|
1. |
Extended‐order statistic filter and evaluation of noise reduction performance |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 1-11
Tatsuya Fujii,
Kaoru Arakawa,
Hiroshi Harashima,
Preview
|
PDF (969KB)
|
|
摘要:
AbstractThe authors investigated construction of an extended‐order statistic filter to remove non‐Gaussian white noise added to image signal. First, they devised a differential‐order statistic filter using signal‐differential ordering instead of a signal‐value ordering and estimated its noise reduction performance. Next, they proposed an extended‐order statistic filter which unifies the three‐order information of time, signal value, and signal differences obtained from a signal within the filter window.Preparing the 3‐dimensionalized filter coefficient space and using filter coefficients chosen sequentially in the three different orders among them allow effective noise reduction performance to be obtained. In addition, as a method for resolving a problem on the design method of filter coefficients, investigation was made using a synthesized training sequence. Extracting parameters necessary to a linear AR model from an objective signal, it is shown that the filter design using a training sequence synthesized from it gives noise reduction performance similar to the one at the
ISSN:1042-0967
DOI:10.1002/ecjc.4430740501
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
2. |
A realization of interpolator with the coefficients of the power of two |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 12-20
Chiaki Todaka,
Noriyoshi Kambayashi,
Preview
|
PDF (580KB)
|
|
摘要:
AbstractThis paper is aimed at the reduction of the hardware complexity and the computational complexity in the interpolation processing. A construction and design method for the interpolator is shown where only the coefficients of the power of two are employed. The interpolator is composed of the cascade connection of a passband compensatory filter, a zero‐interpolator, and a stopband forming filter. Since no multiplier is used in the proposed construction, zero coefficient sensitivity as well as a high‐speed operation are realized. FIR or IIR construction is used for the passband compensatory filter, and the stopband forming filter is constructed by a cascade connection of RRS.The design procedure for the interpolation is as follows. The linear programming is applied first to design the stopband forming filter. Then the passband compensatory filter is designed by the indirect exhaustive method or the least‐square method. It is shown through design examples that the proposed interpolator is useful in reducing the hardware complexity and the computational compl
ISSN:1042-0967
DOI:10.1002/ecjc.4430740502
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
3. |
Sound field analysis of ultrasonic probe with many medium boundaries |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 21-29
Hisao Okada,
Preview
|
PDF (556KB)
|
|
摘要:
AbstractFor the field analysis of an ultrasonic probe, the method based on Huygens' principle often is employed. Most such calculations, however, are concerned with the cases where the sound source and the region of calculation belong to the same medium or belong to a different medium divided by a single planar boundary.This paper reports on the two‐dimensional method of field analysis, which is based on Huygens' principle, but can be applied even if a large number of curved medium boundaries exists between the source and the region of calculation. The method analyzes the field as follows: the vibration amplitude and transmission coefficient on the medium boundary closest to the source are calculated. The product of the vibration amplitude and the transmission coefficient is calculated, and the medium boundary is regarded as the secondary source with the product value as the vibration amplitude. Then the field at the next medium boundary or in the next medium is calculated.The medium boundary is divided into line elements, and the transmission coefficient is calculated by determining the angle of incidence from the phase difference of the vibration at both ends of the element. As an application example of the proposed method of field analysis, the focused field is analyzed which is formed by the acoustic lens, cylindrical vibrator and the phased arra
ISSN:1042-0967
DOI:10.1002/ecjc.4430740503
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
4. |
Separation of mixed voices by acoustic parameter optimization |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 30-38
Tatsuya Morita,
Preview
|
PDF (608KB)
|
|
摘要:
AbstractThe human auditory response is highly adaptable to the environment, enabling voices to be recognized even in a noisy environment. This ability is due primarily to the language processing function in the central auditory nervous system, but the binaural sound locating function also seems to play an important role. From such a viewpoint, the author has presented a method which estimates the sound source locations and separates the voices from more than one speaker using the position information.A problem in this method is that the acoustic space is assumed as a free space, and the voice is separated based on the sound field model. Consequently, in the actual sound space, the estimation of the source location may be incorrect, thereby making the voice separation inaccurate.This study proposes a method whereby the separation index is estimated using the evaluation function based on the cross correlation of the separated voices, and the sound parameters are sought in the direction to minimize the evaluation function to improve the voice separation. The effectiveness of the method is verified by an experiment.
ISSN:1042-0967
DOI:10.1002/ecjc.4430740504
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
5. |
Stationary performance of series‐parallel‐type adaptive digital filter based on a logarithmic estimation |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 39-48
Masaki Kobayashi,
Yosio Itoh,
Preview
|
PDF (503KB)
|
|
摘要:
AbstractThe series‐parallel ADF is constructed for the unknown system by placing the logarithmic estimation ADF in series and the transversal ADF in parallel. Its features are that the examination of the stability is unnecessary since the transfer function as well as its inverse transfer function of the logarithmic estimation ADF are stable, and the number of taps can be reduced compared to the conventional transversal ADF.This paper presents the analysis and verification for the steady‐state estimation accuracy for the tap coefficient and the echo elimination, for the case where an additive noise exists at the near end of the series‐parallel ADF. The construction of ADF and the adaptive algorithm based on the LMS method is described first. Using the Taylor series expansion near the solution, the steady‐state algorithm is discussed. Finally, the derived result is verified by a computer simulation indicating that the theory and simulation agr
ISSN:1042-0967
DOI:10.1002/ecjc.4430740505
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
6. |
LSI module placement methods using neural computation networks |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 49-56
Hiroshi Date,
Terumine Hayashi,
Preview
|
PDF (544KB)
|
|
摘要:
AbstractWhen solving combinatorial optimization problems using the neural networks proposed by Hopfield, the objective function and constraints of (0, 1)‐variable optimization problems must be transformed. Then the problem is that there is no theoretical formula to decide the weighting factors of the terms of the objective function.This paper presents an LSI module placement algorithm which uses neural networks. Then an efficient method is proposed to decide fully adaptive parameters, where these parameters ensure convergence of the solution. Experimental results show that the new algorithm gives better solutions than the min‐cut algorithm, although it is not as good as the simulated annealing algori
ISSN:1042-0967
DOI:10.1002/ecjc.4430740506
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
7. |
Orthogonal periodic sequences with two complex numbers derived from M‐sequences |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 57-64
Tetsuo Kirimoto,
Yoshimasa Oh‐Hashi,
Takashi Hotta,
Preview
|
PDF (523KB)
|
|
摘要:
AbstractThe periodic sequence for which the sidelobe of the autocorrelation function is zero is called an orthogonal sequence. The orthogonal sequence is applied to various problems such as synchronization of communication and radar ranging. As a general method to generate the orthogonal sequence, the method to form the discrete‐Fourier transform of the periodic sequence with a constant amplitude is known.This paper proposes a method of generating the orthogonal sequence with an element of the sequence being composed of two complex numbers (orthogonal periodic sequence with two complex numbers). The method does not use the discrete‐Fourier transform, but maps the elements of the M‐sequence on Galois field GF(2) to complex numbers. Then the properties of the orthogonal sequence are discussed from the geometrical viewpoint and the range for which the orthogonal periodic sequence with two complex numbers exists is indi
ISSN:1042-0967
DOI:10.1002/ecjc.4430740507
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
8. |
Spectral efficiency of M‐ary/spread spectrum multiple access communication systems |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 65-77
Shin‐Ichi Tachikawa,
Gen Marubayashi,
Preview
|
PDF (878KB)
|
|
摘要:
AbstractIn this paper, M‐ary/spread spectrum multiple access communication systems (SSMA) are studied to improve spectral efficiency. A novel method of calculation of the spectral efficiency for several tip waveforms and bit error rate is shown. For M‐ary/ SSMA using orthogonal codes, the spectral efficiency can be improved considerably, however, the code length is becoming too long to achieve satisfactory spectral efficiency. To shorten the code length for the same spectral efficiency, M‐ary/SSMA is adopted using concatenated Reed‐Solomon codes, and the effectiveness i
ISSN:1042-0967
DOI:10.1002/ecjc.4430740508
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
9. |
A construction of M‐channel filter banks with hilbert transformers |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 78-87
Shigeo Wada,
Shin‐Ichi Takahashi,
Preview
|
PDF (608KB)
|
|
摘要:
AbstractThis paper proposes a new design method for the filter bank. The filter bank is a system which can reconstruct a signal via band‐splitting, decimation, interpolation and restoration. It is employed widely to solve various problems in the communication system, and various methods have been presented for its specification, implementation and design.This paper discusses the approximate realization of the filter bank, which is called the M‐channel uniform‐splitting FIR perfect‐reconstruction filter bank, and its design principle is as follows. First, the analysis filter is designed so that each channel has the characteristic which is the translation of the given LPF on the frequency axis. Then the synthesis filter is designed in the time‐domain so that the unit impulse response of the whole system is equivalent to the unit pulse, i.e., that the response of the linear‐phase all‐pass transmission characteristic. The Hilbert transformer plays the important role in the translation of the characteristic. Compared with the conventional design, the proposed method has the merit that no constraint is imposed on the band splitting and the band is partitioned uniformly with an arbitrary number of divisions. Design examples are shown to indicate the usefulness of the pr
ISSN:1042-0967
DOI:10.1002/ecjc.4430740509
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
10. |
Construction of 2‐D digital filters using all‐pass networks with complex coefficients |
|
Electronics and Communications in Japan (Part III: Fundamental Electronic Science),
Volume 74,
Issue 5,
1991,
Page 88-97
Hisamichi Toyoshima,
Shin‐Ichi Takahashi,
Preview
|
PDF (904KB)
|
|
摘要:
AbstractSeveral methods have been presented recently for constructing the one‐dimensional IIR digital filter by combining all‐pass networks. It is known that the filter composed of the all‐pass networks has a better computational efficiency and also low coefficient sensitivity due to the structural boundaries. In addition to the one‐dimensional filters, a construction of the two‐dimensional filter based on the all‐pass network has also been proposed. A problem in that method, however, is the restriction on the realizable characteristics.This paper attempts to overcome that problem, and proposes a new construction of the two‐dimensional digital filter using the all‐pass network. The new approach employs the complex coefficient all‐pass network instead of the conventional real‐coefficient all‐pass network. By this approach, the degree of freedom in the design is enhanced, and the characteristics which have been difficult or impossible to realize in the case of the real coefficient filters, ar
ISSN:1042-0967
DOI:10.1002/ecjc.4430740510
出版商:Wiley Subscription Services, Inc., A Wiley Company
年代:1991
数据来源: WILEY
|
|