電気学会論文誌C(電子・情報・システム部門誌)
Online ISSN : 1348-8155
Print ISSN : 0385-4221
ISSN-L : 0385-4221
133 巻, 12 号
選択された号の論文の27件中1~27を表示しています
特集:電気関係学会東海支部連合大会
巻頭言
特集論文
<電気回路・電子回路>
<情報通信工学>
  • 松月 正樹, 竹田 馨, 細川 裕貴, 野呂 雄一, 竹尾 隆
    2013 年 133 巻 12 号 p. 2143-2148
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    In this study, the positioning accuracy of RFIDs used for managing medical instruments in a hospital by means of a wireless LAN is experimentally investigated. After a relationship between the positioning error and building materials or structures is examined, some methods of reducing the error, including a method utilizing a threshold setting technique, the Bayes' theorem and their combination are discussed, especially for electromagnetically degraded or shielded circumstances such as around a CT room. It has been made clear that the threshold setting and the Bayes' theorem methods are effective in reducing the positioning errors and one can obtain better results when both methods are simultaneously employed with the error being about 1 to 2 m.
  • 高 義礼, 加川 義久, 藤原 修
    2013 年 133 巻 12 号 p. 2149-2154
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    We previously measured discharge currents due to collision of the hand-held metal piece of a charged human through a 50 Ω SMA connector with a 6 GHz or 12 GHz digital oscilloscope, and then proposed an equivalent circuit model to calculate the discharge current. This model enables one to derive a discharge voltage across the spark gap from the measured discharge current, which revealed that at charge voltages bellow 600 V the estimated breakdown fields from the discharge voltages are almost kept constant (2-3×107 [V/m]) regardless of the different approach speeds of the metal piece. In this study, with a 12 GHz wide-band digital oscilloscope, we measured the discharge currents through a commercially available IEC calibration current target in order to estimate breakdown fields at different charge voltages from 200 V to 8,000 V. As a result, we confirmed that at charge voltages below 400 V, the breakdown fields show an almost constant value of (2-3)×107 V/m. It was also found that the breakdown fields at charge voltages from 600 V to 1,000 V decrease with charge voltages, and approximately agree with the Pachen's law, while at 2,000 V to 8,000 V they are smaller compared to the results calculated from the Pachen's law.
  • 鈴木 登季夫, 平田 晃正, 王 建青, 藤原 修
    2013 年 133 巻 12 号 p. 2155-2159
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    In this study, for the aim of validating whole-body average SARs in human volunteers for 0.1-2 GHz vertically polarized far-field exposure, we calculated with the FDTD method power absorption (product of whole-boy average SAR and weight) and square values averaged over height of layer induced currents in anatomical based human models for Japanese adults and children standing on a ground. As a result, we found that there exist strong correlations between them, and their regression coefficients have over 0.8 coefficient of determination at frequencies from 0.1 to 2 GHz, and are almost kept constant at frequencies from 300 MHz to 600 MHz with over 0.95 coefficient of determination. We also calculated the circumference integral of magnetic near-fields around a human-body model to reveal that the squares of layer induced currents can be estimated from the magnetic near-fields with relative errors of 25% from 300 MHz to 600 MHz and 1.5% at 100 MHz. This finding implies that the whole-body average SAR in a real human for vertically polarized far-field exposure could be estimated from the square values averaged over the height of layer induced currents. Experimental validation of this finding is a future subject.
<生体医工学・福祉工学>
  • 早田 有利, 河中 治樹, 小栗 宏次
    2013 年 133 巻 12 号 p. 2160-2166
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Many studies about estimating drivers' state while driving have been done so far. Now, people pay attention to not only estimation, but prediction of the drivers' state while driving in recent years. In this study, we proposed a method to predict the low arousal state early by using indexes which we could collect before the driver started to drive and also even the indexes which we could collect immediately after the driver started to drive. We simulated the before driving condition by control the drivers' sleeping hours. We used multiple linear regression to predict a time when the driver reaches the critical drowsiness level based on the biological information, vehicle operating information, and ability checks collected before and also while driving. As a result, our proposed method demonstrated the possibility to predict a time when the driver reaches the critical drowsiness level.
<システム・計測・制御>
特集研究開発レター
<情報通信工学>
論文
<電子物性・デバイス>
<情報通信工学>
  • Intan Sari Areni, Shinji Tsuzuki, Yoshio Yamada
    2013 年 133 巻 12 号 p. 2184-2189
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    The interference with existing radio broadcasting services because of unintentional radiation is a serious problem in power-line communication (PLC) systems. A radiation detection method based on burst packets, which had been proposed by the authors and was called scan and sum (SAS) method, was a low cost method and used to solve the problem. But its measurement took a long time and the burst packets occupied a channel. To shorten the measurement time, a new method based on periodic packets has been proposed in this paper, which is called periodic packet sound (PPS) method. Although both methods record the radiated signal as a sound data, the PPS does not need to scan the frequency of an AM radio receiver but analyses the sound data by a cross-correlation operation with a reference signal. First, how to generate the reference signal that depends on the modems is described. Second, an optimization of the packet size is studied, and finally it is shown that the detection sensitivity of PPS was maintained to be the same as SAS while the measurement time of PPS was 1/9 shorter than that of SAS. Moreover, the channel occupation of PPS was reduced to be 2% of SAS when the packet size was optimized to be 1500 bytes.
  • 熊谷 崇, 新熊 亮一, 久保 広行, 朝香 卓也, 高橋 達郎
    2013 年 133 巻 12 号 p. 2190-2199
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Nowadays a sharing of the UHF bandwidth used as terrestrial broadcasting by cognitive radio, is discussed. For sharing of the UHF bandwidth, cognitive radio terminals have to understand the surrounding radio environment. Then, the database technologies that collect the radio environment information from various sensing equipments are proposed. Nevertheless, their technologies have a drawback that the data-volume of database is enlargement. Accordingly, a generating and providing method of the minimum radio environment information for the terminals is effective. In this paper, we propose an available-spectrum information providing system for mobile terminals. The proposed system generates the Broadcasting Receiving List (BRL) for every area of the cell phone base stations using spectrum sensing results and base station ID in a database. In case that the terminals use TV white space, they carry out the spectrum sensing except the channels that described to the BRL, and use the clear channels. Therefore, this system can control to use the broadcast-wave in the cell, and keep the safety for broadcasting reception. Moreover, in this system, the terminals reduce the number of times of server-access using their time interval of the base station ID registration.
<生体医工学・福祉工学>
  • 吉川 一輝, 木下 史也, 松浦 康之, 高田 宗樹
    2013 年 133 巻 12 号 p. 2200-2204
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Stabilometry, which is useful for quantitative evaluation of equilibrium function, is generally performed maintaining Romberg's posture. However, equilibrium function deteriorates with advancing age, and most elderly individuals insist that it may be difficult to maintain the Romberg's posture for several seconds in stabilometry. The aim of this study was to compare the stability in the Romberg's posture with that in a posture in which the heels are together. Stabilograms of the young and the elderly were recorded both with heels together and in the Romberg's posture. The distribution of stabilograms in the 2 postures did not differ significantly, although analysis of stabilogram indices such as area of sway and total locus length suggested that the degree of postural sway with heels together was less than that with Romberg's posture. Therefore, there was no significant difference between the mathematical model of body sway with heels together and that with Romberg's posture.
  • 宇野 宏祐, 内藤 玄造, 飛佐 洋平, 吉田 塁, 小川 雄太郎, 小谷 潔, 神保 泰彦
    2013 年 133 巻 12 号 p. 2205-2210
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Brain-computer interface (BCI) has been studied so that handicapped people can communicate and control their environments. However, conventional BCIs are required to be decided what objects are displayed as target devices in advance. BCI combined with Augmented Reality (AR) can display suitable choices in consideration of real space and allows users to control devices intuitively. However, little is known how AR system affects accuracy of BCIs and major causes of the influence are considered as “the background of AR-BCI derived from scenery” and “the change of the number of target choices”. In this research we confirm these two effects, and develop the new method using the non-target object for the situation of few target choices. The results show the background has no effect, and the change of the target decreases the accuracy however our method improves the accuracy of the AR-BCI.
<システム・計測・制御>
<知能,ロボティクス>
  • 李 忠鍾, 寺田 信幸, 秋元 俊成
    2013 年 133 巻 12 号 p. 2243-2250
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    This paper presents a self-localization system using image feature points. Using image feature points in a self-localization system has its limitations in terms of processing time and angle variation. In order for the system to use image feature points, it needs to solve both the angle variation and processing time problems. The image feature points store three-dimensional information by using a stereo camera unit, which is constructed from a web camera. The web camera was used to reduce cost and improve general versatility, and a particle filter was used for improving the performance of the self-localization system. The system was constructed in two steps. First, a map was constructed using the image feature points. Second, a robot run by autonomous locomotion estimated its self-position using this map. The experiment was performed in environments both identical and not identical to the map environment. In the final result, the robot arrived at the goal, regardless of the environment condition.
<メディア情報,ユーザ・インタフェース>
  • 戸田 健, 塙 拓人, 陳 保旻, 藤多 和信, 井手口 直子
    2013 年 133 巻 12 号 p. 2251-2256
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Electronic health record (EHR) system is gradually coming into use at large hospitals, but is not at clinics. This is because its cost is too expensive and there is no merit for patients in the way that the patients check own EHR's. And doctors do not want the patients their health records. Then there is a need for a dermatological doctor to let patients see their affected part images in follow-up to check improvement of symptoms. In this paper, we propose a simple web-based database system of affected part images that both doctors and patients can see. In the prototype system, doctors can take images of affected parts of patients with tagging keywords such as patient's ID, patient's name, disease name, symptom, affected part name and free wards. The images with keyword tags are transmitted to the database. The system is implemented on smartphone and tablet for both doctors and patients quickly and easily use during medical examination.
  • 川崎 智久, 篠崎 隆宏, 古井 貞煕
    2013 年 133 巻 12 号 p. 2257-2263
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    With the recent progress in computer hardware and computer graphics (CG) techniques, applications using 3D virtual space are getting popular. So far, a mouse and a keyboard are generally used in these applications. While a mouse is a very successful input device for continuously controlling 2D objects, it is not necessarily intuitive for controlling 3D objects. In order to control 3D objects such as an avatar or a moving camera in a virtual space, speech interface has a potential to be a more natural and powerful alternative to a mouse. We propose speech based direct manipulation interface based on stretched word-end voice that controls continuous movements of 3D objects. By combining the proposed method with normal word based commands, both continuous movements and discrete actions are seamlessly controlled. Therefore, everything can be controlled using speech. The proposed method is implemented as an interface to the Second Life system. We compare it with a conventional speech based method that specifies start and end timing of motions. Analyses based on human subjects show that the proposed method is superior to the conventional speech based method. Moreover, we show that the best result is obtained when both methods are combined.
<音声画像処理・認識>
  • Shuang Bai, Tetsuya Matsumoto, Yoshinori Takeuchi, Hiroaki Kudo, Nobor ...
    2013 年 133 巻 12 号 p. 2264-2274
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Patch-level context can be used to relieve ambiguity of patch encoding in the bag of visual words method. However, in previous research patches of interest are often combined to their context in a fixed way. In this paper, we propose a novel approach in which the variation of the closeness of the relationship between a patch of interest and its context is taken into consideration, so that the context can be utilized more effectively. Specifically, we associate each patch of interest with other patches and take them as its context. In patch encoding, the patch of interest is first encoded by its N nearest visual words. Then, the encoding is adjusted based on its context by investigating whether its assignments to the visual words get support from the context patches. Furthermore, to explore the image content extensively, given an input image, we create three representations using different context strategies and fuse them in a probabilistic manner. Finally, we evaluated the proposed method on the datasets scene categories 8, scene categories 13 and scene categories 15, respectively. Experimental results demonstrated the effectiveness of the proposed method.
<情報処理・ソフトウェア>
  • 水野 友貴, 松本 啓之亮, 森 直樹
    2013 年 133 巻 12 号 p. 2275-2281
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Improving the reusability of models and automatic code generation efficiency are problems of model driven development. In order to solve these problems, we apply component-based development technologies which have been mainly developed on software implementing level to modeling level. In the proposed approach, functionally relevant model elements are packed as a component, and modeling software is carried out associating with the model components. The role of a model becomes clear by introducing the concept of component, and the reusability of model components is raised by externalizing the dependency between components. In addition, flexible model transformation rules united with the roles of models can be designed. As a result, it is able to generate more source codes automatically. The validity of the proposed approach has been checked by application experiments.
  • Huiquan Zhang, Sha Luo, Osamu Yoshie
    2013 年 133 巻 12 号 p. 2282-2290
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    Activity recognition, which has emerged as a pivotal research topic in pervasive sensing over the last several years, utilizes a collection of data from sensors to capture human behavior, detect anomalies and provide warning or guidance information. This paper presents an approach to explore appropriate clusters in subspace for human activity recognition. The approach includes two major phases: discovery of human activity (extraction of human behavior patterns and generation of human activity clusters), and recognition of human activity (application of similarity function to recognize activities). Different from many existing works, the proposed approach applies a subspace clustering based algorithm to generate clusters of human activity. This approach aims to accumulate human activity by approximating the generated clusters to the activity from a conceptual human perspective. The experiments were implemented using radio-frequency identification (RFID) based systems. The results show that the proposed approach is effective in improving the accuracy of both activity discovery and activity recognition.
資料
<ソフトコンピューティング・学習>
  • 中村 馨, 樋口 貞雄, 大沼 敏治
    2013 年 133 巻 12 号 p. 2291-2296
    発行日: 2013/12/01
    公開日: 2013/12/01
    ジャーナル フリー
    It should be not too much to say that crystal structure determines the various properties of material. However, crystal structure prediction before synthesis experiment has been the fundamental challenge in the field of materials science. Its difficulty originates in the infinite possibility of the crystal structure, including combinations of atomic positions, lattice constants and chemical composition. Although number of global optimization method has been proposed to overcome this difficulty, it is very recent that we can predict stable crystal structure only from chemical composition, thanks to the sophisticated state of art, evolutionary algorithm. In this paper, we will briefly introduce the application procedure of evolutionary algorithm to the crystal structure search to predict the stable crystal structure. By utilizing evolutionary algorithm combined with first-principles calculation, we successfully obtained crystal structures of high-pressure phase of KNbO3 and SrZnO2. Applicability of present method will be demonstrated by introducing two example cases, and further analysis will be shown to quantify the diversity of population and correlation between generated structures.
 
部門記事
 
feedback
Top