-
Nobuhito Matsushiro
Session ID: 03-02-01
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
From a new view point, primary considerations related to chromatic adaptation is performed. In the long history of chromatic adaptation, almost all models have been constructed based on psychometric testing. In this paper, a hypothetical principle of chromatics adaptation is defined as the minimized state of the metameric black component of an image and is defined as the enhancement of the efficiency of the perception. There have been no discussion related to chromatic adaptation from the view point of metameric black control.
View full abstract
-
Koji Takeuchi, Takahiro Nakayama, Masahiro Konda, Koji Kotani, Tadahir ...
Session ID: 03-02-02
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Conventional image compression techniques based on transform coding to frequency domain have a problem of image quality, when images contain sharp edges like character patterns are compresed. In order to solve this problem, we have developed Adaptive Resolution Vector Quantization (AR-VQ) method and a systematic codebook design method applied to all kinds of images without using learning sequences for 4x4 and 2x2 pixel blocks. By using these methods, consequently, we can realize much superior compression performance than the JPEG and the JPEG2000. On the compression of the XGA (1024x768 pixels) images including text, for instance, there exist an overwhelming performance difference of 5 to 40 dB in compressed image quality.
View full abstract
-
Hiroshi Sata, Yoshitaka Maekawa, Mitsuhiro Iyoda
Session ID: 03-02-03
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Vector quantization(VQ) is one of the image compression technologies. VQ has the feature that mosquito-noise does not arise, but it has a problem in compression in a low frequency domain. In order to improve this problem, the generalized harmonic analysis(GHA) codes a low frequency component. The GHA can analyze a signal in a low frequency domain with a good performance. Therefore, this paper proposes the compression scheme of using the GHA in combination with VQ for improving image quality, and estimates the proposed scheme. The proposed scheme codes a low frequency component by the GHA, and applies VQ to the region with the GHA's error more than a threshold.
View full abstract
-
Ikuko Uwano, Masashi Kameda, Koichi Matsuda, Akio Doi, Takashi Inoue, ...
Session ID: 03-02-04
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
It is required for the computer aided diagnosis that 3-dimentional brain area is extracted accurately from head MRI images. The Region Growing method, which has been used in our method, has the following problems; i) a small leakage in ROI (Region of Interest) expands into other region, and ii) an isolated region on a slice image is not able to be extracted in spite of such region is connected 3-dimansionally. Therefore, we propose an efficient extraction method to prevent leakage by the edge detection, and to extract isolated region by two-way Region Growing with restricted processing considering the correlation coefficient between each slice image. As a result, our proposed method has been able to improve the accuracy of extraction compared with our previous method.
View full abstract
-
Kunihiko TAKANO, Ryoji WAKABAYASHI, Yoshiaki OKAMURA , Daisuke NOMURA, ...
Session ID: 03-02-05
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
Here is an image transmitting technique using SSTV system which enables us to make band width within 3.0 kHz by transforming image data into audio signal. In this paper, the SSTV technique is applied to a transmission of three dimensional images, using the hologram which enables us to record in three dimensional space as fringe patterns. As a result, it is found that the excellent three dimensional images are reconstructed from the transmitted data of the hologram adopting SSTV system.
View full abstract
-
A Study of Interactive Motin Generation System
Suma Noji, Hiroki Takahashi, Masayuki Nakajima
Session ID: 03-02-06
Published: May 27, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
This paper writes about a development of an interactive circus system. The system is based on dynamics and developed for edutainment. Acrobatic motion simulator and editing tool are implemented. The system enables users to learn basic physical dynamics easily. And also enables creative character animation work. The system provides recording function for motion editing. Users can save the simulation data and review the animation operating real time camera work.
View full abstract
-
AZAM Bastanfard, HIROKI TAKAHASHI, MASAYUKI NAKAJIMA
Session ID: 03-02-07
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
The realistic modeling and rendering of fabrics is a challenging task in computer graphics. The application of cloth modeling and rendering are found in computer aided design, Multimedia and fashion. This paper discusses the importance of fabric buckling appearance and it presents an image-based method to generate fabric buckling by using Gabor wavelets. Since modeling the geometric details is prohibitive in terms of memory requirement and rendering time, Gabor wavelets proves to be a very useful texture analysis that have the benefits of Fourier and locality. In this paper Gabor filters were configured as a multi resolution filter banks for texture analysis and these filter banks will be synthesized to extract fabric surfaces. The validation of the proposed approach is given by discussion and illustration of experimental results.
View full abstract
-
Hamid LAGA, Hiroki TAKAHASHI, Masayuki NAKAJIMA
Session ID: 03-02-08
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
We present an efficient algorithm for complex 3D object reconstruction from a set of unorganized points in 3D . This algorithm is based on Radial Basis Functions (RBFs) which provide an efficient compact model for the surface description. However, for large data sets, estimating the RBF parameters is memory and time consuming. In this paper we propose a new approach to estimate the RBF parameters for large data sets. Unlike previous approaches our algorithm avoids adding the inside and outside constraints required by the distance function. In a firrst step, our algorithm uses only the boundary constraints and solves for the derivative of the implicit function, so the set of RBF centers, thus the problem size is reduced to one third. This results in significant compression and computational advantages. Then, in a second step, the problem is further optimized by an efficient boundary constraints reduction. We will demonstrate through the experiments that this approach improves the fitting and evaluation times and the required storage.
View full abstract
-
David GAVILAN RUIZ, Hiroki TAKAHASHI , Masayuki NAKAJIMA
Session ID: 03-02-09
Published: July 25, 2003
Released on J-STAGE: August 29, 2003
CONFERENCE PROCEEDINGS
RESTRICTED ACCESS
This work generalizes the concept of blobs defined in the scale-space theory to contain color information. We also propose an image categorization algorithm based on those color blobs. Our approach relies on color quantization using basic color categories and the successive classification of the resulting regions using neural networks. We show that the preattentive information can be used to categorize even low resolution pictures taken with mobile devices. Thus, we propose a framework in a mobile environment where the classification and additional contextual information, such as GPS location, are used to assist user in such tasks as navigation or self-configuration of the device.
View full abstract
