All issues
Predecessor
24.74
Displaying 1-16 of 16 articles from this issue
- |<
- <
- 1
- >
- >|
-
Article type: Cover
Pages Cover1-
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSDownload PDF (18K) -
Article type: Index
Pages Toc1-
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSDownload PDF (85K) -
Kohei UTSUNOMIYA, Enko CHO, Yasushi IWAMOTO, Hiroaki NISHINO, Kouichi ...Article type: Article
Session ID: BCS2000-42/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSMost range finders usually use laser to get the 3D shape of real objects without contacting with them. Such range finders project the laser on the surface of the real objects, acquiring the 3D shape by analyzing the reflected light from the object. However, the entire object data cannot be gotten by a single scan due to some reasons such as occlusion and diffused reflection on the surface. The lacked parts need to be covered to get the entire object data by registering multiple range data. In this paper, we propose an approach to registering multiple range data. This method employs two step registration procedures to solve some problems such as increase in calculation time and inferior resolution data issued in traditional methods.View full abstractDownload PDF (1071K) -
Tetsuya KINEBUCHI, Miroru MORI, Akira SUZUKI, Akio SHIO, Sakuichi OHTS ...Article type: Article
Session ID: BCS2000-43/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSThis paper introduces a method for eliminating the inclination of faxed images. The angle of inclination is calculated by extracting the location of two specific marks(black rectangles)per page. The specific marks are distinguished from similar marks by modified generalized Hough transform that uses the number of white pixels surrounding black pixels and by measuring the relative position of other marks. An experiment using the proposed method finds that it can eliminate the inclination of faxed images up to the angle of 0.25°View full abstractDownload PDF (843K) -
Hiroyuki MURATA, Hideto NISHIKADO, Hironori YAMAUCHIArticle type: Article
Session ID: BCS2000-44/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSA new restoration method for blurred image has been proposed and this genetic algorithms(GA) oriented method has got the better result compared with conventional reverse filter methods. Many kinds of techniques to find its two-dimensional reverse filters have been studied for the blurred images that their deterioration processes can be presented by the convolution with the original image and the paired point spread function(PSF). In this paper, a new method to find this inverse filter directly by using GA are proposed and experimentally studied. Two sophisticated ideas; a proposal for huge answer space problem caused by the real number coefficient, and three kinds evaluation function for high efficient convergence, made the GA practical. This method has its possibility to repair many sorts of images of which blurred process is unknown.View full abstractDownload PDF (1171K) -
Kohei INOUE, Keisuke SEKO, Kiichi URAHAMAArticle type: Article
Session ID: BCS2000-45/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSA method is presented for enhancement of contrast of color images. The dynamic range of luminance is compressed by utilizing multiscale smoothing of the original image by a nonlinear smoothing filter. Each component of RGB is varied with histogram equalization preserving their ratio to the luminance. At first, the logarithm of the luminance is calculated at each pixel and it is processed by the mode filter with triple scales. Obtained three images with different smoothness are next combined linearly and expanded exponentially to produce the luminance with enhanced contrast. At last, the ratio of obtained luminance to its original value is multiplied to each component of RGB and the range of resulted RGB is expanded to[0, 255]by the histogram equalization. Nevertheless this method contains nonlinear level transformation, the ratio of each component of RGB to the luminance is preserved.View full abstractDownload PDF (1993K) -
Naohiro OZAKI, Yankan WANG, Makoto FUJIMURA, Hideo KURODAArticle type: Article
Session ID: BCS2000-46/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSWe propose the noise elimination method that uses the characteristics of human vision. This method increases the image quality with simple process. At first, a noisy image is transformed to some bands, and the edges are detected in LL-band of image. And the noise in image is eliminated by use of the location of edge. The experiment shows that the proposed method have the good MOS(Mean Opinion Score) in the case of the image with only sharp edge.View full abstractDownload PDF (1393K) -
Takashi MATSUO, Makoto FUJIMURA, Yankan Wang, Hideo KURODAArticle type: Article
Session ID: BCS2000-47/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSThe correspondence point detection in the stereo image is an indispensable thing in 3 dimension coordinate calculation. In this report, authors proposed the correspondence point detection method of the stereo image that is able to detect many correspondence points independent a lighting condition. Proposed method uses the color degree as the characteristic quantity of the correspondence point. Also, proposal method detectes the correspondence point in the image unit for the increase of the number of the correspondence point. The experiment shows that many correspondence points are detected.View full abstractDownload PDF (923K) -
Tainxing LIU, Hideo KURODA, Makoto FUJIMURA, Yankang WANGArticle type: Article
Session ID: BCS2000-48/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSIn this paper we put forward an image compression algorithm partitioning an image into submatrices, and dynamically stratifying submatrices into FD-submatrices. According to our compression logic, we generate the compression tree; First, we make use of the decision theory of linear programming to define the initial parameters; and we classify the submatrices by using these parameters. Second, each of the submatrices is stratified to generate the FD-submatrix and the transitional submatrix. FD-submatrix can be actually expressed by the average of all the elements in it and the position submatrix having 0 and 1 as its elements. This means that we attain the aim of the image compression. The feasibility of our algorithm has been tested and verified by the experimental results.View full abstractDownload PDF (796K) -
Hirofumi MORI, Ichiro MATSUDA, Susumu ITOHArticle type: Article
Session ID: BCS2000-49/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSThis paper proposes a new lossless coding scheme for color still images using adaptive color transformation and minimum-rate predictors. First, RGB signals are converted into three components utilizing a lossless color transformation technique of the ladder network type, and these components are encoded separately by the same coding method. In order to improve coding efficiency, the scheme partitions a color image into square blocks composed of 8×8 pels, and classifies these blocks into 16 classes. Then not only the linear predictors but coefficients of the color transformation are optimized individually in each class so that the total coding rate of the three components can have a minimum. Simulation results indicate that the proposed coding scheme is superior to the conventional scheme, which utilizes the color transformation used in JPEG2000, for many kinds of color images in terms of a coding rate.View full abstractDownload PDF (629K) -
Yoshimitsu KUROKI, Yoshifumi UESHIGE, Teiji OHTAArticle type: Article
Session ID: BCS2000-50/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSThe predictive coding techniques employ the linear predictors in general. The predictor coefficients are obtained by solving the normal equations in minimization of the total squared prediction residuals. This paper addresses the performance of the simple predictors as follows:(1)The predicted values are computed by weighted summation of the neighboring three pixels, which are identical to the pixels used in JPEG and JPEG-LS. (2)The predicted values are obtained with shift operation, addition, and/or subtraction. (3)The predictor coefficients sum up to one. The comparison of the total squared residuals between those simple predictors and the predictors obtained by solving normal equations are discussed in a theoretical approach. The estimations for actual images are also described.View full abstractDownload PDF (846K) -
Yoshifumi UESHIGE, Yoshimitsu KUROKI, Teiji OHTAArticle type: Article
Session ID: BCS2000-51/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSIn this paper, we describe a method of video image segmentation using fractal coding. We apply 3D-IFS which includes overlapped range blocks. By increasing the number of range blocks, we obtain more accurate region segmentation. Inverse process of 3D-IFS is iteratively applied to coordinates of pixels. Characteristics of the orbits are applied. (i)The number of calculation of the orbits depends on their initial points. (ii)Some orbits converge into periodic sequences of the points. The simulation results show segmentation results are improved in low contrast boundary and so on.View full abstractDownload PDF (772K) -
Isao NAGAYOSHI, Tsuyoshi HANAMURA, Hiroyuki KASAI, Hideyoshi TOMINAGAArticle type: Article
Session ID: BCS2000-52/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSIn this paper, we focus on the MPEG-2 transcoder to provide the flexible video transmission by converting the compressed video bit streams, and study on the new coding scheme for realizing bit stream scalability by preserving and re-using the differenatial data generated by transcoding process. First, we show problems in the scalability by encoding and decoding process, and propose the scalable coding method by separating and composing of MPEG-2 bit streams. Especially, we aim at the improvement of coding efficiency, we consider the separating and composing algorithm. Next, we point that it is possible to estimate the changing of DCT coefficient block that generated by re-quantization process, and show the coding algorithm of this changing data by classificating coefficient value and re-constructing method of original DCT block. Last, by simulation experiments, we show that the proposed method prevents the increase of total bit-count by separating process, moreover it brings re-compressing effect. And by cascading the bit stream separating processor, it can easily realize the multi layer scalabirity, we show the effectivensss of proposed scalable coding method.View full abstractDownload PDF (1060K) -
Takamasa YOKOYAMA, Yukihiro BANDOU, Sei-ichiro KAMATAArticle type: Article
Session ID: BCS2000-53/MIP2000-1
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSIn video processing, compression techniques are required to improve the efficiency of transmission and storage of video data. Data compression methods can be classified in two basically different categories: lossy compression method and lossless compression method. Lossless compression method is demanded in medical image processing and materials of broadcast. We investigate a novel video lossless compression method using the Hilbert curve. We map the three-dimensional data to one-dimensional data along the Hilbert curve and then apply adaptive linear prediction coding to the one-dimensional data.View full abstractDownload PDF (860K) -
Article type: Appendix
Pages App1-
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSDownload PDF (89K) -
Article type: Cover
Pages Cover2-
Published: November 22, 2000
Released on J-STAGE: June 23, 2017
CONFERENCE PROCEEDINGS FREE ACCESSDownload PDF (65K)
- |<
- <
- 1
- >
- >|