For the purpose of realizing a task of clearing the table by a robot, a recognition and handling system of tableware is developed based on the “RECS” (Robot Environment Compromise System) concept. RECS concept makes it possible to develop the multipurpose robot by setting the environment so as that the robot can recognize and act skillfully to the object needed. As the setting of the environment, in this paper barcode marks are attached to tableware in order to simplify the image processing and realize the reliable and fast recognition of tableware. A position and orientation measuring method is proposed, which utilizes image data of four barcode marks and can calibrate camera parameters rapidly in every time when the camera position is changed. By using four marks, the reliability of reading barcode is also improved. A grasping hand is newly developed, which is composed of a simple link mechanism and is controlled easily by only driving the last joint of an articulated robot. This hand can cope with the change of configuration and size of the tableware and can allow the positioning error of an articulated robot, which is mainly caused by the image processing error. Experiments of recognizing and handling tableware are carried out and it is proved that this system is effective to a practical task of cleaning the table.

View full abstract

In this paper, we propose a method for the practical robot vision based on an approach referred to as “mark-based”. So far we developed 2D mark-based vision in which a mark is attached to every object in the environment to make object recognition simple and straightforward. In this method, object shape recognition is replaced with simpler mark recognition, and objects under planar motion are recognized in real-time. In order to get 3D information of objects, we developed 3D mark-based vision. To generalize the method to 3D environment, a combination of 3 marks and a barcode is introduced. A system of 3D mark-based vision is implemented and experimental results show the usefulness of the proposed method.

View full abstract

This paper deals with a task-assignment architecture for cooperative transport by multiple mobile robots in unknown environment. The architecture should satisfy three features: to deal with the variety of tasks in time and space; to deal with a large amount of tasks; and, to decide behavior in real-time. The authors propose the following approach: (1) We consider the unit of task (task instance) as the form that should be done in short time by one robot. Based on environmental information, task instances are dynamically generated using task templates; (2) Priority of task instances are evaluated dynamically based on the number of robots and the configuration in the work space. In addition, it is avoided to generate too many task instances by suppressing movement; (3) Main part of the architecture consists of two real-time planner: priority-based task-assignment planner solved by using a linear programming method and motion planners based on short-time estimation. Effectiveness of the proposed architecture is verified by a cooperative transport simulation in unknown environment.

View full abstract

This paper focuses on navigation systems for autonomous robots by using distributed map management systems, which are realized by putting “Intelligent Data Carrier (IDC)” in an environment. This paper provides the mathematical model describing the behavior of an autonomous robot based on a stochastic model and proposes the evaluaiton method for an effect of layout of the IDCs. The IDCs arrangement problem is formulated based on the mathematical model. This problem is combinatorial problem and difficult to solve. In order to overcome combinatorial complexity, the relaxation problem is introduced. This paper provides the design method for IDCs arrangement problem using the relaxed problem. A numerical example shows an optimal arrangement of IDCs can be obtained by using the provided method.

View full abstract

Much research has been done worldwide on support systems for visually impaired persons. There are still many problems for representing real-time information that changes around a user. In this paper, we intro-duce a visual support system that presents a three-dimensional visual information using three-dimensional virtual sound. Three-dimensional information is obtained from analyzing images captured by small stereo cameras, and objects which interest the user are recognized. The user can hear the three-dimensional virtual sounds, which correspond to the position and movement of objects using Head Related Transfer Functions (HRTFs). The user's auditory sense is not spoiled using a bone conduction headset, which dose not block environmental sound. The proposed system is useful in places where the infrastructure is incomplete, and when the situation changes in real-time. We plan to use it, for example, for walking assistance and sports. As experimental results, we found that there are many back and front recognition mistakes, and active operation is needed to recognize the sound position more correctly.

View full abstract

It is important to save the dissipated energy of manipulators for improving the environment of the earth which is being warmed up by CO₂ gas emitted from thermal power plants.
This paper describes an optimal trajectory which minimizes the dissipated energy in PTP motion of a 3-link manipulator. The two-point boundary-value problem is derived to get the optimal trajectory from Euler canonical equation. A generalized Newton-Raphson method is time-consuming to converge the optimal solution because the manipulator has many links and strong non-linearity. After using the method one time for short operating time which is liable to converge, a shooting method is applied to solve the optimal trajectory with various operating times. It is important for these iteration methods to select the starting velocity functions so as not to fall into a local optimal solution. In order to attain to a globally optimal path, this paper proposes a starting velocity function which the heavier link is accelerated toward gravitational direction and decelerated toward anti-gravitation.
It is shown from simulation that the proposed optimal path can lessen the dissipated energy in comparison to conventional path.

View full abstract

The paper discusses the problems in teleoperation systems for mobile robot in a multi-robot environment and the utilization of a virtual world in such systems. In order to achieve smooth operation of the mobile robot through the communication link, we should consider the data transfer in the systems. To compensate the incomplete data sets, virtual images can be generated by computer graphics when the information on the working environment can be acquired beforehand. In this paper, we construct a teleoperation system with the virtual world. Performance of the system is examined through the experiments with an actual mobile robot. And it is shown that the virtual robot can be operated by a human operator in almost the same manner as the teleoperated real robot

View full abstract

Wc analyzed the linear systemic properties of intracellular Calcium ion transport across the Sarcoplasmic Reticulum (SR) membrane in the cellular plasma which is characterized by multiple control inputs acting on multiple steps of the transport. The temporal behavior of the Ca transport system was expressed as thirteen rate equations. Linear systems analysis disclosed that the transport system is stable but uncontrollable. The singular values, H^∞ and H₂ norms of the system increased as the transport stage proceeded and the number of input site increased. System optimization reduced the singular values, H^∞ and H₂ norms of the system.

View full abstract

In this paper, we study a modified normalized least mean square (NLMS) algorithm for updating coef-ficients of an adaptive FIR digital filter (ADF). In the modified adaptive algorithm, filter coefficients are updated with NLMS algorithm for each sample, but the coefficients are freezed when the input signal is smaller than a given threshold (constant). This modified NLMS has been known, but not been analyzed.
In this paper, we call the modified NLMS as conditioned NLMS (C-NLMS) and analyze the convergence characteristics. As a result, the optimum threshold value is obtained. The simulation results and theo-retical analyses show the effectiveness of the C-NLMS with the proposed threshold, and a good agreement between both results. The stability of the NLMS algorithm in the presence of small input signal is improved. The convergence speed of the C-NLMS ADF under the noisy circumstances is also faster than that of the unconditioned ordinary NLMS ADF in small tap number case.

View full abstract

An attention is being given to the new system of distributing commodities for the core. It becomes possible to integrate some cores on a chip allowing free choice from the cores developed by various foundry in a quest of the core (mega-cell) standardization efforts. On such trends, virtual socket interface (VSI) alliance has been formed in 1996, and since then hopes run high that the new system of distributing commodities for the cores of large scale ASIC, including problems associated with the intellectual property (IP) of the cores, will be established in the not-too-distant future. In a VSI chip integrated from free choice of the cores developed by various foundry, we face much larger risks of the unauthorized duplication and usage. For this reason, a guarantee of the protected IP of cores should be offered.
In this paper, we propose a structure-protected IP of a core by public key cryptosystem focused as an information security techniques. The necessary structure of the multiplier/divider for the power residue arithmetic embedded in each cores is also essential for the arithmetic in what we call system LSI, so it won't be allowed to remove the arithmetic block and that assures a tight structure-protected IP of the core. Hardware only RSA encryption and its authentication system is also explained briefly.

View full abstract

This paper is concerned with the robust stabilization and robust L₂ disturbance attenuation of a class of cascaded nonlinear systems with structural uncertainties. The design tools are construction of a Lyapunov function for robust stabilization and a storage function for robust L₂ gain performance. The cascaded system consists of two subsystems called the x-subsystem and the ξ-subsystem. It is shown that if for each subsystem there exist Lyapunov functions satisfying Hamilton-Jacobi inequalities related with nominal subsystems, then a robust stabilizing controller can be constructed such that the sum of the functions becomes a Lyapunov function for the cascaded system with uncertainties. This concept is extended to the construction of the storage function and the controller for the robust L₂ disturbance attenuation problem by introducing an appropriate weight into the sum of the subsystem's Lyapunov functions. It is also shown that these results can be extended to solve robust design problems for the ξ-subsystem with a triangular structure.

View full abstract

In this paper, we study the numerical algorithm for solving coupled algebraic Riccati equations with γ. The main result in this paper is to propose a new algorithm for solving coupled algebraic Riccati equations by making use of a Lyapunov iteration. The proposed algorithm is of the reduced-order and can be implemented as a synchronous parallel algorithm. As a matter of fact, the proposed algorithm is based on the succes-sive approximations technique of dynamic programming. Under some assumption and the condition that guarantees a disturbance attenuation level γ which is larger than the maximum of the optimal disturbance attenuation level γ^*, the Lyapunov iterations are constructed such that the proposed algorithm converges to the positive semidefinite solution of the coupled algebraic Riccati equations. Furthermore, in order to show the effectiveness of the proposed algorithms, numerical examples are included.

View full abstract

This paper discusses dynamically controlled network segmentation for the purpose of adapting flexibly to unpredictably variable and unevenly distributed traffic and also utilizing network resources efficiently over radio frequencies, for application to a wireless ad-hoc community network (WACNet). This paper formulates the optimization criterion for the network segmentation as a graph partitioning problem and presents a heuristic computation method to obtain the optimum solution in a limited time range. The results computed by employing the heuristic method show that even if loss due to the segmentation is counted, a considerably high level of effects can be expected.

View full abstract

This paper deals with a new edge detection algorithm of color image by using RGB color histogram. From the investigation for edge characteristic based on the histogram, it became clear that the edge part has large color variations influenced by noise and is the minor area in the RGB color histogram. In order to detect this minor area stably we propose a new method that we can retrieve the low bit areas from the image constructed by summing up the multi-gray histogram images which are obtained from the RGB color histogram. The validity of proposed method is verified by some examples of application to real scenes.
The proposed method has advantages in that edge detection is not affected by gradient magnitude and features of curved surface can be selectively detected by variable detection parameters. Because of the low cost of calculation, we can easily implement this algorithm within an image frame accumulation process, performed before objective image-processing.

View full abstract

This paper presents a hierarchical compression method for 3-D medical image data with contour extrac-tion for improving compression performance. The 3-D image is regarded as a set of 2-D images, Snakes with complex AR model is applied to one of the 2-D images to extract the contour, and contour extraction in the neighboring images is efficiently done by propagating the one result to neighboring images. The contour data and the 3-D image data inside the extracted contour are coded by the DPCM method and a hierarchical coding method, respectively. The proposed compression method is applied to a 3-D medical image to show the effectiveness.

View full abstract

In real time communications on LAN such as FDDI utilizing the synchronous class of Timed Token Protocol (TTP), there exists a problem that the transmission delay time of synchronous data with Poisson arrivals at LAN nodes from upper layer largely increase in the range that the average load of synchronous data is close to the synchronous bandwidth. It is caused by the fact that TTP has not a function to process excess data that arrive at LAN nodes earlier than the timing determined with pre-assigned synchronous bandwidth
In this paper, in order to solve the problem, we propose another data transfer protocol “Advanced Timed Token Protocol (ATTP)” of which data transfer mechanism is modified from TTP and improves the transmission delay time of synchronous data including above mentioned excess data by rotating token more quickly on LAN than TTP. Furthermore we show the effectiveness of the protocol by simulations on the model of time critical communication for manufacturing control system.

View full abstract

Recently, the utilization of radio equipments along railways have been advanced rapidly in Japan. The importance of these radio systems rise in train control, information services and other communication systems. But the current frequency band width in UHF band assigned to Japanese railway companies is narrow to satisfy future various requirements for railway works and customer services. In this paper, we first present a construction of a next generation radio communication system using the microwave band radio between running trains and ground radio base station, as the system which can cope with the increase of the communication traffic volume in the railway communication. Secondly, we study the propagation characteristics on radio links to introduce this communication system into actual conventional lines. Concretely, we consider that the typical railway environments are distinguished to several types on analyses by complication of environments along a railway track. For each type, we calculate the propagation characteristics on a proper model. Also, we examine influences of rain which have an important effect on circuit quality. Computer simulation indicates that these analyses are effective to rationalize this communication system using the microwave band radio.

View full abstract

This paper deals with a method for a herd of mobile robots to self-organize information of a common coordinate system. In this method, it is assumed that each robot moves around autonomously having its own coordinate system and updating it using dead reckoninig. Encountering each other, two robots perform 1 to 1 bilateral proximity communication (called “conversation”) to exchange and modify their coordinate systems to be identical. After enough times of conversations, a common coordinate system is organized to be shared among all the robots. In this paper, we describe the scheme precisely and confirm its validity by theoretical analysis, computer simulations and experiments employing real mobile minirobots.

View full abstract

The simulated annealing is an optimization algorithm, by means of which near-optimal solutions can be searched in a solution space. The solution space is defined by a set of neighbor relations between solutions and a schedule of the temperature. Thus how to construct and determine them is a key to the success of application of the algorithm. However, it has not been studied in detail how to effectively analyze and eval-uate them. In this paper, we propose a new method to visualize a solution space explored in the algorithm. First, a transition probability between two solutions, which cannot be expressed in the existing methods, is considered. Its visual expression is also considered. Next, based on transition probabilities, a “space plot” can be obtained. It can reflect the behavior of the algorithm at a given temperature, and indicate the quality of a given set of neighbor relations. Several examples of space plots are also shown to demonstrate the effectiveness of the proposed method.

View full abstract

The aim of this paper is to apply the feature of lip motion in continuous image to personal identification. At first, four feature points are selected around the lip region. Next, the following are respectively computed: distance between the upper and lower and distance between the right and left. Finally, personal identification is carried out using the features that made by the measurements in continuous image.
By using the Mahalanobis distance, twenty persons are accurately distinguished. The experimental result indicates the effectiveness of the feature of lip motion to personal identification.

View full abstract

The effects of posture on the coordination between respiration and the onset of mastication were investi-gated in adult human subjects. When subjects sat on a chair with the torso upright and head facing forward (UP posture), the probability of the onset of mastication occurring in a particular time-bin (20% of the control respiratory cycle per bin) did not vary throughout the respiratory cycle. In contrast, in the DB posture (on hands and knees with the head facing down and back toward the knees), the onset of mastication occurred significantly more often near the transition phase from inspiration to expiration. These results suggest that the coordination between respiration and the onset of mastication is affected by posture.

View full abstract

A current-mode circuit to extract the feature of the target objects is proposed in this letter. From a color image with complex background, the proposed circuit can realize the feature extraction by using fuzzy scheme. Applying the fuzzy processing to RGB components of a color image, the proposed circuit can realize low sensitivity to changes of amount of light as well as noise of the background. Furthermore, thanks to the circuit design by using current-mode techniques, the proposed circuit can achieve low sensitivity to both temperature variations and supply voltage variations. The validity of the proposed algorithm and the circuit design is confirmed by computer simulations. The proposed pattern recognition circuit is integrable by a standard CMOS technology.

View full abstract

Parallel Simulated Annealing (PSA) is an effective method for some commbinatorial optimization problems. In this letter, We apply PSA to the graph partitioning problem and the maximum clique problem in a PC-cluster environment, and show the results of our simulations.

View full abstract