摘要:

AbstractSince the mobility of the hole is low in GaAs, P‐channel MESFET is of low speed. This makes it difficult to utilize the feature of the complementary logic circuit. Consequently, a direct‐coupled logic circuit is usually constructed using N‐channel MESFET. On the other hand, using the load‐current‐controlled logic circuit proposed in this paper, the speed and the power consumption can be improved compared with the directcoupled logic circuit. In other words, by limiting the load current according to the input state, the current counteracting the transition of the logic state can be decreased drastically. This speeds up the logic transition, thereby increasing the logic swing and reducing the power consumption.In this paper, the whole‐region model of the field‐effect transistor is introduced, and the logic swing of the GaAs ring‐oscillator is determined using Poincaré's mapping. Then the transient response of the load‐current‐controlled logic circuit is determined by simulation, leading to high speed and l

ISSN:0882-1666    

DOI:10.1002/scj.4690190301

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractLaser speckle photography is employed as a noncontact technique for measurement of minute in‐plane displacement of objects. With this technique, the direction and magnitude of displacement in a tiny portion of an object are expressed as the angle and spacing of Young's fringes. There has been a need for an image data processing procedure which can quantify and calculate displacement values. This paper describes a new algorithm, developed for this purpose, which employs projected profiles of the original image density and also the maximum entropy method (MEM). Specifically, the fringe angle is found by comparing projected density profiles obtained from various angles. MEM is used to calculate fringe spacing. Results of an experiment carried out to verify the accuracy of this algorithm confirm that it is relatively unaffected by optical noise and provides sufficient measurement accuracy for practical application even when there are few fringes. Accuracy to within ±2% is achieved for a displacement magnitude of 150 μm and to within ±1% for the displacement direc

ISSN:0882-1666    

DOI:10.1002/scj.4690190302

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractAristotle's syllogism is known in Japan as the “three‐stage inference.” However, it is not correctly understood in general, nor is it in the mainstream of the theory of logic now due to the widespread use of predicate logic. Syllogism is a logic system obtained by the direct analysis and description of the logical structure of linguistic knowledge and notion. Consequently, it includes several aspects suited to the knowledge information processing. The authors have proposed “tuple logic” as the logical framework for representing general knowledge described by a natural language, etc., and explored the possibility for the semantic representation and inference, which are different from those of predicate logic. This paper aims at the realization of a natural and efficient knowledge information processing by combining syllogism into tuple logic as the additional inference mechanism. The technique for efficient realization is investigated. In particular, an efficient validity‐testing procedure for syllogistic expressions is given by improving Lukasiewicz' decision method. The procedure is an application of a graph algorithm. When the set of premises is consistent, the inference can be performed in linear time, which is a useful property in the actual knowledge processing. As an example, a question‐answering system is demonstrated which can perform an inference for a certain range of Japanese sentences including

ISSN:0882-1666    

DOI:10.1002/scj.4690190303

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractA feature‐extracting model for the auditory nervous system was constructed using neurons with nonlinear inhibitory mechanisms as the elements. The operation was verified by computer simulation. The model is composed of the bandpass filters corresponding to basilar membrane, the half‐wave rectifiers corresponding to the hair cells, and the feature‐extracting neurons. The feature detector is a multilayered network, where several neural layers composed of a one‐dimensional array of neurons of the same response are connected in cascade. Each neuron performs the spatio‐temporal addition for the excitatory and inhibitory inputs through spatial connections in the input‐side layer and the first‐order delay. Then, the neuron produces an output through the nonlinear inhibitory mechanism (shunt inhibition) and the analog threshold characteristics. Considering the physiological knowledge and characters of the speech, three kinds of feature‐detecting neural groups were constructed. These are: CF group to detect the interval where the frequency is kept constant; FM group to detect the interval where the frequency is varying; and the fricative noise group to detect the fricative noise. These three correspond to the features of the vowel formant, the transition from consonant to vowel or the contracted sound, and the features of most of the consonants, respectively. Computer simulation was performed with actual speech input, and it was verified that the proposed model correctly extracts the various features containe

ISSN:0882-1666    

DOI:10.1002/scj.4690190304

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractThis paper proposes a method of extracting regions of car number plates using a parameter‐restricted Hough transformation (PRH). The PRH algorithm, in which the range of the parameter plane is restricted, reduces the computation time and the storage capacity more effectively than the usual Hough transformation. In this method, some features of car number plates are use

ISSN:0882-1666    

DOI:10.1002/scj.4690190305

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractIn most studies of the optimum allocation problem of the data file in the distributed database system, access to individual files is considered independent, ignoring the relation among files. This assumption may not be appropriate in actual cases. In this paper, the expression for the query cost is simplified as long as the essential aspect of the problem is maintained. A formulation is made based on a clear distinction between the part of the file which can be considered as independent in the individual access, and the part which is not. Then for each formulation, the computational complexity in the optimum file allocation is discussed. It is shown that the problem for the former can be answered in a polynomial time, while the latter is NP‐har

ISSN:0882-1666    

DOI:10.1002/scj.4690190306

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractIn the recent workstation for CAD and CG, a color map display is often used, which can display 256 colors selected from 224(approx. 16000000) colors using a look‐up table. Even in the computer system with such a color display with such a limited number of colors, it is desired that the continuous color image should be processed. In most cases, by determining colors which are important to the image (representative colors), the color image can approximately be displayed without much degrading the image quality. This paper proposes a new method for the approximate display of the color image by 256 colors. In the method, RGB color space is successively partitioned to determine 256 representative colors, so that the quasi‐optimum solution is obtained in the sense of the least‐mean‐square error. By the proposed method, a highly accurate approximate image is obtained with several times faster speed, compared with the traditional methods. The work memory required in the program is only 100 to 200 kBytes, which is less than in the traditional

ISSN:0882-1666    

DOI:10.1002/scj.4690190307

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractThis paper discusses the speeding‐up of the color image quantization procedure. The color image quantization aims at the most faithful representation of the original image using a limited number of colors. The color image quantization is composed of two processings which are the construction of the color table and the construction of the quantized image. In the construction of the quantized image. a method is usually employed in which the color of each pixel is compared with all colors in the color table. However, this method requires a long processing time, occupying more than 90 percent of the total processing time. From such a viewpoint, this paper aims at the speeding‐up of the color image quantization, by reducing the processing time for construction of the quantized image. A method is proposed for speeding up, where the pixel and the color table are connected indirectly using the look‐up cube composed of 32 × 32 × 32 equipartitioned RGB space. By this method, the number of comparisons between the pixel and the color table is

ISSN:0882-1666    

DOI:10.1002/scj.4690190308

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractThis paper describes a method of extraction information of the position of an object three‐dimensionally via a stereograph. In stereography which is a general method of obtaining three‐dimensional information from two‐dimensional images, it is difficult to find a corresponding point of an object in multiple images photographed from different positions. In most cases, this is based on the local characteristic of edges which are features of an image, hence the results are not consistent. This paper proposes a method in which the correspondence between the pair images is found by comparing the structures of feature‐regions represented by the relationships to the adjacent edgelines (e.g., parallelism or crossing). The usefulness of the method was confirmed by applying it to an indoo

ISSN:0882-1666    

DOI:10.1002/scj.4690190309

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY

摘要:

AbstractIn order to determine the shape and the motion of the object from its projection, it suffices to derive and solve the equation relating the parameters specifying the shape and the motion of the object with the features measured on the image. The equation is usually nonlinear, and is difficult to solve in the analytic way. This paper notes that there is no inherent coordinate system on the image, and any two coordinate systems are equivalent. By combining the features of the image, coordinate‐rotation‐invariant parameters (invariances) are derived. By this scheme, the geometrical meaning of the image characteristics is made clear, and the analytic solution is obtained in a natural way. The idea is demonstrated for the cases of the optical flow analysis and the surface recovery by texture, as examp

ISSN:0882-1666    

DOI:10.1002/scj.4690190310

出版商:Wiley Subscription Services, Inc., A Wiley Company    

年代:1988

数据来源: WILEY