摘要:

AbstractThis paper describes a method by which a moving robot with vision can recognize the environment and behave in a proper way by analyzing the consecutive images taken during movement in an artificial environment such as a building or plant. In this paper, the environment is assumed as a corridor in a building. The following two general features of the corridor are used as the premises for the image analysis: (1) the wall of the corridor is planar and is perpendicular to the floor; (2) the floor is flat. Since it is difficult to determine simultaneously the five three‐dimensional motion parameters, a method proposed in this paper determines successively the parameters by partitioning them into the rotation and translation components. The rotation is determined based on such points in the image as vanishing point and infinity point, which are not affected by the translation. Then the translation is determined from the optical flow and the rotation. The position information in the corridor is determined further. The proposed method features versatility since it does not employ the geometrical model represented by numerical values, utilizing only the properties valid in most of the artificial environment. An experiment is made in the corridor of a buildin

ISSN:0882-1666    

DOI:10.1002/scj.4690190201

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

年代:1988

数据来源: WILEY

摘要:

AbstractThis paper proposes a new public‐key cryptosystem based on the difficulty of solving a system of nonlinear equations. The proposed cryptosystem has the following features:1)The public‐key is a nonlinear transform from a plaintext to a ciphertext in the form of rational functions.2)The complexity of both encryption and decryption isO(m2), wheremis the plaintext length.3)Digital signature is possible.The two previously proposed systems based on the matrix decomposition and the squared matrix are special cases of the proposed system. The reliability of the cryptosystem when nonlinearity is limited to the polynomial form is discussed. Next, a publickey cryptosystem based on the difficulty of solving a system of nonlinear equations with rational functions is proposed, its decryption algorithm is studied, and the conditions for this cryptosystem to ensure reliability are derived. Finally, the computational complexity of encryption and decryption, the description volume of public and secret keys, and the possibility of digital signature are stud

ISSN:0882-1666    

DOI:10.1002/scj.4690190202

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

年代:1988

数据来源: WILEY

摘要:

AbstractThis paper discusses the psychological validity of the associative model HASP, which was propsoed by one of the authors. The fan effect and the priming effect are important properties of the retrieval process in human memory. The fan effect is a phenomenon that when the number of items associated with a concept is increased, a longer time is required to select an item from those items. In HASP, one of the multiply matched associative items is selected by the competitive circuit. The fan effect can be ascribed to the delay in the selection due to the increased number of competitions. The priming effect is a phenomenon where the preceding information processing affects the succeeding processing. Depending on the relation between the consecutive information, the effect can be accounted for by the activation of the preceding information left in HASP. It is also shown that the priming effect in the unconscious processing condition, where the cognition of the preceding information is prevented, can also be accounted for. Furthermore, it is shown that the processing in HASP is well matched to the two‐level theory in lexical retrieval, where the semantic selection process for the lexically ambiguous word is composed of the processes of activating all meanings and of selecting the one matched to the contex

ISSN:0882-1666    

DOI:10.1002/scj.4690190203

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

年代:1988

数据来源: WILEY

摘要:

AbstractSince bytecodes in Smalltalk‐80 are designed for stack machines, stack operations occupy half the total operations so that it is difficult to find low‐level parallelism. To perform fast processing with low‐level parallelism, we not only designed high‐level intermediate codes (I‐codes) which correspond to message passings on a one‐to‐one basis, but also introduced routines for determining multiple‐object pointers in parallel. We have developed both an I‐codes interpreter and a bytecodes interpreter on a low‐level parallel computer MUNAP and evaluated them.The results show that:(1)the static number of I‐codes is close to that of bytecodes, but I‐codes halved the number of dynamic execution steps of bytecodes;(2)the number of dynamic execution steps for basic cycles decreased from 30 percent for bytecodes to 17 percent for I‐codes so that the number of total execution steps was decreased 19 percent;(3)the number of execution steps for determining object pointers was decreased 12 percent because 69 percent of the object pointers were determined in parallel; and(4)compared with a single processor using bytecodes, the number of execution steps was 0.56 for bytecodes and

ISSN:0882-1666    

DOI:10.1002/scj.4690190204

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

年代:1988

数据来源: WILEY

摘要:

AbstractA new method is proposed in which all the characteristic vectors are classified into subvectors. Then a rough classification of handwritten characters is performed by unifying classification information obtained in each subvector all over again. In rough classification, “soundness” of classification and “quickness” of processing are required. To perform rough classification effectively, it is necessary to store considerable classification information in a dictionary. On the other hand, to implement processing quickly, it is necessary to decrease the amount of computation by condensing the classification information.The method herein not only realizes efficient rough classification through decreased redundancy by condensing classification information independently for each subvector of the characteristic vector, but also realizes a sounder rough classification by unifying classification information obtained in each subvector all over again. In the classification experiments, the amount of processing could be reduced to one‐fourth or one‐fifth that in the previous method to examine all the character classes without reducing a successful classification rate; this demonstrates the effectiveness of

ISSN:0882-1666    

DOI:10.1002/scj.4690190205

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

年代:1988

数据来源: WILEY

摘要:

AbstractWith the recent tendency toward miniaturization and price‐down of tablet digitizers, they are expected to be used as the input device for handwritten characters and drawings. As the preprocessing for the point coordinate sequence read from a tablet, only empirical methods have been applied. This paper discusses the preprocessing from the viewpoint of signal processing, and a quantitative evaluation was made for the following two points: (1) directional stability of the preprocessed coordinate sequence; and (2) follow‐up (fidelity) of the preprocessed point sequence to the original point sequence. As a result, a performance superior to the traditional methods was obtained by the following four phases: 1 Adaptive sampling, where sharp corners are stressed, while the sampling points are set uniformly for slow pen movement; 2 interpolation, where the point interval is made uniform for fast pen movement; 3 smoothing by the moving‐average filter; and 4 thinning to eliminate redundant points. The point sequence after the proposed preprocessing has the following features: (1) The original point sequence is followed, being little affected by the pen speed; (2) the directional stability is achieved even for 32 directional quantizations in the straight‐line segmentation used by the authors; and (3) features such as sharp corners are pr

ISSN:0882-1666    

DOI:10.1002/scj.4690190206

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

年代:1988

数据来源: WILEY

摘要:

AbstractThe time‐optimal control problem for a dc servomotor with an amplitude limitation in the manipulated variable is a complicated nonlinear problem. The problem has been investigated extensively from a theoretical viewpoint. However, little study has been made on the practical digital control system, even for a small‐scale controller using a microcomputer. This paper discusses the time‐optimal position control for a dc servomotor from such a viewpoint, and proposes a design method of the system based on the linear state feedback with the predetermined optimum sequence of a manipulated variable and the optimum sequence of the state trajectory. In other words, the optimal sequence of a manipulation variable and the optimum sequence of the state trajectory are predetermined by the linear programming, and the results are stored in ROM. In the actual control, the optimum sequence is read out from ROM to generate the desired sequence. The linear state feedback can reduce the effects of the external disturbance and the modeling error caused by the system fluctuation. Traditionally, the effects of the temporary disturbance and the modeling error have been a problem from the viewpoint of practical implementation, processing speed, and stability. The control minimizing these effects is the most remarkable feature of the proposed method. This method is suited to the controller using the microcomputer, and can be effective for the reduction of complexity and cost of the contr

ISSN:0882-1666    

DOI:10.1002/scj.4690190207

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

年代:1988

数据来源: WILEY

摘要:

AbstractThe blood‐flow measurements using nuclear magnetic resonance (NMR) signals are categorized roughly into three groups depending upon the nature of underlying information: intensity, frequency or phase. Among them, the method utilizing phase information is attracting particular attention. Since the phase is sensitive to the instrument imperfections, it has been customary to carry out image calculation with corrective image data measured separately. In this case, not only does multiple imaging require time, but image displacement due to the patient's movement comes into question. In this study, the quantitative relationship between the instrument imperfections and the phase shift of image was clarified, a corrective method was established, and the effect of a gradient magnetic field contained in the pulse sequence on a moving material was analyzed. Consequently, it has become possible to determine the velocity of blood flow perpendicular to the cross section with single scan. Finally, the experimental results of velocity measurement using the blood‐flow phantom was compared with the results of simulat

ISSN:0882-1666    

DOI:10.1002/scj.4690190208

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

年代:1988

数据来源: WILEY

摘要:

AbstractThe study of how sentences written in natural language are understood includes very important problems in examining human intelligence. At present, however, these problems have not been solved in detail. This paper defines the understanding of sentences as an extraction and organization of information requested for the next intelligent act from sentences.The act of question answering based on the input sentences is used and the basic process of sentence understanding, knowledge representation for sentence understanding, and the method of application for each of these, are examined.On the basis of this examination, an experimentally developed system is reported.In this system, sentences are input one by one. First, a sentence is transferred into the semantic representation in the form of a tree structure. This structure represents the dependency relationships between words. This semantic representation is read in sequential order by the basic processes of verb and noun phrases through matching the semantic representation with the definitive knowledge of the word and the general knowledge of the word in the form of a proposition.Finally, the proper meaning of polysemic words is identified and the meanings of omitted words and phrases are inferred. Also, some of the background information necessary to sentence understanding is identified.The results of these inferences are used to construct a frame. They are then connected to the other input sentences and organized to obtain the context information.After the input sentences have been analyzed, an interrogative sentence is input into the system, the system searches for an answer by referring to the context information, generates the semantic representation of the answer, and outputs its response in Japanese.

ISSN:0882-1666    

DOI:10.1002/scj.4690190209

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

年代:1988

数据来源: WILEY

摘要:

AbstractA new self‐organizing neural network model is proposed and evaluated for the mechanism of temporal pattern recognition of the auditory system. The model is constructed based on the hypothesis that total recognition of temporal patterns approximately of the length of words, is carried out by hierarchical identification and integration of the temporal relations of the constituent features. The model has a hierarchical structure in which short‐term memories storing spatial patterns, the circuits extracting temporally transient components of the pattern, and the feature detection circuits to identify spatial patterns are iteratively cascaded. After the circuit is self‐organized by repetitive presentation of training patterns, the model can correctly identify the training patterns and their temporally compressed and stretched patterns. It is also indicated that the short‐term memory function at each layer of the model is essential to the acceptance of temporally deformed patterns. Studies to expand the model to top‐down processing are also

ISSN:0882-1666    

DOI:10.1002/scj.4690190210

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

年代:1988

数据来源: WILEY