
  • Hitoshi KIYA
    Type: Article
    Session ID: ME2011-139
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    In general, image signals are handled as discrete signals with finite length. In addition, the visibility of images is mainly controlled by the phase information. In this manuscript, image signals are considered from the standpoint of discrete signals with finite length, and a phase representation of images and its properties are described. Especially, some of the latest study results regarding the relationship between phase information and DCT-signs are presented. A DCT-sign based visibility control method, a DCT sign correlation and their applications are also shown as examples of the study results.
    Download PDF (103K)
  • Kazuki KUMAGAI, Terumasa AOKI
    Type: Article
    Session ID: ME2011-140
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    We study on a method of how to create 3DCG animation of aircrafts easily. Our target in this paper is an animation of air show from many kinds of aircraft animations. In the air show, there are two kinds of performances in general. One is group flight and the other is sole flight. We focus on the latter. We propose a method to generate flight course of those who aren't principal aircrafts automatically during solo flight. This method satisfies synchronization and smoothness of the curvature.
    Download PDF (522K)
  • Athudou Inokuma, Kazuhito Murakami
    Type: Article
    Session ID: ME2011-141
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Although there exists a method to generate high dynamic range image using multi-exposure images, the performance is not good if a moving object exists in an image. This paper proposes a method to generate high dynamic range image by detecting a moving object from multi-exposure images. This paper also discusses a method to evaluate it by sharpness factor.
    Download PDF (685K)
  • Takuya Yamamura, Masayuki Tanabe, Kan Okubo, Norio Tagawa
    Type: Article
    Session ID: ME2011-142
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Tissue Harmonic Imaging (THI) is used for obtaining high spatial resolution medical ultrasound images. THI has a high range resolution because of its wide bandwidth compared with fundamental imaging. Additionally an azimuth resolution of THI is also high. However, amplitude of harmonic components is smaller than that of a fundamental component. Additionally, frequency dependent attenuation is severe especially for harmonic components. These phenomena mean that THI is susceptible to noise. In this study, using a fundamental component and intensity characteristics of harmonic components with respect to the distance from a sound source, we propose a method for improving the SNR of THI with keeping its high-resolution nature. We show the effectiveness of the proposed method through simulations.
    Download PDF (579K)
  • Type: Appendix
    Pages App1-
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (85K)
  • Type: Appendix
    Pages App2-
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (85K)
  • Type: Appendix
    Pages App3-
    Published: December 05, 2011
    Released: September 21, 2017
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (85K)
feedback
Top