scispace - formally typeset
Search or ask a question

Showing papers in "Electronics and Communications in Japan Part Iii-fundamental Electronic Science in 1995"


Journal ArticleDOI
TL;DR: Based on Lin's method of constructing the divergence, a new divergence called Hermite-Hadamard divergence is introduced and the property of the proposed divergence, as well as its relation to Lin's inequality, are discussed.
Abstract: The f-divergence is a general class of divergences which includes various divergences used in representing the difference between two probability densities. It was introduced by Csiszar. Lin, on the other hand, in 1991 introduced a divergence based on a new idea. the inequalities relating the divergence to the existing two divergences are shown, and the results are applied to the problems in information theory. This paper discusses the relation between the inequality concerning Lin's divergence and the inequality concerning the f-divergence. Based on Lin's method of constructing the divergence, a new divergence called Hermite-Hadamard divergence is introduced. the property of the proposed divergence, as well as its relation to Lin's inequality, are discussed. Using the proposed divergence, an inequality which is improved further compared to the inequality relating the Kullback diver gence and Hellinger distance is derived. Those results are derived in this paper.

44 citations


Journal ArticleDOI
TL;DR: In this paper, the authors measured the quality of falling snow particles as a help in analyzing radio attenuation during snowfalls by means of image processing and the weight of all snow particles that have fallen to the ground.
Abstract: It is important to measure the quality of falling snow particles as a help in analyzing radio attenuation during snowfalls. Since the quality of snow particles depends almost entirely on their density (the amount of water they contain), the snow particle density reflects their quality. to permit automatic measurement of the average densities of falling snow particles over long periods of time, the diameters and fall velocities of snow particles are measured by means of image processing. Simultaneously, the weight of all snow particles that have fallen to the ground was measured directly by means of an electronic balance. the density of the falling snow particles is calculated by dividing into this total weight the total volume of snow particles that passed through a unit volume (obtained from the diameter and velocity of snow particles). the diameters, velocities, and number of snow particles are determined by image processing. Snowfall rates calculated from the image processing data are compared with those measured directly and are found to agree.

23 citations


Journal ArticleDOI
TL;DR: It is found that a relatively higher success rate for the memory retrieval can be achieved by using the association model instead of the conventional model and the traveling salesman problem can be solved using parameter control reminiscent of the annealing in the Boltzmann machine.
Abstract: A chaos neural network with externally controlled parameters is proposed and a few applications to an association model and a combinatorial optimization problem are given. It is found that a relatively higher success rate for the memory retrieval can be achieved by using the association model instead of the conventional model (burdened by chaotic wandering) because unacceptable trapping at a spurious state is avoided and the traveling salesman problem can be solved using parameter control reminiscent of the annealing in the Boltzmann machine.

18 citations


Journal ArticleDOI
TL;DR: In this article, the authors used potential as a way to systematically represent the conditions contributing to collision avoidance such as a threat of collision with other aircraft and degrees of deviation from the original course.
Abstract: This paper describes ground-based avoidance for horizontal maneuvers using a concept of potential as a collision avoidance aircraft procedure. First, mathematical formulation is introduced by using potential as a way to systematically represent the conditions contributing to collision avoidance such as a threat of collision with other aircraft and degrees of deviation from the original course. The flight course with a lower potential is considered safer. In the choice of a specific avoidance course, an avoidance pattern generation method by means of potential gradient is shown. the future location of the aircraft is sampled at a certain time interval. the shift of the sampling point corresponding to the potential gradient at each sampling point or the force applied to each sampling point (attractive or repulsive) is repeated. It was confirmed that by means of this method, avoidance patterns without danger of collision can be obtained with only horizontal avoidance except for special cases. However, when the sampling points are symmetric or when more than three aircraft are involved, the procedure is not adequate and some provisions are needed. to this end, the sampling point arrangement in the initial stage before the avoidance pattern is formed and the three-dimensional avoidance including vertical avoidance can be considered.

17 citations


Journal ArticleDOI
TL;DR: It is verified as a result that the proposed method can well suppress the increase of the prediction error due to the outliers, improving greatly the accuracy of the predicted value compared to the conventional method.
Abstract: In recent years, there have been intensive studies on the short-term prediction method for time-series data with chaotic properties. the prediction for the chaotic time-series data is executed, for example, as follows. Based on the single-variable time-series data, the trajectory of the attractor is reconstructed in the multidimen sional space and the change of the trajectory is predicted by a polynomial approximation. However, when there exists a noise such as the outlier in the time-series, the prediction error is increased drastically. To solve this problem, this paper proposes a prediction method which is robust against the outlier. More precisely, the following three elaborations are added to the conventional method: (1) the coefficients of the polynomial are calculated by Biweight's estimation method; (2) the order of the polynomial is determined automatically examining the residual error; and (3) the data with a large prediction error are replaced by the predicted value and the next prediction is applied. an evaluation experiment is executed for the data where the pseudorandom variable with the Cauchy distribution is superposed on the chaotic time-series data generated by the Henon map and the Lorenz model. It is verified as a result that the proposed method can well suppress the increase of the prediction error due to the outliers, improving greatly the accuracy of the prediction compared to the conventional method.

17 citations


Journal ArticleDOI
TL;DR: In this paper, a computer experiment using plasma has been carried out with an electromagnetic particle code, in which not only the simple three-wave coupling but also the waveparticle interaction are included.
Abstract: The concept of a solar power station (SPS) was proposed 25 years ago. During the intervening years, long-distance, high-power transmission microwave technology has advanced. Realization of wireless power transmission in space is now no longer a dream. It has been predicted from three-way coupling that a static plasma wave is excited when an intense microwave passes through space plasma. the microwave power transmission rocket experiment in an ionosphere carried out by the authors (MINIX, Microwave Ionosphere Nonlinear Interaction experiment) has confirmed this fact. In this experiment, a phenomenon that cannot be explained by the three wave resonance theory has been discovered. In this paper, a computer experiment using plasma has been carried out with an electromagnetic particle code. Static plasma wave excitation and the saturation phenomenon are explained in which not only the simple three-wave coupling but also the wave-particle interaction are included.

15 citations


Journal ArticleDOI
TL;DR: A layered neural network is proposed for exact edge detection where the inputs consist of three pixel values and a local variance in a 2-D mask and this network can detect edges and suppress the noise in an image at the same time, and its performance is adjusted by learning.
Abstract: An algorithm is developed for the restoration of an image degraded by a known two-dimensional (2-D) shift-invariant point-spread function and corrupted with white noise. A layered neural network and the Hopfield network are used for the edge detection, and the restoration and smoothing of a blurred image, respectively. In particular, a layered neural network is proposed for exact edge detection where the inputs consist of three pixel values and a local variance in a 2-D mask. This network can detect edges and suppress the noise in an image at the same time, and its performance is adjusted by learning. Finally, some examples are given to illustrate the utility of the proposed algorithm.

13 citations


Journal ArticleDOI
TL;DR: In this method, the coding of the static image is reduced to the optimization problem where the error between the original image and the decoded image is minimized and the distortion function is defined so that the noise generated by encoding is suppressed from the viewpoint of the whole image.
Abstract: When the original signal to be encoded has a strong correlation in the neighborhood, as in the case of the natural image, it can be expected that the original signal is represented by a smaller number of bits. This paper proposes the following low-bit coding system. the analog image is converted into the digital image by A-D conversion in the transmitter based on the dynamics of the locally interconnected discrete-time cellular neural network (DTCNN). the digital image is passed through a lowpass filter in the receiver so that the original analog image is restored. In this method, the coding of the static image is reduced to the optimization problem where the error between the original image and the decoded image is minimized. the distortion function is defined so that the noise generated by encoding is suppressed from the viewpoint of the whole image. A high-quality image is reconstructed by the dynamics of the multivalued neuron, which is an extension of the output of the neuron to multiple values. In this method, a parallel processing dynamics is employed, where a complex function can be realized from simple devices, and it is expected to realize the digital conversion in real-time of the analog information which is input in parallel, as in the case of the light input to the retina.

13 citations


Journal ArticleDOI
TL;DR: The purpose of this study is to verify that EEG is composed of frequency components with chaotic characteristics, and it is found that the fractal properties exist in most of the frequency components.
Abstract: The purpose of this study is to verify that EEG is composed of frequency components with chaotic characteristics. EEG data are decomposed into frequency components, and the chaos and fractal analyses are applied. More precisely, the running spectrum of EEG is derived and the chaos analysis is applied using the Lyapunov spectrum and the correlation dimension. As a result, the existence of the chaotic phenomenon is verified for most of the frequency components used as the objects of the analysis. It is found also that the fractal properties exist in most of the frequency components. the multifractal analysis is applied to the EEG data, and the generalized dimension of q-th order is determined. It is seen as a result that the reconstruction attractors of EEG are placed by a nonuniform fractal distribution, which is represented by a large number of scaling indices concerning the dimension.

12 citations


Journal ArticleDOI
TL;DR: In this article, the optimal ADF coefficient and the least-mean-square error (LMSE) in the subband adaptive system were theoretically derived and the derived theoretical value is compared to the results of the experiment and the effectiveness of the proposed model and the theoretical expression are demonstated.
Abstract: It is well known that the subband adaptive system is useful; however, the study of the theoretical analysis of the system does not seem sufficient. the theoretical analysis is problematic because of the structural constraint in which the input to the unknown system is not the same as the input to the adaptive filter. This paper attempts theoretical analysis of the subband adaptive system and proposes a new equivalent model that removes the foregoing constraint. Then the theory of the Wiener filter is applied to the equivalent model. the optimal ADF coefficient and the least-mean-square error (LMSE) in the subband adaptive system are theoretically derived. Lastly, the derived theoretical value is compared to the results of the experiment and the effectiveness of the proposed model and the theoretical expression are demonstated.

11 citations


Journal ArticleDOI
TL;DR: In this article, the generating condition of the electrostatic plasma wave phenomena excitated in the plasma medium by a large signal electromagnetic wave is theoretically investigated and the nonlinear coupling coefficient between the large signal EM wave and the ionospheric plasma wave is derived.
Abstract: It is well known that refraction, Faraday rotation, scintillation and absorption of microwaves take place when a microwave passes through an ionospheric plasma. However, when the amplitude of the microwave is large, nonlinear interactions occur with plasma in addition to these plasma interactions with a small signal microwave. Various phenomena not conceivable for weak microwaves used in communication can take place. Especially, it is predicted that the three-wave resonance, which is the most fundamental nonlinear wave-wave interaction, can be applied to the nonlinear wave-wave interaction between the plasma and a large signal microwave. In this paper, the generating condition of the electrostatic plasma wave phenomena excitated in the plasma medium by a large signal electromagnetic wave is theoretically investigated and the nonlinear coupling coefficient between the large signal electromagnetic wave and the electrostatic plasma wave is derived.

Journal ArticleDOI
TL;DR: The model is proposed and response characteristics of the single neuron are proposed and it is shown that the network model with loop structure can select time interval sequence peculiar to the network structure even if random noise is superimposed.
Abstract: In this paper, we analyze some possible functions of networks where delays of pulse propagation and asynchronous events of pulse occurrence are considered, on the basis of a model of asynchronous chaotic neural networks. First, the model is proposed and response characteristics of the single neuron are numerically analyzed. Second, it is shown that the network model with loop structure can select time interval sequence peculiar to the network structure even if random noise is superimposed.

Journal ArticleDOI
TL;DR: It is shown as a result that the proposed coding system realizes the same speech quality at 16 kbit/s as that of the 32-k bit/s ADPCM, even though there is no algorithm delay.
Abstract: This paper proposes the ADPCM speech coding system using mel-cepstrum as the parameter for the short-term prediction. In the proposed coding system, the backward type adaptive prediction is applied based on the adaptive mel-cepstrum analysis. Operations such as noise shaping and postifitering are executed through the melcepstrum. Consequently, it is expected that the noise shaping and the postfiltering have the effect which is matched to the human auditory characteristics. This paper presents the configuration of the coding system with some improvements such as the addition of the pitch predictor, and the quality of the coded speech is evaluated. It is shown as a result that the proposed coding system realizes the same speech quality at 16 kbit/s as that of the 32-kbit/s ADPCM, even though there is no algorithm delay.

Journal ArticleDOI
TL;DR: A method is developed for an accurate measurement of the three-dimensional shape of the vocal tract, using magnetic resonance imaging (MRI), aiming at the analysis of the features ofThe vocal tract shape in the utterance by children, males and females.
Abstract: In this paper, a method is developed for an accurate measurement of the three-dimensional shape of the vocal tract, using magnetic resonance imaging (MRI), aiming at the analysis of the features of the vocal tract shape in the utterance by children, males and females. to derive the precise data for the shape of the vocal tract, the data for the vocal tract of the oral cavity is derived from the coronal MR image from the lips to the atlas. the data for the vocal tract from the glottis to the palate are derived from two kinds of horizontal MR images. High-speed photography is used, and the imaging time is 2.9 s for each slice. the central line of the vocal tract is estimated from the mid-sagittal MR image, and the vocal tract length is calculated. Based on the mid-sagittal MR image, the three-dimensional shape of the vocal tract is reconstructed from the vocal tract contour in the three kinds of images. To process the large amount of data, a method is developed that automatically extracts the vocal tract contour from the MR image. A correction method is proposed for the teeth shape which cannot be imaged by MRI, based on the MR image of the dental impression. the error of the measuring method is evaluated quantitatively using a model for the vocal tract. the transfer function for Japanese /a/ are estimated by the proposed measuring method, and the result is compared to the result of LPC analysis of the speech recorded at the site. It is seen as a result that the errors for F1, F2, and F3 are 8, 17, and 141 Hz, respectively.

Journal ArticleDOI
TL;DR: In this article, the authors proposed a new approach which is totally different from the forementioned conventional methods, i.e., the global optimization method for the unconstrained nonlinear optimization problem where chaos is introduced.
Abstract: Most of the actual optimization problems are nonlinear and have many peaks (nonconvex). With the widespread use of the high-speed and large-capacity computers as the background, it has recently been felt highly necessary to derive the global optimal solution for the optimization problem which is nonlinear and has multiple peaks. It is one of the most important topics of research in the field of optimization. The major global optimization method developed until now can be divided into the trajectory method, the function transformation method, and the simulated annealing method. This paper proposes a new approach which is totally different from the forementioned conventional methods, i.e., the global optimization method for the unconstrained nonlinear optimization problem where chaos is introduced. the proposed method is based on the dissipative system where the inertia term and the nonlinear damping term are added to the conventional gradient method. By appropriately adjusting the characteristics of the nonlinear damping term, the generation of chaos can itself be controlled. Then, by overriding the barrier of the energy function between the local minima, the process converges to the global optimal solution. Finally, the proposed method is applied to the typical multipeaked nonlinear optimization problem with two and ten variables and it is shown that the global optimal solution can be derived by adequately adjusting the parameter of the nonlinear damping term.

Journal ArticleDOI
TL;DR: The hysteresis net is considered in which all crossconnections are uniform and the control paameters are restricted to two, and an associative memory is proposed which has the ideal memory-recall function for the information.
Abstract: The hysteresis net is considered in which all crossconnections are uniform and the control paameters are restricted to two. an analysis is presented concentrating on the case where the state specified by the system is a stable equilibrium. A complete result is obtained for the type and number of the stable equilibria of the systm, as well as the convergence region for each stable equilibrium. As an application example of the hysteresis net, an associative memory is proposed which has the ideal memory-recall function for the information. the associative memory is composed of three units which are the hysteresis net. the initial value generator, and the output circuit. It has the function that almost completely solves the problems in the traditional associative memory. In other words, all the contents which are prespecified as desirable are securely memorized and the false stable memory content is not produced. When incomplete information is given, the desired memory with the smallest Hamming distance is securely recalled. the proposed associative memory can very easily be implemented, and the desired content can always be stored using the switches.

Journal ArticleDOI
TL;DR: In this article, a new practical and simple method of critical flicker frequency (CFF) measurement and analysis is considered, where the CFF at the start of the measurement is used as the initial value and the relative change is measured in a short time with high accuracy.
Abstract: Aiming at the quantitative measurement of the fatigue of the television observer, a new practical and simple method of critical flicker frequency (CFF) measurement and analysis is considered. As to the measurement, a method is considered where the CFF at the start of the measurement is used as the initial value and the relative change is measured in a short time with a high accuracy. It is verified experimentally that the measurement by the descending series of the method of limit is suitable. As to the analysis, two methods are compared. One is to normalize the average of measured values of all the observers, and the other is to normalize the measured value of each observer and then to calculate the average. It is shown as a result that the latter can represent more clearly the change of CFF. Then the effectiveness of the proposed method in the measurement of the fatigue of the television (TV) observer is examined. CFF of the actual TV image observer is measured, including the binocular three-dimensional (3-D) TV. As a result, it is verified statistically that CFF depends on the observation time and the image display method. the result corresponds well to the subjective evaluation of the fatigue. Thus, it is concluded that CFF can be a useful measure in assessing quantitatively the fatigue of the TV image observer, including the binocular 3-D TV. It is expected that the image quality and the improvement effect of the observation condition can be examined quantitatively.

Journal ArticleDOI
TL;DR: In this article, the authors proposed the optimal design for the separable-denominator two-dimensional digital filter, using the genetic algorithm (GA) for the specification in the frequency domain.
Abstract: This paper proposes the optimal design for the separable-denominator two-dimensional digital filter, using the genetic algorithm (GA). the design proposed is for the specification in the frequency domain. the simple GA, which is the most basic among GA, is used in the proposed optimal design. the parameters of the transfer function of the separable-denominator twodimensional digital filter are encoded into bit sequences, and then GA is applied to design the filter. By the use of the separable-denominator transfer function, the stability of the filter is completely guaranteed. The approximation error of the proposed method is compared to that of the conventional method through design examples. the convergence of GA is discussed. It is shown that the proposed optimal design can easily cope with the modification of the design specification of the two-dimensional digital filter. It is shown also that the filter can be designed using the maximum error or the absolute error sum as the error evaluation function in addition to the mean-square error.

Journal ArticleDOI
TL;DR: In this article, an adaptive α-trimmed mean filter based on local statistics is proposed, which has an excellent performance in the preservation of the detailed signal components, and the performance of the filter is evaluated.
Abstract: The nonlinear filter is useful in restoring an image which is degraded by additive noise. the α-trimmed mean filter is a generalization of the median filter. In the α-trimmed mean filter, the data within the filter window are examined and a specified number of samples is selected as close to the median value from the ordered statistical viewpoint. Then the mean of the selected samples is output. This mechanism is powerful in eliminating the impulsive noise and suppressing the non-impulsive noise. On the other hand, it should be noted that the image and other signals are composed of various components (such as the flat part and the edge) and it is indispensable that the number of samples to be averaged in the α-trimmed mean filter should be made variable according to the local information (i.e., adaptive filtering) to realize a highly precise restoration of the signal. From another aspect, the α-trimmed mean filter has a problem in that the detailed behavior of the signal cannot be preserved when the filter window is large. From such a viewpoint, this paper newly proposes an adaptive α-trimmed mean filter based on the local statistics, which has an excellent performance in the preservation of the detailed signal components. the performances in the preservation of the edges and the elimination of the noises are evaluated. Finally, the filter is actually applied to the image restoration and the practical usefulness of the proposed filter is demonstrated.

Journal ArticleDOI
TL;DR: The operational characteristics are assumed so that the output for the normative input satisfies the specified conditions and the constraint is derived from the specification and the template is designedSo that the corresponding cost function is minimized.
Abstract: The cellular neural network (CNN) is composed of a planar placement of cells which consists of a (piecewise-linear) nonlinear element and controlled current sources. It features a simple structure which is close to that of the retina and is expected to be utilized in pattern recognition and image processing. In CNN, each cell is connected to the neighborhood cells by the same pattern, and by adjusting the connection pattern, CNN with various functions can be designed. the connection pattern is called the cloning template. It is very important in the development of new CNN to establish the design method for the cloning template. In this paper, the operational characteristics are assumed so that the output for the normative input satisfies the specified conditions. the constraint is derived from the specification and the template is designed so that the corresponding cost function is minimized. the simplex method is used as the optimization technique, which features a simple algorithm. As application examples, noise-remover CNN as well as the maze-tracing CNN are designed and satisfactory results are obtained. the design method is reported in this paper.


Journal ArticleDOI
TL;DR: The proposed technique is applied to the real data, as the inference example of the method, and the strategy for presenting the items is compared to item response theory, and discussions are made from the theoretical viewpoint.
Abstract: This paper aims at estimation of the state of the knowledge (understanding) of the individual learner by executing a test composed of a small number of items. The following approach is proposed: (1) a teacher model is introduced, and the inference process for the understanding state of the learner is represented using the Bayesian network; (2) as the network information of the items based on the teacher model, EVINI (expected value of item network information) is proposed. As the test information, EVTIN (expected value of test information with network) is proposed; and (3) a strategy for item selection based on EVINI is proposed. The following advantages are realized in practice: (1) only a small amount of data is required to construct the model; (2) the goal of inference is to know the understanding state of the individual learner, and the result provides the feedback from an educational viewpoint; (3) the knowledge of the teacher about the domain can be utilized in the inference process, which helps to improve the prediction efficiency; and (4) it is easy to add or modify the knowledge of the teacher model. In this paper, the proposed technique is applied to the real data, as the inference example of the method. the strategy for presenting the items is compared to item response theory, and discussions are made from the theoretical viewpoint.

Journal ArticleDOI
TL;DR: In this article, a technique for minimizing the normalized noise variance at the output of two-dimensional (2D) recursive digital filters is developed where error spectrum shaping of the filter is performed by means of higher-order feed-back.
Abstract: A technique for minimizing the normalized noise variance at the output of two-dimensional (2-D) recursive digital filters is developed where error spectrum shaping of the filter is performed by means of higher-order feed-back. First, the optimal solution of error feedback coefficients is derived. Then suboptimal error feedback coefficients are found in which the error feedback coefficients are symmetric. Here, both the optimal and suboptimal solutions are found analytically by solving the minimization problem of a quadratic form. Finally, a numerical example is given to check the validity of the proposed algorithm and it is shown that the proposed technique is effective even if the optimal or suboptimal coefficients are rounded to a power of 2.

Journal ArticleDOI
TL;DR: In this paper, a design method for 2-D maximally flat diamond-shaped filters with rectangular support of impulse response is proposed, which combines the derived conditions to form a system of linear equations with respect to the independent filter coefficients.
Abstract: Two-dimensional (2-D) diamond-shaped filters are important filters with many applications and are known to be designed and implemented efficiently as 2-D half-band filters. On the other hand, filters with different number of taps in both directions are often suitable for a specific application rather than filters with equal number of taps. This paper proposes a design method for 2-D maximally flat diamond-shaped filters with rectangular support of impulse response. First, the maximally flat conditions at the origin are considered together with equations which make the filter approximately diamond-shaped. Then, by combining the derived conditions to form a system of linear equations with respect to the independent filter coefficients, designs for the difference in the number of taps 2, 4, and 6 are demonstrated. Finally, together with giving several design examples, comparison with other design methods are made and the usefulness of the suggested design method is verified.

Journal ArticleDOI
TL;DR: In this paper, a new zero-crossing representation is proposed in which the zero crossing points are extracted from the multiscale wavelet transform, and the information concerning the maximum value (the minimum value when negative) between two consecutive zero cross points as well as its position is combined.
Abstract: This paper discusses a reconstruction method of the image signal from the zero-crossing representation. First, the one- and two-dimensional multiscale wavelet transforms are described which have the completeness and the translation invariance. A new zero-crossing representation is proposed in which the zero-crossing points are extracted from the multiscale wavelet transform. the information concerning the maximum value (the minimum value when negative) between two consecutive zero-crossing points as well as its position is combined. Then it is shown that the original signal can be reconstructed with a stable and fast convergence from the zero-crossing representation of the signal through the iterated projections to the convex set. The effect of the wavelet basis filter on the convergence in the reconstruction is investigated, and a necessary condition for the optimal basis filter is derived considering the convergence. Finally, a new basis filter is designed using the B-spline function and the usefulness of the proposed method is demonstrated for the one-dimensional signal and the image signal. It is shown also that the signal reconstruction procedure converges with stability, even if there exists an error at the extracted zero-crossing point.

Journal ArticleDOI
TL;DR: In this article, the convergence of the blind equalization algorithm is dependent on the condition number of the correlation matrix of the input sequence, which is not known to the best of our knowledge.
Abstract: A major problem with blind equalization algorithms based on the distribution matching principle is that they need a long time to accomplish their convergence. What affects convergence has not been discovered to date. the similarity between the least mean-square (LMS) algorithm (widely used for adaptive equalization) and the blind algorithms is considered here. It is expected that the convergence of the blind algorithms is dependent on the condition number of the correlation matrix of the input sequence. Prefiltering methods, including coefficient-fixed type and coefficient-adaptive type, are derived for blind equalization. In these methods, the prefilters are realized by a prediction error filter, which has the ability to compensate for amplitude distortion induced by a channel. Since the prefilters have such ability, the blind equalizers to be cascaded with them are required only to compensate for phase distortion. As a result, the burden imposed on the blind equalizers is reduced. the prefilters output a near-white sequence and lead to an improvement in the convergence of the blind algorithms, being degraded as the condition number is increased. the effective ness of the proposed methods is validated by computer simulations.

Journal ArticleDOI
TL;DR: A method is proposed which utilizes the nonlinear input-output relation of the fuzzy inference and the spatial distribution of the image to be adapted to each object to improve the discrimination rate for the ultrasonic image.
Abstract: This paper discusses the method for the semiautomatic discrimination of the biological tissue and organs, based on the medial ultrasonic data. the ultrasonic image is difficult to be evaluated on the absolute basis, since multiple physical information is included simultaneously and the image is affected greatly by the individuality of the object. This method utilizes the feature parameters of the image for various tissues, and the discrimination is executed on the relative basis by comparing the feature parameters. In the past method, the general procedure is to utilize the first-order statistics such as the average image density. In the case of a tumor in the liver, for example, there does not exist a remarkable density difference between the image of the tumor and the surrounding normal tissues, which prevents a satisfactory discrimination of the tumor and other images. This paper aims at the discrimination of tissue images with small density differences, based on the textural feature parameters (the second-order statistics) contained in the image. However, the crisp-like discrimination is difficult since there is an overlapping ambiguous area between the histograms of the feature parameters of the tissues. to improve the discrimination rate for the ultrasonic image, a method is proposed which utilizes the nonlinear input-output relation of the fuzzy inference and the spatial distribution of the image to be adapted to each object. Using the data obtained from the test phantom for the ultrasonic diagnostic equipment and the biological data obtained from four subjects, the effectiveness of the proposed method is demonstrated.

Journal ArticleDOI
TL;DR: A Haar wavelet transform with interband prediction is proposed that permits high-speed processing and which is suitable for coding as a subband decomposition technique since it generates little block distortion in the decoded images.
Abstract: Although the discrete cosine transform (DCT) and wavelet transform have been used as effective techniques for reducing the redundancy of image waveforms, they have the problems of high complexity and causing block distortion. In this paper, a Haar wavelet transform with interband prediction is proposed that permits high-speed processing and which is suitable for coding as a subband decomposition technique since it generates little block distortion in the decoded images. the interband prediction method of this proposed transform method uses the derivative of the low-band waveform to predict the high-band waveform each time a band is split into two sub-bands. This results in a transform process that gives a prediction residual signal with lower entropy. During the inverse transform, the prediction coefficients obtained from the forward transform are used to predict the high-frequency waveforms and then, by adding the prediction residual, the original high frequency waveforms are reconstructed. Because the high-frequency waveforms are generated that smoothly interpolate the up-sampled, low-frequency waveforms, any block effects from this prediction technique are difficult to see in the decoded image. A detailed analysis of the proposed interband prediction process is performed by interpreting the proposed transform as a subband decomposition process that is based on the use of a symmetric short-kernel filter (SSKF) filter bank; then experiments are conducted that demonstrate its performance as a subband decomposition for coding of real image data and the possibility of progressive build-up of the decoded images. Results show that this transform method is effective for coding image data.

Journal ArticleDOI
TL;DR: In this paper, a nonlinear filtering technique based on the fuzzy control rule is presented and the usefulness of the filter in the noise elimination from the image signal is demonstrated, where the coefficient of the smoothing filter is determined according to the properties of the data in the window.
Abstract: When a random noise is superposed on a signal with an abrupt change (i.e., edge), as in the case of the image signal, there is a large overlap of the frequency component between the abrupt changes (edge) of the signal and the noise. Then it is impossible to realize both the noise elimination and the edge preservation sufficiently by the conventional linear filter, such as the mean filter. Thus, a nonlinear technique is required in order to eliminate the noise and preserve details of the original signal where the coefficient of the smoothing filter is determined according to the properties of the data in the window. There can be more than one observed value in the filter window that can be used to specify the properties of the data. It is difficult, however, to represent precisely the properties of the image signal using a single observed value. In addition, the observed values contain ambiguities since they are calculated from the local image data with superposed noise. In other words, it is required to determine the filter coefficients based on multiple obseved values containing ambiguities. For such a problem, the fuzzy control rule, which is utilized effectively in control problems, can be applied to calculate the control variable (i.e., filter coefficient) adequately by the nonlinear technique from the multiple observed values containing ambiguities. In this paper, the nonlinear filtering technique based on the fuzzy control rule is presented and the usefulness of the filter in the noise elimination from the image signal is demonstrated.

Journal ArticleDOI
TL;DR: The nth order statistical number of degrees of freedom is defined as the equivalent number of uncorrelated points in the square finite region in the homogeneous random field as well as the correlation area and the correlation length, which are derived from the definition.
Abstract: One of the important features that characterize the physical random field or the image texture is Taylor's correlation length. However, there has not been presented a clear-cut theory as to the foundation of the correlation length from the viewpoint of probability theory or statistical theory. Another aspect is that the feature is one-dimensional, and, therefore, the application to the random field is limited to the isotropic random field. This paper defines the nth order statistical number of degrees of freedom as the equivalent number of uncorrelated points in the square finite region in the homogeneous random field. From the definition, the correlation area and the correlation length are derived. the correlation area is referred to as the average area of the region statistically affected by an arbitrary point on a two-dimensional field. the correlation length represents the effective distance between adjacent independent or uncorrelated points on the line parallel to the x- or y-axis. the first-order correlation length is the same as the conventional Taylor correlation length. the first-order correlation area is a natural extension of Taylor's correlation length to the two-dimensional random field. the correlation length along the principal axis, which is independent of the definition of the coordinate axis, is also derived. Those features are based on the theory of statistics and are expected to be useful for the analysis of the actual random physical field and the texture analysis of the image.