ITE Technical Report
Online ISSN : 2424-1970
Print ISSN : 1342-6893
ISSN-L : 1342-6893
Search
Advanced search
OR
Browse
Search
33.23
Displaying 1-24 of 24 articles from this issue
Show all abstracts
    • |<
    • <
    • 1
    • >
    • >|
  • [title in Japanese]
    Article type: Cover
    Pages Cover1-
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_Cover1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (14K)
  • [title in Japanese]
    Article type: Index
    Pages Toc1-
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_Toc1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (81K)
  • Fixation Map and Number of Observers
    Shingo TANITA, Masahiro ISHII, Zheng TANG
    Article type: Article
    Session ID: IST2009-17/ME2009-93
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    Usability testing is widely used to determine the quality of visual materials such as web site designs. To visualize the eye tracker's data, gaze fixations are plotted in 2-D plane. It is necessary to employ techniques to extract characteristics for this is signal averaging of data obtained from many observers.This study investigated convergence of intersection between individuals against number of observers.
    View full abstract
    Download PDF (810K)
  • Remarkable area presumption using local extrema
    Yoshinori Takigawa, Masahiro Ishii
    Article type: Article
    Session ID: IST2009-18/ME2009-94
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_5
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    The object that stands out might seem to come up immediately jumping into eyes. Such a phenomenon is called a pop out. Saliency Map to which a remarkable area is excellently presumed by various images uses Laplacian of Gaussian. Then, it proposes the technique for presuming a remarkable area by using the extremal point of the image in the present study. Moreover, the discrimination ability of the color falls not only eyesight deteriorates by the perimetric view that is considered when I move eyes next. Then, the influence that color information gives to the perceptual salience of the image area is examined at the same time. I show an image to a person and measure eyes and calculate an actual gaze area. The degree of similarity of the remarkable area that had been presumed to be the actual gaze area was calculated. As a result, it was a result that the accuracy of the extreme value point map that added the color element in the test image was the best.
    View full abstract
    Download PDF (734K)
  • An Index of Evaluation of Effect of Jaggy Elimination by Pseudorandom Pixel Placement
    Yui MAEDA, Junichi AKITA
    Article type: Article
    Session ID: IST2009-19/ME2009-95
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_9
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    The pixels in the conventional image systems are located in lattice places, and the lattice placement of pixels causes the jaggy in the image, which often cause the severe defect in the image quality. The conventional approaches to overcome this jaggy problem are such as the increased number of pixels and anti-aliasing, however, the number of pixels is limited by the physical pixel size and the quantity of the image information, while the anti-aliasing intrinsically causes the blurred image. The authors have been proposing and evaluating the image systems with pseudorandom pixel placement to decrease the effect of the jaggy. In this report, we discuss the comparison of the subjective evaluation index of the jaggies, which is the final goal of the image quality evaluation, with the objective evaluation index of the jaggies, which can be calculated from the image information. We also discuss the jaggy elimination effect by pseudorandom pixel placement by using these two evaluation indexes.
    View full abstract
    Download PDF (611K)
  • Oblivious Affine Compensation Using Three 2-D Barcodes
    Masashi SHIMIZU, Masaaki FUJIYOSHI, Hitoshi KIYA
    Article type: Article
    Session ID: IST2009-20/ME2009-96
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_13
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    This paper proposes an image registration method using 2-D barcodes for images digitized by scanners. The proposed method never requires the original image corresponding to the image to be compensated, whereas correlation based methods require the original image. The proposed method puts three 2-D barcodes to an original image. 2-D barcodes encode information to be used in an image registration, such as, positions of barcodes, relations among barcodes, and so on. The image with 2-D barcodes is printed and redigitized by a scanner. By using the 2-D barcode that is robust to the geometric deformation, 2-D barcodes can be decoded. Therefore, information for image registration is obtained before compensation. The proposed method compensates an affine distortion of the image by using decoded geometric relations.
    View full abstract
    Download PDF (992K)
  • A Quality Layer Construction Given a Target PSNR for JPEG2000
    Atsumu TANAKA, Kiyoshi NISHIKAWA, Hitoshi KIYA
    Article type: Article
    Session ID: IST2009-21/ME2009-97
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_17
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We propose a one-pass rate control method given a target PSNR for JPEG2000 coded images. The target PSNR could be set at each layer when the processed image has layer structure. The proposed method is based on PCRD-opt algorithm, therefore computational costs for rate controling would not increase compared to those of the rate control method given a target bit-rate. Besides, bit-rate could be controllable under the condition that an upper limit of bit-rate and a target PSNR are simultaneously given. By applying the proposed method to a video sequence compression, we could prevent to generate low PSNR frames. Also, by employing the proposed method with a rate control method given a target bit-rate, it is possible to efficiently reduce the amount of coded data.
    View full abstract
    Download PDF (562K)
  • A projection system using a screen with low-reflection and high-transmission for displaying artificial shadows
    Koichi NAKAHASHI, Masahiro ISHII, Zheng TANG, Kazuya YAMASHITA
    Article type: Article
    Session ID: IST2009-22/ME2009-98
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_21
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We developed a system for displaying artificial shadows calculated by the computer. As beginning of our project, we intended to minimize the latency and to display only artificial shadows. However, we noticed the problem that the light reflects and generates real shadows on the device. To solve this problem, we constructed the device using a screen with low reflection and high transmission. When this system can interfere the colors and the shapes of shadows, it will be applicable for various approach such as media art and entertainment.
    View full abstract
    Download PDF (994K)
  • Integration time control and image synthesis by local brightness distribution for wide dynamic range imaging
    Shuichi FUJISAKI, Satoko KAGAMI, Takayuki HAMAMOTO
    Article type: Article
    Session ID: IST2009-23/ME2009-99
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_25
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We propose a method of wide dynamic range imaging by variable-integration-time imaging and reconstruction processing based on local brightness distribution of objects. In the proposed method, local contrast is improved by allocating each pixel's integration time when the contrast with surrounding pixels is maximum. Additionally, the reconstruction processing adapted to local brightness distribution realizes wide dynamic range imaging. Further, we construct a system to achieve the proposed method in real-time.
    View full abstract
    Download PDF (937K)
  • High speed and high accuracy motion vector detection by on-sensor pixel matching
    Yoshihiro KAWASHIMA, Kenichi NAKAYAMA, Takayuki HAMAMOTO, Kazuya KODAM ...
    Article type: Article
    Session ID: IST2009-24/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_29
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We propose a method of motion vector detection at high frame rate on a smart image sensor, and describe the design of the smart image sensor on which the method is implemented. In the proposed method, motion vector can be detected by single pixel matching with limited search range for pixels on the detected edge. Additionally, two continuous results of motion detection are compared to reduce effects of noise with controlling the processing interval. Then, the detection accuracy is much improved.
    View full abstract
    Download PDF (700K)
  • Contrast improvement of in-vivo imaging using a complementary-metal-oxide-semiconductor image sensor
    Kiyotaka Sasagawa, Ayato Tagawa, Hiroki Minami, Toshihiko Noda, Takash ...
    Article type: Article
    Session ID: IST2009-25/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_33
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    Fluorescent imaging by using a complementary-metal-oxide-semiconductor image sensor is expected to be useful for observation of freely moving animals. In the in-vivo imaging of a mouse brain with an implanted image sensor, high contrast images are obtained only at around the excitation light source because of its inhomogeneous illumination. In this paper, we report on improvement of imaging dynamic range and contrast based on multiple readout and the excitation light distribution image.
    View full abstract
    Download PDF (712K)
  • The development to a sensor of RF CMOS technology : Measurement System on Chip
    Akio KITAGAWA
    Article type: Article
    Session ID: IST2009-26/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_37
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    The RF/AMS-CMOS technology begins to have an impact on the leading-edge sensor technology. The trend of CMOS technology is surveyed in terms of the novel sensor and the measurement system on chip. For example, the free radical sensor, nano-displacement sensor, excretion-monitoring sensor is reported.
    View full abstract
    Download PDF (1875K)
  • Super Resolution for Text Images
    Shunichi KIMURA, Ikken SO, Masanori SEKINO, Satoshi KUBOTA, Kanya ISHI ...
    Article type: Article
    Session ID: IST2009-27/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_45
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    This paper addresses the problem of recovering a super-resolved image from (an) office document image(s) that is/are obtained through document scanner and has/have several texts. We propose a method that enables super-resolution even if we get only one input image. Firstly, text images are extracted from (an) input image(s). Secondly, these text images are classified such that one class has same text images. Finally, super-resolution algorithm is adopted for each class to create a high-resolution text image. Simulation results show that this method can restore sufficient text quality. Also, we exploit a new image-registration algorithm using quadric hyper-surfaces and a fast super-resolution algorithm especially for text images.
    View full abstract
    Download PDF (748K)
  • Video Mosaicing using High-speed Vision
    Keitaro YAMAMOTO, Takashi KOMURO, Masatoshi ISHIKAWA
    Article type: Article
    Session ID: IST2009-28/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_49
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    In this paper, we propose a new kind of image mosaicing system using high-speed vision. We constructed the system that generated a smooth mosaic image by using the least distorted image which is nearest to its center and by simplifying the algorithm using the property of high-speed vision.
    View full abstract
    Download PDF (1260K)
  • Forest visualization using tree survey data and rectification of fish-eye photos
    Masayuki KITAHAMA, Mamoru KUBO, Ken-ichiro MURAMOTO
    Article type: Article
    Session ID: IST2009-29/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_53
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    The objective of this study is to describe three-dimensional forest geographical information system by creating a fusion of tree survey data, satellite image and omni-directional fish-eye photos. First, we describe three-dimensional forest space which is obtained from making a tree object from tree surveying data and mapping a satellite image onto its tree crown part. In certain viewpoint of three-dimensional forest space, the panoramic image consisting of 24 pieces projected planes are obtained from turning eyesight whose angle of elevation is 60 degrees and horizontal viewing angle is 15 degrees. Second, we convert a range of elevation angle 60 degrees of omni-directional fish-eye photos which is photographed at a same spot into panoramic image, and overlap the panorama image which made from the three-dimensional forest space.
    View full abstract
    Download PDF (1055K)
  • Image measurement and snow particle classification in snowfall monitoring
    Kazuki SETO, Mamoru KUBO, Ken-ichiro MURAMOTO, Yasushi FUJIYOSHI
    Article type: Article
    Session ID: IST2009-30/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_57
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    The snowfall is the weather phenomenon that shape and kind of the particles are not constant, and there are various snow particles. The classification of snow particle into snowflake and graupel is important for understanding of snow particle formation. Although qualitative discussion has been made of the formation processes based on the data of surface observation, there still remain various problems unsolved. This mainly comes from a lack of quantitative and continuous observation of snow particles, and method for classification into snowflakes and graupels is not determined. To monitor the snowfall phenomenon, snow particles are classified into the snowflake and graupel by the image measurement. Natural falling snow particles were photographed with the digital camera automatically. Particle image feature values were extracted and particles were classified into snowflakes and graupels by discriminant analysis, and the ratio of the two types was calculated at time intervals of one minute.
    View full abstract
    Download PDF (719K)
  • A space- and wavelength-division-multiplexing free-space optical communication system using a CMOS image sensor with multi-point optical signal acquisition
    Keiichiro KAGAWA, Jun TANIDA
    Article type: Article
    Session ID: IST2009-31/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_61
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We are developing an intuitive and fast indoor free-space optical communication system in which a dedicated CMOS image sensor with functions of ordinary image capturing and concurrent multi-point optical data acquisition to offer space- and wavelength-division-multiplexing. As its components, prototypes of the dedicated CMOS image sensor and a wide-angle beam steering optics for uplink have been developed. We have confirmed beam deflection, and concurrent optical data acquisition at data rate of 10-Mbps/ch from two nodes at the hub.
    View full abstract
    Download PDF (783K)
  • In-air Typing Interface for Mobile Devices using a Single Camera
    Yuki Hirobe, Kazunori Funahashi, Takehiro Niikura, Yoshihiro Watanabe, ...
    Article type: Article
    Session ID: IST2009-32/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_65
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We propose vision-based interface for mobile devices that supports text entry by tracking 3D motion of a human fingertip using a single camera and by recognizing small keystroke gesture in the air. Since the fingertip near the camera moves fast in the image, a high-frame-rate camera is used for stable tracking. In order to separate keystroke gesture from the others, frequency filter is applied to scale change in the fingertip images. We evaluated the accuracy of estimated fingertip position, detection performance of keystroke action and actual typing usability of the proposed interface.
    View full abstract
    Download PDF (870K)
  • Acceleration of Closing for Use in a Make-up Simulation
    Ikumi MAIHARA, Akira TSUKADA
    Article type: Article
    Session ID: IST2009-33/ME2009-10
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_69
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We are researching and developing a real-time make-up simulator that helps accelerate and optimize selection for cosmetics buyers. Our task was to improve the speed and smoothness of the simulator. For this, our report focuses on foundation. Virtual foundation is applied to a still image of the user resulting in a mask image which is then extracted and passed through a blur filter and then blended with the original. Using Spiral Labeling when extracting the mask image allows the Closing process to take place within the boundaries of the relevant facial area resulting in a speeding up of the imaging process.
    View full abstract
    Download PDF (612K)
  • Cross-media retrieval system based on association between heterogeneous media
    Masaki ISHIHARA, Takayuki BABA, Susumu ENDO, Shuichi SHIITANI, Yusuke ...
    Article type: Article
    Session ID: IST2009-34/ME2009-11
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_73
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    Among the current information retrieval systems, generally the search query and the retrieval target are the same media type, such as using keyword query for retrieving documents, or using image query for retrieving similar images. In order to search and collect desired information between different kinds of media, it is necessary to retrieve across various media types, without being constrained by the media type of the search query and the retrieval target. The authors have developed a prototype system that is able to achieve cross-media retrieval by associating between different types of media via multimedia which composed text, image, or sound.
    View full abstract
    Download PDF (734K)
  • Dynamics Analysis of Facial Expression Changes for Person Identification
    Hidenori TANAKA, Hideo SAITO
    Article type: Article
    Session ID: IST2009-35/ME2009-11
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_77
    CONFERENCE PROCEEDINGS FREE ACCESS
    Show abstractHide abstract
    We propose a new method for analyzing dynamics of facial expressions to identify persons. In the proposed method, facial feature points are extracted using an Active Appearance Model in the first frame of continuous image sequences. They are then tracked using Lucas-Kanade based method. Next, an interval from the beginning time to the ending time of facial expression changes is extracted. Finally, a feature vector is obtained. In the identification phase, an input vector is classified by calculating the distance between an input vector and the reference vectors using DP matching. We show the effectiveness of the proposed method using smile image sequences from MMI Facial Expression Database.
    View full abstract
    Download PDF (816K)
  • [title in Japanese]
    Article type: Appendix
    Pages App1-
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_App1
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (82K)
  • [title in Japanese]
    Article type: Appendix
    Pages App2-
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_App2
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (82K)
  • [title in Japanese]
    Article type: Appendix
    Pages App3-
    Published: June 15, 2009
    Released on J-STAGE: September 20, 2017
    DOIhttps://doi.org/10.11485/itetr.33.23.0_App3
    CONFERENCE PROCEEDINGS FREE ACCESS
    Download PDF (82K)
    • |<
    • <
    • 1
    • >
    • >|
feedback
 Top