-
Article type: Cover
Pages
Cover1-
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Article type: Index
Pages
Toc1-
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Hideto Motomura
Article type: Article
Session ID: IPU2003-1
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
An algorithm of multi-primary color separation, which can control display system colorimetrically, is proposed in this paper. The proposed method is available for not only RGB 3-primary display but also over-4-primary display. Colorimetric accuracy is evaluated with 6-primary DLP^<TM> projector. CIELAB color difference between target colors and reproduced colors is 0.8 on average and 2.9 at maximum.
View full abstract
-
M. Nanba, Y. Takiguchi, K. Osada, K. Miyakawa, S. Okazaki, T. Yamagish ...
Article type: Article
Session ID: IPU2003-2
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
To clarify the role of the mesh electrode in a field-emitter image sensor, two types of exoerimental sensors with/without the mesh electrode were fabricated and their characteristics comoared. The results agree with the known effect that a mesh electrode suppresses the electron beam soread that causes the resolution characteristics to deteriorate. They also revealed that 1)the sensor with the mesh electrode could readout the signal charges keeping the HARP'S surface Dotential low, and 2) the mesh electrode grounded with a large caDacitor reduced the noise mainly due to the horizontal driving Dulses of the cathode electrodes.
View full abstract
-
Toshihisa WATABE, Hiroshi OHTAKE, Masahide GOTO, Hirotaka MARUYAMA, Ke ...
Article type: Article
Session ID: IPU2003-3
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We have been studying a solid-state HARP image sensor in developing a highly sensitive solid-state imaging device. This device reauires a CMOS image sensor with a high signal-to-noise ratio (SNR) to make full use of the ootential of HARP film. We developed the device by using a charge-transfer circuit. However, row-to'row and pixel-to-pixel fixed pattern noise (FPN) appeared in images we took. We tried to solve these oroblems by applying a new ooeration method to the oixels for reading and resetting signals. Experimental results showed that the prototype's row-to-row and pixel-to-pixel FPN were improved by 40 and 2 1 dB, respectively, comoared with a conventional sensor that has a charge-transfer circuit.
View full abstract
-
Hiroshi YOSHIKAWA, Yasuhiro TAKAKI
Article type: Article
Session ID: IPU2003-5
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We have already orooosed a natural 3 D display which is based on high-density generation of directional images. In order to display real 3D scenes, the most promising 3D camera technique is the generation of high-density directional images from multiple images caotured by horizontally aligned multiple cameras by using an interpolation algorithm. In this study we examined four different simple interpolation algorithms. An algorithm that utilizes one reoresentative object distance as a oriori was found to be the most effective one. This technique enables a fast interpolation. There, however, is a tradeoff!jetween the number of cameras and the depth of3D objects. The allowable depth for 3D objects is renorted.
View full abstract
-
Takaaki KUDO, Yasuhiro TAKAKI
Article type: Article
Session ID: IPU2003-6
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
We have already constructed a natural 3D display which generates 64 directional images simultaneously. We are also developing a 3D interactive processor to enable the 3D display to display 3D moving images and to irodate 3D images interactively. lt is implemented in a computer array which consists of 64 PC's. ln this study we report the video-rate synchronization of the computer array by using the broadcast mechanism ofTCP/IP socket communications on the ethernet.
View full abstract
-
Satoshi KURIHARA, Tatsuya SAITO, Kenji KOHIYAMA
Article type: Article
Session ID: IPU2003-7
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper, we will propose a new textured 3D model scanning methodology based on IBM &R, and its implementation. Up to now, we have orooosed the scanning methodology, which was the initial version of it, used slit light sources and took photographs of a object like CT scan. By this methodology we could get high-resolution 3D model simplicity, but it had the following oroblem: the inherent color information of the object could not get by the slit light source. So, this time, we will propose a new methodology to enable to simultaneously use general light sources at the scanning ohase by the slit light sources.
View full abstract
-
Yasuhiko Watanabe, Toshiyuki Kurahashi, Yoshikatsu Kishida, Yukio Toku ...
Article type: Article
Session ID: IPU2003-8
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
This paper describes two new methods for improvement of accessibility of contents retrievals from distributed archives in network. One method uses user driven parameters such as access histories or previously defined conditions. The another method uses information driven parameters, e.g. links for related information for the retrieved contents. The trial system is realized to evaluate these methods using meta server for accessing multiple image databases. Experimental results show the improvement of accessibility to distributed digital archives through these methods.
View full abstract
-
Shunrou Fujiwara, Akio Doi, Koichi Matsuda, Masashi Kameda
Article type: Article
Session ID: IPU2003-9
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
In this paper, we propose a method to extract a 3D geometry model automatically from a 3D image data(volume data) by the simple model which has a curved surface such as cylinder or sohere. The geometry model which is used in our method is the deformable model based on a energy-minimizing method. By using the geometry model, the commitational cost is less than the orevious method because the shape is as slimilar as that of the resion in a volume data to extract. Consequently, our method is useful fbr extracting a 3D geometry model like the shape of cvlinder and that of sohere fom a volume data.
View full abstract
-
Article type: Appendix
Pages
App1-
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS
-
Article type: Appendix
Pages
App2-
Published: February 07, 2003
Released on J-STAGE: September 20, 2017
CONFERENCE PROCEEDINGS
FREE ACCESS