掲載種別：研究論文（学術雑誌）   出版者・発行元：ALife Robotics Corporation Ltd.  

DOI： 10.5954/icarob.2021.os11-3

researchmap

担当区分：責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

DOI： 10.1109/ICPR48806.2021.9412924

researchmap

その他リンク： https://dblp.uni-trier.de/db/conf/icpr/icpr2020.html#DaoudMMKMFD20

掲載種別：研究論文（学術雑誌）   出版者・発行元：Elsevier BV  

DOI： 10.1016/s0167-8140(21)01686-8

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）  

In this paper, we propose a method for predicting dose distribution images of patients with Nasopharyngeal carcinoma (NPC) from contoured computer tomography (CT) images. The proposed system is based on our previous method [1]. The first phase is to obtain the feature maps of 2D dose images of each beam from contoured CT images of a patient by convolutional deep neural network model. In the second phase, dose distribution images are predicted from the obtained feature maps by the integration network. Our modified system predicted dose distribution images accurately. From the experimental results using 80 NPC patients' images, the average number of pixels that satisfy the dose constraints of tumors and OARs regions is 81.9 % and 86.1 %, respectively. The proposed system had a global 3D gamma passing rates varying from 82.1 % to 97.2 % for all regions and an overall mean absolute errors (MAEs) was 1.0 ±1.2. From the obtained results, our modified system is superior to the results obtained in our previous system results and conventional methods. Contribution-The use of the predicted 7-beam weights, as input, into our CNN network leads to improve the predicted dose distribution. Contribution-The use of the predicted 7-beam weights, as input, into our CNN network leads to improve the predicted dose distribution.

DOI： 10.1109/ICIEVicIVPR48672.2020.9306610

Scopus

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：INDERSCIENCE ENTERPRISES LTD  

This paper presents an automatic segmentation technique for identifying nasopharyngeal carcinoma regions in CT images. The proposed technique is based on the region growing method by which an initial seed is automatically generated. A probabilistic map representing a chance of being the tumor pixel in each CT image will be created and used for initial seed determination. This map is generated from three probabilistic functions established upon location of the tumor considered, intensities of the tumor pixels, and asymmetry of organs respectively. A representative of potential tumor pixels will be selected as an initial seed. The experimental results showed that seeds were correctly determined with the percent accuracy of 84.32%. These seeds were grown in preprocessed CT images for identifying the nasopharyngeal carcinoma regions subsequently. The results showed that, for no metastasis cases, perfect match and corresponding ratio were 85.03% and 52.44% respectively and 29.26% and 28.03% correspondingly for metastasis cases. This resulted from a single seed generated in each CT image. It was unable to identify more than one tumor region.

Web of Science

researchmap

担当区分：責任著者   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：SPIE  

DOI： 10.1117/12.2519819

researchmap

担当区分：筆頭著者   記述言語：英語   掲載種別：研究論文（学術雑誌）  

DOI： 10.1007/s00138-018-0972-5

Web of Science

researchmap

掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Springer  

DOI： 10.1007/978-3-030-39770-8_6

researchmap

掲載種別：研究論文（学術雑誌）  

DOI： 10.1016/j.compmedimag.2019.101644

researchmap

掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Springer  

DOI： 10.1007/978-3-030-39770-8_8

researchmap

掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Springer  

DOI： 10.1007/978-3-030-32486-5_16

researchmap

掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Springer  

We propose a new method for constructing a spatiotemporal statistical model of the distribution of anatomical landmarks (LMs) of a human embryo. This method exhibits potential for the quantitative assessment of the extent of anomalies and is important in the research of congenital malformations. However, a few of the LMs might not be observed at a specific developmental stage because large morphological deformations exist during the early stages of development. It is difficult for conventional statistical shape analysis methods to handle missing LMs in the training dataset. The basic concept of the proposed method is to conduct statistical analyses by predicting and completing the coordinates of the missing LMs. We demonstrated the proposed method in the context of spatiotemporal statistical modeling of 10 LMs on the brain surface using 37 embryonic subjects with Carnegie stages of 19–22. We conducted a comparative study of the spatiotemporal statistical models between four different prediction methods, and we found that deformable surface mapping was the best prediction method in terms of model generalization and specificity.

DOI： 10.1007/978-3-030-32689-0_10

Scopus

researchmap

その他リンク： https://dblp.uni-trier.de/db/conf/miccai/clip2019.html#ShinjoSTYHMMMS19

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）  

© 2018 TSI Press. This paper presents a new method for generating object mesh models composed of hexahedra. The proposed method is based on a self-organizing deformable model (SDM) which is a deformable surface model guided by competitive learning and an energy minimization approach. Extending the SDM, the proposed method generates a hexahedral mesh model of a target object by fitting a cuboid composed of rectangular voxels to the object. Moreover, the shape of each hexahedron in the model is corrected by dividing the hexahedron into sub-hexahedra and moving the nodes of the hexahedron. From our experimental results, the proposed method obtains the hexahedral mesh model which consists of many regular hexahedra while recovering the shape of the object.

DOI： 10.23919/WAC.2018.8430386

Web of Science

Scopus

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：Elsevier {BV}  

© 2018 Elsevier B.V. Background and Objective: This paper proposes a new method for mapping surface models of human organs onto target surfaces with the same genus as the organs. Methods: In the proposed method, called modified Self-organizing Deformable Model (mSDM), the mapping problem is formulated as the minimization of an objective function which is defined as the weighted linear combination of four energy functions: model fitness, foldover-free, landmark mapping accuracy, and geometrical feature preservation. Further, we extend mSDM to speed up its processes, and call it Fast mSDM. Results: From the mapping results of various organ models with different number of holes, it is observed that Fast mSDM can map the organ models onto their target surfaces efficiently and stably without foldovers while preserving geometrical features. Conclusions: Fast mSDM can map the organ model onto the target surface efficiently and stably, and is applicable to medical applications including Statistical Shape Model.

DOI： 10.1016/j.cmpb.2018.01.028

Web of Science

Scopus

PubMed

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

We propose a new method for analyzing human grasping motion using an infrared camera. This technique simplifies the teaching of motion to robots based on the observation of human motion. In this method, the contact area on the object is extracted by observing the thermal signature captured by an infrared camera. To understand the intention of human behavior, we propose a grasping identification method using 3D thermal signatures. In addition, we perform a principal component analysis on the contact area and the center of gravity for the contact area of each finger. This method expresses the grasp motion space with a small number of parameters and can be used to enable easy correspondence between human and robot hands. We confirm experimentally that the proposed method is effective for teaching motion to robot.

Web of Science

researchmap

記述言語：日本語   出版者・発行元：日本医用画像工学会  

手術手技訓練シミュレーターや手術ナビゲーションなどの手術支援システムの開発には，臓器の三次元形状モデルによる臓器変形の推定が必要不可欠である．臓器変形の推定には有限要素解析がしばしば用いられるが，これには膨大な計算量を必要とするため，臓器変形の効率的な推定には向いていない．この問題を解決するため，著者らは，有限要素解析による三次元臓器変形推定を深層なニューラルネットで効率的に代替する方法の研究を行っている．本稿では，有限要素解析を模した深層なニューラルネットによる臓器変形の実時間推定の研究例を示しつつ，ディープラーニング手術支援システムへの応用可能性について述べる．

DOI： 10.11409/mit.35.206

CiNii Article

researchmap

担当区分：責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE COMPUTER SOC  

This paper proposes a new method for mapping volume models of human organs onto a target volume with simple shapes. The proposed method is based on our modified Self-organizing Deformable Model (mSDM) which finds the one-to-one mapping with no foldovers between an arbitrary object surface model and a target surface. By extending mSDM to apply to organ volume models, the proposed method, called volumetric SDM (vSDM), establishes the one-to-one correspondence between the volume model and its target volume. At the same time, vSDM preserves geometrical properties of the original model before and after the mapping. In addition, vSDM allows to control the mapping of interior structures of the organ model onto specific regions inside the target volume. These characteristics of vSDM enables to easily find a reliable correspondence between different volume models via a common target volume.

DOI： 10.1109/ICPR.2016.7899963

Web of Science

Scopus

researchmap

担当区分：責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）  

This paper proposes a new method for describing parenchymatous organs by the set of volumetric primitives with simple shapes. The proposed method is based on our modified Self-organizing Deformable Model (mSDM) which maps an object surface model onto a target surface with no foldovers. By extending mSDM to apply to organ volume models, the proposed method, volumetric SDM (vSDM), finds the one-to-one correspondence between the volume model and its target volume. During the mapping, vSDM preserves geometrical properties of the original model while mapping internal structures of the model onto their corresponding primitives inside of the target volume. Owing to these characteristics, vSDM enables to obtain a new volume representation of organ volume models which simultaneously (1) represents by simple primitives the shapes of the whole organ and its internal structures and (2) describes the relationship among the external surface and internal structures of the organ.

DOI： 10.1007/978-3-319-51237-2_4

Scopus

researchmap

担当区分：責任著者   記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：SPIE  

DOI： 10.1117/12.2217367

Scopus

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：ELSEVIER  

Daily life assistance is one of the most important applications for service robots. For comfortable assistance, service robots must recognize the surrounding conditions correctly, including human motion, the position of objects, and obstacles. However, since the everyday environment is complex and unpredictable, it is almost impossible to sense all of the necessary information using only a robot and sensors attached to it. In order to realize a service robot for daily life assistance, we have been developing an informationally structured environment using distributed sensors embedded in the environment. The present paper introduces a service robot system with an informationally structured environment referred to the ROS-TMS. This system enables the integration of various data from distributed sensors, as well as storage of these data in an on-line database and the planning of the service motion of a robot using real-time information about the surroundings. In addition, we discuss experiments such as detection and fetch-and-give tasks using the developed real environment and robot. (C) 2015 The Authors. Published by Elsevier B.V.

DOI： 10.1016/j.robot.2015.07.010

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）  

DOI： 10.3390/s150409438

Web of Science

Scopus

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

Daily life assistance for elderly individuals in hospitals and care facilities is one of the most urgent and promising applications for service robots. Especially, a fetchand- give task is a frequent and fundamental task for service robots to assist elderlys daily life. In hospitals and care facilities, this task is often performed with a movable platform such as a wagon or a cart to carry and deliver a large amount of objects at once. Thus the navigation and control of not only a service robot but also a movable platform must be planned safely. In addition, a robot motion planning to hand over an object to a person safely and comfortably according to his/her posture is also an important problem in this task, however this has not been discussed so much. In this work, we present a coordinate motion planning technique for a fetch-and-give task using a wagon and a service robot. Handover motion is also planned by considering the manipulability of both a robot and a person. Experiments of a fetch-and-give task using a service robot are successfully carried out.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

DOI： 10.1109/IROS.2015.7353688

Web of Science

Scopus

researchmap

掲載種別：研究論文（学術雑誌）  

This paper proposes a parameterization method of vertebra models by mapping them onto the parameterized surface of a torus. Our method is based on a modified Self-organizing Deformable Model (mSDM) [1], which is a deformable model guided by competitive learning and an energy minimization approach. Unlike conventional mapping methods, the mSDM finds the one-to-one mapping between arbitrary surface model and the target surface with the same genus as the model. At the same time, the mSDM can preserve geometrical properties of the original model before and after mapping. Moreover, users are able to control mapping positions of the feature vertices in the model. Using the mSDM, the proposed method maps the vertebra model onto a torus surface through an intermediate surface with the approximated shape of the vertebra. The use of the intermediate surface results in the stable mapping of the vertebra to a torus compared with the direct mapping from the model to the torus.

DOI： 10.1007/978-3-319-14148-0_16

Scopus

researchmap

掲載種別：研究論文（国際会議プロシーディングス）  

This paper proposes a parameterization method of vertebra models by mapping them onto the parameterized surface of a torus. Our method is based on a modified Self-organizing Deformable Model (mSDM) [1], which is a deformable model guided by competitive learning and an energy minimization approach. Unlike conventional mapping methods, the mSDM finds the one-to-one mapping between arbitrary surface model and the target surface with the same genus as the model. At the same time, the mSDM can preserve geometrical properties of the original model before and after mapping. Moreover, users are able to control mapping positions of the feature vertices in the model. Using the mSDM, the proposed method maps the vertebra model onto a torus surface through an intermediate surface with the approximated shape of the vertebra. The use of the intermediate surface results in the stable mapping of the vertebra to a torus compared with the direct mapping from the model to the torus.

DOI： 10.1007/978-3-319-14148-0_16

Scopus

researchmap

掲載種別：研究論文（国際会議プロシーディングス）  

This paper proposes a method to identify the spatial distribution patterns of cone cells related with blood vessel in a given retina image. We define three types of the distribution patterns between cones and vessels. Positive correlation distribution (PCD) and negative correlation distribution (NCD) indicate that the cones tend to be close to or far from the vessels. While the cone cells do not have significant correlation with vessels, the cone distribution is regarded as the random distribution (RD). In our method, RD is modeled by many virtual retina images, each of which is generated by the vessels extracted from the original retina image and the virtual cells are selected randomly from the image. Using the virtual images, we estimate the distribution range of RD. When the distribution of the original cells is above the upper limit or below the lower limit of the RD distribution, the cell distribution is NCD or PCD. Otherwise, the cell distribution is regarded as RD.

DOI： 10.1109/WAC.2014.6936144

Scopus

researchmap

記述言語：英語   出版者・発行元：一般社団法人電子情報通信学会  

The digital human modeling is one of the challenging topics in the last few decades. We have been studying on the digital brain atlas of Japanese. Our system for generating the brain atlas uses the latest techniques in both human anatomy and computer science. This paper reports the detail of generating the brain model by the 3D contour models of the brain surface and internal neural structures. Especially, we propose a new method for registering two sub-block models of the brain by combining a nonrigid registration and dynamic programming.

CiNii Article

CiNii Books

researchmap

担当区分：筆頭著者,　責任著者   記述言語：英語   掲載種別：研究論文（研究会，シンポジウム資料等）  

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：MDPI AG  

This paper describes a new method of measuring the position of everyday objects and a robot on the floor using distance and reflectance acquired by laser range finder (LRF). The information obtained by this method is important for a service robot working in a human daily life environment. Our method uses only one LRF together with a mirror installed on the wall. Moreover, since the area of sensing is limited to a LRF scanning plane parallel to the floor and just a few centimeters above the floor, the scanning covers the whole room with minimal invasion of privacy of a resident, and occlusion problem is mitigated by using mirror. We use the reflection intensity and position information obtained from the target surface. Although it is not possible to identify all objects by additionally using reflection values, it would be easier to identify unknown objects if we can eliminate easily identifiable objects by reflectance. In addition, we propose a method for measuring the robot's pose using the tag which has the encoded reflection pattern optically identified by the LRF. Our experimental results validate the effectiveness of the proposed method.

DOI： 10.3390/s140407524

Web of Science

Scopus

PubMed

researchmap

その他リンク： http://orcid.org/0000-0002-4219-7644

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

We propose a grasp stability index which takes into account the elastic deformation of fingertips. In this index, an evaluation value of grasp stability is derived as the minimum energy which causes slippage of a fingertip on a grasped object surface. Elastic potential energy of the fingertips and gravitational potential energy of the object are considered in this index. It is ensured that any fingertip does not slip on a grasped object surface if the external energy added to the object is less than the evaluation value. This index has good features as follows. (i) Contact forces corresponding to the grasping state are taken into account. (ii) It is possible to derive a condition of stable grasp including the kinetic energy of a grasped object. The effectiveness of this index is verified through several numerical examples.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

This paper presents a grasp planner which allows a robot to grasp the constricted parts of objects in our daily life. Even though constricted parts can be grasped more firmly than convex parts, previous planners have not sufficiently focused on grasping this part. We develop techniques for quadric surface approximation, grasp posture generation, and stability evaluation for grasping constricted parts. By modeling an object into multiple quadric surfaces, the planner generates a grasping posture by selecting one-sheet hyperbolic surfaces or two adjacent ellipsoids as constricted parts. When a grasping posture being generated, the grasp stability is evaluated based on the distribution of the stress applied to an object by the fingers. We perform several simulations and experiments to verify the effectiveness of our proposed method.

DOI： 10.1109/IROS.2014.6942895

Web of Science

researchmap

出版者・発行元：IEEE  

DOI： 10.1109/EMBC.2013.6610170

PubMed

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Springer  

This paper proposes a method for estimating patient-specific material parameters used in the finite element analysis which simulates soft tissue deformation. The estimation of suitable material parameters for a patient is important for a navigation system for endoscopic surgery. At first, many data of soft tissue deformation are generated by changing the material parameters. Next, using Principle Component Analysis, each data with high dimensional is converted into the lower vector. The relationship between the material parameter and the deformation is found in the lower potential space. © 2013 Springer-Verlag.

DOI： 10.1007/978-3-642-38637-4_12

Scopus

researchmap

掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：Springer  

DOI： 10.1007/978-3-642-38637-4_11

Scopus

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

This paper describes a method of measurement and estimation of human behaviors in a room together with the layout of objects on the floor. The information obtained by the method is essential for a service robot working in a human daily life environment. The method uses only one laser range finder (LRF) installed in the room and a strip of mirror attached to a side wall close to a floor. The area of sensing is limited to a plane parallel to and just a few centimeters above the floor, thus covering the whole room with minimal invasion of privacy of a resident while reducing occlusion. Processing both distance and reflectance acquired by the LRF from the surface of the existing objects allows us to exclude immediately distinguishable clusters and to focus on the analysis of remaining clusters. The human behavior models that we propose are effectively used to estimate human behavior based on the limited LRF data. Our experimental results validate the effectiveness of the proposed method.

DOI： 10.1109/ROBIO.2013.6739791

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

This paper proposes a new method for mapping a tissue surface model onto an arbitrary target surface while preserving the geometrical features of the tissue surface. In our method, firstly, the tissue model is roughly deformed by using Self-organizing Deformable Model. Since the deformed model may contain folded patches, the folded patches are removed. Moreover, by Free-Form Deformation (FTD), and the area- and angle-preserving mapping, the model is mapped onto the target surface while preserving geometrical properties of the original model. From several experimental results, we can conclude that the proposed method can map tissue models onto arbitrary target surface without foldovers.

DOI： 10.1109/EST.2013.34

Web of Science

Scopus

researchmap

掲載種別：研究論文（学術雑誌）  

To estimate the internal structures of the patient brain, several methods map a brain atlas to a patient brain shape using some landmarks, which are selected manually from the brain shape. The determination of the correspondence between small sulci of arbitrary two brains is complex. Moreover, the relationship between the surface shape and internal structure of the brain is unclear. Therefore, even if the surface the deformed atlas is fitted to the patient brain shape, the accurate internal structures of the brain can not always be obtained. To solve these problems, this paper proposes a method of selecting landmarks to estimate the internal structures of the patient brain by deforming brain atlas. Firstly, the brain shapes are represented by a simple shape. Secondly, a small number of initial landmarks are selected manually from the contours of the approximated brain shape and the internal structures which can be identified from MR images. Thirdly, some new landmarks are generated automatically on the contours based on the initial landmarks. Finally, the brain atlas is deformed non-rigidly using these landmarks. To evaluate our method, we estimate the patient brain structure using 4 methods. From the experimental results using 10 brain images, our method can estimate the patient brain structure reliably and stably using a small number of landmarks compared with the method using the landmarks on the original brain shape.

Scopus

researchmap

掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：SPIE  

DOI： 10.1117/12.911171

Web of Science

Scopus

researchmap

掲載種別：研究論文（国際会議プロシーディングス）  

Daily life assistance for elderly people is one of the most promising scenarios for service robots in the near future. In particular, the go-and-fetch task will be one of the most demanding tasks in these cases. In this paper, we present an informationally structured room that supports a service robot in the task of daily object fetching. Our environment contains different distributed sensors including a floor sensing system and several intelligent cabinets. Sensor information is sent to a centralized management system which processes the data and makes it available to a service robot which is assisting people in the room. We additionally present the first steps of an intelligent framework used to maintain information about locations of commodities in our informationally structured room. This information will be used by the service robot to find objects under request. One of the main goal of our intelligent environment is to maintain a small number of sensors to avoid interfering with the daily activity of people and to reduce as much as possible the invasion of their privacy. In order to compensate this limited available sensor information our framework aims to exploit knowledge about people's activity and their interaction with objects to infer reliable information about the location of commodities. This paper presents simulated results that demonstrate the suitability of this framework to be applied to a service robotic environment equipped with limited sensors. In addition we discuss some preliminary experiments using our real environment and robot. © 2012 IFAC.

DOI： 10.3182/20120905-3-HR-2030.00046

Scopus

researchmap

記述言語：日本語   出版者・発行元：日本定位・機能神経外科学会  

三次元的整合性と細胞構築を両立させる、より簡便なヒト脳座標アトラスの作成法について検討した。日本人の献体脳(89歳男性)を用い、非接触型3次元デジタイザーにより脳全体の初期形状を記録した後、振動刃ミクロトームにより100μmの切片を作成した。その組織切片を染色した後、スキャナで電子化し、3Dニューロン再構築解析ソフトウェアを用いて再構築した。この方法は標本作製に伴う機械的変形が全くみられず、三次元的整合性と細胞構築を両立させ、より簡便にヒト脳アトラスを作成できた。

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：ELSEVIER IRELAND LTD  

Research on the human brain has undoubted significance, but our knowledge on its detailed morphology is still limited. We have developed a simple method for reconstruction of large-sized brain tissues of the human. Fixed brains were cut into blocks (maximum size 7 cm x 7 cm x 1 cm), embedded and post-fixed in gelatin just one overnight before obtaining complete serial sections with a vibrating microtome. Quality of stained materials was sufficient to create three-dimensional histological maps, where digital reconstructions from adjoining blocks could be accurately combined. The present method will facilitate both direct examination of the human brain and construction of its histological database. (C) 2010 Elsevier Ireland Ltd and the Japan Neuroscience Society. All rights reserved.

DOI： 10.1016/j.neures.2010.03.005

Web of Science

PubMed

CiNii Article

researchmap

記述言語：日本語   出版者・発行元：日本定位・機能神経外科学会  

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

本論文では,ニューラルネットワークを用いて,軟性臓器モデルの変形をシミュレートする新たな手法を提案する.提案手法は,基本的なモデルの変形(以後,変形モードと呼ぶ)の組合せに基づいて,モデルの変形を推定する.つまり,変形モードをあらかじめ非線形有限要素法で求め,臓器に加わった外力と,それに対応する変形モードの関係をニューラルネットワークで学習する.学習したニューラルネットワークは,非線形有限要素解析によりモデルの振舞いを推定することを模倣する.実験結果より,提案手法は,非線形有限要素解析とほぼ同程度の精度を保ちつつ,計算コストを大幅に削減することができた.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：The Robotics Society of Japan  

This paper describes the on-demand patient service system which uses RFID tags and a wheelchair robot in medical facilities. In this system, the robot can identify a patient and obtain his/her diagnosis schedule by the RFID tag in his/her registration card. According to the schedule, the robot can carry the patient to his/her target examination room. To construct this system, we employ the idea of the intelligent environment, called "Robot Town". Many sensors are distributed in Robot Town, and robots can easily move and work by using the sensors. We executed a proving experiment with our system under the scenario that assumed an outpatient came to the unfamiliar and large medical facility. In addition, to evaluate our system, we executed the questionnaire survey to nurses. As the result of survey, we know that their expected points are an improvement of the daily work efficiency, safe transportation of patients and an improvement of an amenity level in a hospital. And, we also know that their worried points are safety of the robots, the abilities of detection to the unexpected emergency situations, costs and difficulties of maintenance of systems, and responsibilities in case of accidents.

DOI： 10.7210/jrsj.27.877

CiNii Article

CiNii Books

researchmap

その他リンク： https://jlc.jst.go.jp/DN/JALC/00339803556?from=CiNii

記述言語：英語   掲載種別：研究論文（学術雑誌）  

BACKGROUND: To date, several training and evaluation systems for endoscopic surgery have been developed, such as virtual-reality simulators and box trainers. However, despite current advances in these objective assessments, no functional brain studies during learning of endoscopic surgical skills have been carried out. In the present study, we investigated cortical activation using near-infrared spectroscopy (NIRS) during endoscopic surgical tasks. STUDY DESIGN: A total of 21 right-handed subjects, comprising 4 surgical experts, 4 trainees, and 13 novices, participated in the study. Suturing and knot-tying tasks were performed in a box trainer. Cortical activation was assessed in all subjects by task-related changes in hemoglobin (Hb) oxygenation using NIRS. RESULTS: In surgical experts and novices with no experience of endoscopic surgical training, we found no changes in oxy-Hb, deoxy-Hb or total-Hb levels in any of the frontal channels. In surgical trainees and one novice with experience of endoscopic surgical training, we found significant increases in oxy-Hb and total-Hb levels in most of the frontal channels. There were significant differences in oxy-Hb and total-Hb levels in CH-19 between surgical experts and trainees (p = 0.02 for both), and between surgical trainees and novices with no experience of endoscopic surgical training (p = 0.008 for both). Furthermore, additional training increased oxy-Hb levels in the frontal cortex of novices with no experience of endoscopic surgical training but had no such effect on surgical experts. CONCLUSIONS: The present data suggest that NIRS is a feasible tool for assessing brain activation during endoscopic surgical tasks, and may have a large impact on the future development of teaching, training, and assessment methods for endoscopic surgical skills.

DOI： 10.1007/s00464-008-0316-z

PubMed

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：SPRINGER  

Background To date, several training and evaluation systems for endoscopic surgery have been developed, such as virtual-reality simulators and box trainers. However, despite current advances in these objective assessments, no functional brain studies during learning of endoscopic surgical skills have been carried out. In the present study, we investigated cortical activation using near-infrared spectroscopy (NIRS) during endoscopic surgical tasks.Study design A total of 21 right-handed subjects, comprising 4 surgical experts, 4 trainees, and 13 novices, participated in the study. Suturing and knot-tying tasks were performed in a box trainer. Cortical activation was assessed in all subjects by task-related changes in hemoglobin (Hb) oxygenation using NIRS.Results In surgical experts and novices with no experience of endoscopic surgical training, we found no changes in oxy-Hb, deoxy-Hb or total-Hb levels in any of the frontal channels. In surgical trainees and one novice with experience of endoscopic surgical training, we found significant increases in oxy-Hb and total-Hb levels in most of the frontal channels. There were significant differences in oxy-Hb and total-Hb levels in CH-19 between surgical experts and trainees (p = 0.02 for both), and between surgical trainees and novices with no experience of endoscopic surgical training (p = 0.008 for both). Furthermore, additional training increased oxy-Hb levels in the frontal cortex of novices with no experience of endoscopic surgical training but had no such effect on surgical experts.Conclusions The present data suggest that NIRS is a feasible tool for assessing brain activation during endoscopic surgical tasks, and may have a large impact on the future development of teaching, training, and assessment methods for endoscopic surgical skills.

DOI： 10.1007/s00464-008-0316-z

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）  

BACKGROUND: Laparoscopic surgeons require extended experience of cases to overcome the lack of depth perception on a two-dimensional (2D) display. Although a three-dimensional (3D) display was reported to be useful over two decades ago, 3D systems have not been widely used. Recently, we developed a novel 3D dome-shaped display (3DD) system, CyberDome. STUDY DESIGN: In the present study, a total of 23 students volunteered. We evaluated the effects of the 3DD system on depth perception and laparoscopic procedures in comparison with the 2D, a conventional 3D (3DP) or the 2D high definition (HD) systems using seven tasks. RESULTS: The 3DD system significantly improved depth perception and laparoscopic performance compared with the 2D system in six new tasks. We further found that the 3DD system shortened the execution time and reduced the number of errors during suturing and knot tying. The 3DD system also provided more depth perception than the 3DP and 2D HD systems. CONCLUSIONS: The novel 3DD system is a promising tool for providing depth perception with high resolution to laparoscopic surgeons.

DOI： 10.1007/s11548-009-0282-5

PubMed

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：SPRINGER HEIDELBERG  

Background Laparoscopic surgeons require extended experience of cases to overcome the lack of depth perception on a two-dimensional (2D) display. Although a three-dimensional (3D) display was reported to be useful over two decades ago, 3D systems have not been widely used. Recently, we developed a novel 3D dome-shaped display (3DD) system, CyberDome.Study design In the present study, a total of 23 students volunteered. We evaluated the effects of the 3DD system on depth perception and laparoscopic procedures in comparison with the 2D, a conventional 3D (3DP) or the 2D high definition (HD) systems using seven tasks.Results The 3DD system significantly improved depth perception and laparoscopic performance compared with the 2D system in six new tasks. We further found that the 3DD system shortened the execution time and reduced the number of errors during suturing and knot tying. The 3DD system also provided more depth perception than the 3DP and 2D HD systems.Conclusions The novel 3DD system is a promising tool for providing depth perception with high resolution to laparoscopic surgeons.

DOI： 10.1007/s11548-009-0282-5

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：SPRINGER-VERLAG BERLIN  

This paper presents a new method for simulating the deformation of organ models by using a neural network. The proposed method is based on the idea proposed by Chen et al. [2] that deformed model can be estimated from the superposition of basic deformation modes. The neural network finds a relationship between external forces and the models deformed by the forces. The experimental results show that the trained network can achieve a real-time simulation while keeping the acceptable accuracy compared with the nonlinear FEM computation.

Web of Science

PubMed

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

This paper presents a parallel algorithm of the Level Set Method named the Parallel Fast Level Set Method, and its application for real-time 3D reconstruction of human shape and motion. The Fast Level Set Method is an efficient implementation algorithm of the Level Set Method and has been applied to several applications such as object tracking in video images and 3D shape reconstruction using multiple stereo cameras. In this paper, we implement the Fast Level Set Method on a PC cluster and develop a real-time motion capture system for arbitrary viewpoint image synthesis. To obtain high performance on a PC cluster, efficient load-balancing and resource allocation algorithms are crucial problems. We develop a novel optimization technique of load distribution based on the estimation of moving direction of object boundaries. In this technique, the boundary motion is estimated in the framework of the Fast Level Set Method, and the optimum load distribution is predicted and performed according to the estimated boundary motion and the current load balance. Experiments of human shape reconstruction and arbitrary viewpoint image synthesis using the proposed system are successfully carried out.

DOI： 10.1109/ROBOT.2008.4543332

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE COMPUTER SOC  

This paper presents a new technique for projecting a 3D object mesh model onto a surface of another target object. The mesh model adapts its shape to the target surface, and is called Self-organizing Deformable Model(SDM). The SDM algorithm works by combining a competitive learning and an energy minimization. The framework of the SDM makes it possible to map a mesh model onto various kinds of target surfaces. This characteristic can not be seen in other methods for surface parameterization, and it enables us to apply the SDM to some different fields in computer vision and computer graphics. Also the SDM can reconstruct shapes of target objects similar to general deformable models.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

This paper proposes a new method for integrating range images with different resolutions to generate a whole surface model of a 3D object. Generally, a modeling method using range images needs an overlapping area between the images for the registration process. Accordingly, there are redundant data in the overlapping area. In order to produce a seamless and non-redundant model, we integrate the data, and represent the overlapping area with triangular patches. However, few conventional methods have paid attention to integrating images with different resolutions. We propose a new integration method in which geometric and topological information are used. Experimental results show that our proposed method is efficient for integrating multiple range images.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG  

As a representative of the linear discriminant analysis, the Fisher method is most widely used in practice and it is very effective in two-class classification. However, when it is expanded to a multi-class classification problem, the precision of its discrimination may become worse. A main reason is an occurrence of overlapped distributions on the discriminant space built by Fisher criterion. In order to take such overlaps among classes into consideration, our approach builds a new discriminant space by hierarchically classifying the overlapped classes. In this paper, we propose a new hierarchical discriminant analysis for texture classification. We divide the discriminant space into subspaces by recursively grouping the overlapped classes. In the experiment, texture images from many classes are classified based on the proposed method. We show the outstanding result compared with the conventional Fisher method.

DOI： 10.1093/ietisy/e88-d.10.2380

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：SPRINGER-VERLAG BERLIN  

This paper presents a new method for projecting a mesh model of a source object onto a surface of an arbitrary target object. A deformable model, called Self-organizing Deformable Model(SDM), is deformed so that the shape of the model is fitted to the target object. We introduce an idea of combining a competitive learning and an energy minimization into the SDM deformation. Our method is a powerful tool in the areas of computer vision and computer graphics. For example, it enables to map mesh models onto various kinds of target surfaces like other methods for a surface parameterization, which have focused on specified target surface. Also the SDM can reconstruct shapes of target objects like general deformable models.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：SPRINGER-VERLAG BERLIN  

We have developed a new framework for scale invariant texture analysis using multi-scale local autocorrelation features. The multiscale features are made of concatenated feature vectors of different scales, which are calculated from higher-order local autocorrelation functions. To classify different types of textures among the given test images, a linear discriminant classifier (LDA) is employed in the multi-scale feature space. The scale rate of test patterns in their reduced subspace can also be estimated by principal component analysis (PCA). This subspace represents the scale variation of each scale step by principal components of a training texture image. Experimental results show that the proposed method is effective in not only scale invariant texture classification including estimation of scale rate, but also scale invariant segmentation of 2D image for scene analysis.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（学術雑誌）   出版者・発行元：IEICE-INST ELECTRONICS INFORMATION COMMUNICATIONS ENG  

This paper presents a new method for embedding digital watermarks into Bezier polynomial patches. An object surface is supposed to be represented by multiple piecewise Bezier polynomial patches. A Bezier patch passes through its four-corner control points, which are called data points, and does not pass through the other control points. To embed a watermark, a Bezier patch is divided into two patches. Since each subdivided patch shares two data points of the original patch, the subdivision apparently generates two additional data points on the boundaries of the original patch. We can generate the new data points in any position on the boundaries by changing the subdivision parameters. The additional data points can not be removed without knowing some parameters for subdividing and deforming the patch, hence the patch subdivision enables us to embed a watermark into the surface.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

As the representative of the linear discriminant analysis, the Fisher method is most widely used in practice and it is very effective in two-class classification. However, when it is expanded to multi-class classification problem, the precision of its discrimination may become worse. One of the main reasons is an occurence of overlapped distributions on a discriminant space built by Fisher criterion. In order to take such overlap among classes into consideration, our approach builds a new discriminant space with hierarchical tree structure for overlapped classes. In this paper, we propose a new hierarchical discriminant analysis for texture classification. We can divide a discriminant space into subspace by recursively grouping overlapped classes. In the experiment, texture images of many classes are classified based on the proposed method, and we show the outstanding result compared with the conventional method.

Web of Science

researchmap

記述言語：英語   掲載種別：研究論文（国際会議プロシーディングス）   出版者・発行元：IEEE  

To construct a 3D model of an object in a computer, a range finder is used to obtain range images of the object. The range finder takes range images from many view points to eliminate the influence of occlusions. A registration algorithm is, then, used to merge the range images.
When the range finder is mounted on a mobile robot and measures while moving, range images taken by it will be distorted. We extended the registration algorithm to remove the deformation. The extended algorithm can remove it and is used to construct a SD model even for a case where the speed parameters of the robot are not well known.

Web of Science

researchmap

記述言語：日本語   掲載種別：研究発表ペーパー・要旨（全国大会，その他学術会議）  

researchmap

担当区分：筆頭著者   記述言語：英語   掲載種別：記事・総説・解説・論説等（学術雑誌）   出版者・発行元：Springer Singapore  

DOI： 10.1007/978-981-16-4325-5_20

researchmap

掲載種別：速報，短報，研究ノート等（学術雑誌）  

researchmap

記述言語：日本語   掲載種別：研究発表ペーパー・要旨（全国大会，その他学術会議）  

J-GLOBAL

researchmap

担当区分：最終著者   掲載種別：速報，短報，研究ノート等（学術雑誌）  

researchmap

掲載種別：速報，短報，研究ノート等（学術雑誌）  

researchmap

担当区分：筆頭著者  

researchmap

担当区分：筆頭著者,　責任著者  

researchmap

担当区分：最終著者   記述言語：日本語   掲載種別：速報，短報，研究ノート等（学術雑誌）  

researchmap

担当区分：筆頭著者  

researchmap

記述言語：日本語   掲載種別：研究発表ペーパー・要旨（全国大会，その他学術会議）  

researchmap

記述言語：日本語   掲載種別：研究発表ペーパー・要旨（全国大会，その他学術会議）  

researchmap

記述言語：日本語   掲載種別：研究発表ペーパー・要旨（全国大会，その他学術会議）  

researchmap

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：電子情報通信学会  

CiNii Article

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語  

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：電子情報通信学会  

CiNii Article

CiNii Books

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper presents a framework of a task management system for the informationally structured architecture, ROS-TAM, and information structured task management experiments using different types of robots and service tasks. The proposed system interprets user's request, plans a proper robot service, issues a series of robot commands suitable for the structure of each robot, and executes the robot service by task executing machine, SMACH. We conducted a go-and-fetch task with two types of robots.

DOI： 10.1299/jsmermd.2015._1A1-N01_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper proposes a small, lightweight, and easily-relocated sensor terminal named "Portable" for acquiring a variety of environmental information. The Portable is equipped with a variety of sensors including a pyroelectric sensor, a proximity sensor, a sound pressure sensor, a thermometer, a hygrometer, a gas sensor, a flame sensor, and a laser range finder. We introduce three typical applications: abnormality detection, pedestrian tracking and automatic reconfiguration of several Portables for pedestrian tracking.

DOI： 10.1299/jsmermd.2015._1A1-H05_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper presents a viewpoint planning method for an active object recognition by using the shape information of objects. The proposed method uses the certainty that a given object is classifed into one category. The certainty is represented by a probabilistic model. Moreover, the categorization possibility of unknown objects are quantied by the entropy indicating the ambiguity of the recognition. When an object is observed, the certainty is calculated by using the measured data. For each viewpoint, the entropy is computed based on the certainty. We choose as a next viewpoint the viewpoint with minimum entropy. From the experimental results using real objects, the proposed method can achieve en efficient object recognition compared with the method which selects viewpoints randomly.

DOI： 10.1299/jsmermd.2015._2A1-T05_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper reports a new shortcut method of the path planning for keeping clearance between the path and the obstacles. The bonding volumes which have the clearance to each robot link are generated and used for estimating the rough distance of the robot and the obstacles. The effectiveness of the method is verified in simulation.

DOI： 10.1299/jsmermd.2015._2P1-G05_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：公益社団法人 日本生体医工学会  

Finite element method (FEM) is a well-known technique for modeling object physical behaviors. The FE analysis uses polyhedral mesh model, called FE model, of the object to be analyzed. The accuracy of the FE analysis depends on the quality of the polyhedral elements of the FE model. Here, theoretically, the FE analysis using hexahedral elements is superior to that of tetrahedral elements. However, hexahedral element construction needs to be satisfied with more constraints simultaneously compared with the tetrahedral element. Owing to the reason, there are few methods for generating hexahedral FE model of objects with complex shapes. This paper presents an automatic method for generating hexahedral FE model of human tissues. The proposed method is based on a Growing Self-organizing Deformable Model (GSDM) that is a deformable volumetric model. Practically, give a tissue surface model, we use as the initial GSDM the hexahedral model of the cuboid converted with the tissue. The hexahedral FE model is obtained by deforming the GSDM to fit the GSDM to the tissue surface and improve the quality of each element shape.

DOI： 10.11239/jsmbe.53.S253_03

CiNii Article

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper presents a system for estimating indoor human behavior using laser range finders on the floor. The proposed method uses hierarchical hidden Markov model(H-HMM) composed of an action estimate layer and a behavior estimate layer. The former is constructed by two kinds of HMMs: one is the HMM for estimating each action, and the other is the HMM for deciding the human action considering the action continuity. In the latter layer, one HMM learns each behavior by using as the features the relative relationship among the actions and the furniture. Our behavior estimation using such features enable to recognize the behaviors robustly even thought the indoor environment is changed.

DOI： 10.1299/jsmermd.2015._1A1-W04_1

CiNii Article

CiNii Books

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語  

J-GLOBAL

researchmap

記述言語：日本語  

J-GLOBAL

researchmap

記述言語：日本語  

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

複雑な形状をもつ人体組織同士を比較する際,それぞれのメッシュモデルを,形状が単純な目標曲面にいったん写像して,写像先で差異を比較する手法がある.このとき,対象組織で共通の解剖学的特徴が,目標曲面上で同じ位置にあると,他の部位でも対応付けが容易になる.また,組織形状に近い曲面を目標曲面として選ぶことで,写像が単純で直感的になり,解析しやすくなる.しかし,従来の写像法では,写像先を直接的に制御できず,また,従来法の目標曲面は平面や球面のみであり,形状を自由に設定するのは困難である.そこで,本論文では,特に脳表モデルに対し,モデルの写像先を制御しながら,脳表に適した形状の目標曲面へ写像する新たな手法を提案する.まず,自己組織化可変モデル変形法を用いて,モデルを目標曲面上へ写像する.この変形法を用いることで,写像の直接的制御や,モデルと同一位相をもつ形状の目標曲面が使用可能となる.この際,隣接していないモデルの頂点が,目標曲面上で同じ位置に写像されている折り畳みが生じている場合があり,この折り畳みを除去する.次に,モデルの幾何情報の一つである,モデルの表面積に対する各パッチの面積比を写像前後で保存しつつ,目標曲面に脳表を写像する.6個の脳表モデルを用いた実験を行い,提案手法は,特徴領域を特定の位置に写像しつつ,脳表モデルを目標曲面へ滑らかに写像できることを確認した.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

人体組織に対する変形解析手法として,有限要素法(Finite Element Method: FEM)がある.FEMでは,解析対象物体を,要素と呼ばれる小領域の集合で表現した,有限要素(FE)モデルを用いる.要素としては,一般に,生成が容易である四面体要素が用いられるが,六面体要素を用いた場合,より高次の関数で要素内の変形を近似できるため,高精度の解析が可能となる.しかし,表面形状が複雑な物体に対する,六面体FEモデルの自動生成法は,確立されていない.本論文では,競合学習によって変形する成長型自己組織化可変モデル(Growing Self-Organizing Deformable Model: GSDM)を用いた,六面体FEモデルの自動生成法を提案する.既存のFEモデルとの比較実験から,提案手法で生成したFEモデルを用いた変形解析は,より高速,高精度であることが示された.

CiNii Article

CiNii Books

researchmap

記述言語：日本語  

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

本研究では,与えられた眼底画像から,血管に対する錐体細胞分布を自動的に推定する手法を提案する.血管と錐体細胞の相対的位置関係を,提案手法では,錐体細胞から血管への最近距離に基づいて,血管に対する錐体細胞の分布を3種類に分類する.多くの錐体細胞が血管の近傍に分布している場合,錐体細胞の分布と血管の位置に正の相関があるとみなす.一方,錐体細胞が血管より離れて分布している場合,錐体細胞の分布と血管の位置に相関性が低く,負相関とする.どちらの分布にも分類されない分布は,ランダム分布とみなす.シミュレーションによってランダム分布の擬似眼底画像を多数生成し,その累積度数分布を求める.これと,実際の錐体細胞の累積度数分布を比較し,上の種類のいずれかに判別する.

CiNii Article

CiNii Books

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper describes a method for tracking identified persons in a room using the accelerometer attached to each footwear and the laser range finder (LRF) placed on the floor. Even though the LRF measures positions of foot on the floor, it is difficult to identify each person since the floor sensor acquires only the outline of foot. In order to identify the foot, the accelerometer is attached to the footwear. When the feet contacts the floor, it starts to be measured as a blob by the LRF and the acceleration becomes constant. The system detects the timing of such state changes and finds the correspondence between blobs and accelerometers.

DOI： 10.1299/jsmermd.2014._1P2-H01_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

We aim to develop a multi-fingered hand which grasps various objects. We develop a finger equipped with soft fingertip and two layers nails. The first layer of the nails is thin and long. It is possible to insert it into the bottom of the object. The second layer of the nails is thick and short. It supports the elastic force of soft fingertip. We develop a planner for selecting a grasp style according to the height of the object. The planner can grasp a low height object by using a multi-fingered hand. We demonstrate the feasibility of the proposed method with simulation. In addition, it is shown that the developed hand can grasp various objects through experimental results.

DOI： 10.1299/jsmermd.2014._3P2-R01_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

This paper introduces an information processing architecture named ROS-TMS for informationally structured environment. This architecture enables to handle several practical service tasks such as state estimation in a room using accumulated information in a database, and automatic planning and execution for suitable service tasks. As an application of the proposed ROS-TMS, we present an emergency detection and alert system using distributed sensors and a service robot.

DOI： 10.1299/jsmermd.2014._1P2-H02_1

CiNii Article

CiNii Books

researchmap

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本ロボット学会  

This paper describes a method of measuring moving objects and estimating human behaviors in a room using only one laser range finder (LRF) installed in the room and a strip of mirror attached to a side wall close to a floor. The area of sensing is limited to a plane parallel to and just a few centimeters above the floor, thus covering the whole room with minimal invasion of privacy of a resident while reducing occlusion. The important feature of the measurement consists in processing of both distance and reflectance acquired by the LRF from the surface of the existing objects. This enables immediate distinction of clusters of objects made of different materials in the analysis of the scene cluttered with objects. The human behavior models are effectively utilized to estimate human behavior from LRF data. The experimental results validate the effectiveness of the proposed method.

DOI： 10.7210/jrsj.31.769

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

複雑な形状を持つ人体組織同士を比較するために,各メッシュモデルを,形状が単純な目標曲面に一旦写像して,写像先で差異を比較する手法がある.この時,対象組織で共通の解剖学的特徴が,目標曲面上で同じ位置にあると,他の部位でも対応付けが容易になる.また,組織形状に近い曲面を目標曲面として選ぶことで,写像が単純で直感的になり,解析し易くなる.しかし,従来の写像法では,写像先を直接的に制御できず,また,その目標曲面は平面や球面のみであり,形状を自由に設定するのは困難であった.そこで,本研究では,特に脳表モデルに対し,モデルの写像先を制御しながら,脳表に適した形状の目標曲面へ写像する新たな手法を提案する.6個の脳表モデルを用いた実験を行い,特徴領域を特定の位置に写像しつつ,脳表モデルを任意曲面へ滑らかに写像できることを確認した.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

We developed a service robot that picks up everyday objects. The robot is equipped with a small manipulator and a Kinect sensor. A role of the robot is to collect objects which lie scattered on the floor of a room before a cleaning robot (Roomba) works. The robot uses not only a sensor mounted on it but also one laser range finder (LRF) installed in the room. By this, the robot can find objects and move to the grasping point efficiently.

DOI： 10.1299/jsmermd.2013._2A2-Q08_1

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：公益社団法人 日本生体医工学会  

To estimate the internal structures of the patient brain, several methods map a brain atlas to a patient brain shape using some landmarks, which are selected manually from the brain shape. The determination of the correspondence between small sulci of arbitrary two brains is complex. Moreover, the relationship between the surface shape and internal structure of the brain is unclear. Therefore, even if the surface the deformed atlas is fitted to the patient brain shape, the accurate internal structures of the brain can not always be obtained. To solve these problems, this paper proposes a method of selecting landmarks to estimate the internal structures of the patient brain by deforming brain atlas. Firstly, the brain shapes are represented by a simple shape. Secondly, a small number of initial landmarks are selected manually from the contours of the approximated brain shape and the internal structures which can be identified from MR images. Thirdly, some new landmarks are generated automatically on the contours based on the initial landmarks. Finally, the brain atlas is deformed non-rigidly using these landmarks. To evaluate our method, we estimate the patient brain structure using 4 methods. From the experimental results using 10 brain images, our method can estimate the patient brain structure reliably and stably using a small number of landmarks compared with the method using the landmarks on the original brain shape.

DOI： 10.11239/jsmbe.51.390

CiNii Article

CiNii Books

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

We propose a method for localizing furniture on caster and wheelchair using accelerometers and a floor sensor. The floor sensor measures positions of objects on the floor using laser range finder. However, it is difficult to specify each object since the floor sensor acquire only outline of the object. The accelerometer is attached to the moving furniture and integrated with the floor sensor for localizing and specifying it. In this paper, we extend our previous system for localizing multiple objects and verify effectiveness for a robot working in an everyday environment.

DOI： 10.1299/jsmermd.2012._1A1-R04_1

CiNii Article

CiNii Books

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：日本生体医工学会  

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

我々は,以前,ニューラルネットワークを用いて実時間有限要素解析法と,それに基づき実質臓器の変形を推定するシステムを構築した.本論文は,その手法を拡張し,大変形を伴う胃の物理的振る舞いをシミュレートする手法を提案する.胃のような管腔臓器は,形状だけでなく体積も変化するため,その変形パターン数は実質臓器と比べ爆発的に増加する.このように膨大なパターンを持つ対象を,1つのニューラルネットワークで学習させることは困難である.この問題を解決するために,我々は,複数のニューラルネットワークを用いて胃の変形を推定する.まず,変形の類似性に基づいて変形パターンをいくつかのサブデータセットに分類する.また,各サブデータセット内の冗長なデータを削除した後,各サブデータセットを用いて1つのニューラルネットワークを構築する.実験結果から,提案手法は,ネットワークの学習の高速化が実現でき,また,構築したシステムは,非線形有限要素法とほぼ同程度精度を保ちつつ,実時間で胃の変形を推定することが可能となった.

CiNii Article

CiNii Books

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：日本VR医学会  

内視鏡外科手術に適した立体映像提示システムとして、ドーム型スクリーンを用いた立体映像提示システム(3DD)を開発した。更に本システムが有する奥行き情報の直感的な提示が内視鏡外科手術における鉗子動作に与える影響について、従来の2D CRTモニターと比較し検討した。対象は内視鏡外科手術の経験のない医学生16人と日本内視鏡外科学会認定医の資格を持つ外科医3人とした。縫合結紮手技中の鉗子動作の解析結果より、3DDでは医学生においても他の方向に比べ前後方向に移動距離が大きいという外科医と同様の傾向がみられた。また、鉗子の軸周りの回転を多用していることからも、2Dに比べ鉗子動作の幅を増加することが確認された。

DOI： 10.7876/jmvr.8.29

CiNii Article

researchmap

その他リンク： http://search.jamas.or.jp/link/ui/2011149858

記述言語：日本語  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：日本工業出版  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：日本生体医工学会  

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

気道確保法の一つである気道挿管では,通常まず喉頭鏡を使って喉頭展開を行い,声門の位置を目視により確認する.しかし実際の医療現場では,上気道閉塞など様々な要因で,声門の位置を目視により確認しづらい場合がある.この不完全な確認が原因で食道へ誤挿管した場合,気道が確保されず危険なだけでなく,無理な目視のために頸椎や歯牙損傷などの合併症を引き起こす危険性がある.安全・確実な気道挿管の実現に向けて,我々は,スタイレット先端に小型カメラを搭載した自動気管内挿管システムを開発することを自指している.本論文では,その要素機能として,カメラから取得される画像から,挿管チューブが気道あるいは食道に挿管されているかを自動的に識別する方法を提案する.本手法は,気道画像には気道周囲の輪状軟骨が特徴的に観察されることから,まずこの環状模様の記述に適した特徴量を定義し,それに基づいた気道・食道識別器をAdaBoostによって構築する.実験の結果,97.6%の高い識別率で気道・食道の判別が可能であり,提案手法の有効性が確認できた.

CiNii Article

CiNii Books

researchmap

記述言語：日本語  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人日本消化器外科学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

環境の監視等を目的として,マルチカメラシステムによる多視点行動認識手法を提案する.本手法は次の3つの性質を持つ.第一は各カメラの視野を配慮した相補的な情報統合,第二はキャリブレーションフリーなシステムであること,第三は異種センサの組み込み可能性である.これらの性質は各カメラに付随した識別器による認識結果を統合するという枠組により実現される.具体的には,AdaBoostを用い各カメラに付随した識別器を取捨選択および統合しながら,マルチカメラシステム全体として高精度な識別器を自動構成する.本稿ではさらに,このカメラ識別器の取捨選択の際に局所領域選択を導入することを検討する.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：(一社)日本コンピュータ外科学会  

J-GLOBAL

researchmap

記述言語：日本語   出版者・発行元：(一社)日本内視鏡外科学会  

researchmap

記述言語：日本語   出版者・発行元：一般社団法人日本消化器外科学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

レンジセンサにより取得した幾何モデルにカラーセンサで撮影したテクスチャ画像を貼り付けて表示するテクスチャマッピングを容易に実現するには,テクスチャ画像と幾何モデルのみからカラー・レンジセンサ間の相対位置関係を知ることが望ましい.本論文では,幾何拘束に基づく大域的手法とエッジの対応付けに基づく局所的手法の組合せにより,センサ間の相対位置・姿勢を初期値の変動にロバストにかつ高精度に推定し,テクスチャ画像と幾何モデルの位置合せを実現する手法を提案する.本手法はまず,テクスチャ画像から稜線と平面領域を抽出する.次に,この稜線と平面領域を幾何モデルに逆投影し,対象における幾何拘束条件を推定しつつ,この拘束条件のもとでセンサ間の相対位置・姿勢の初期推定値を求める.最後に,テクスチャ画像と幾何モデルの各エッジ間の対応付けに基づき,センサ間の相対位置・姿勢を決定する.実験では,エッジ間の対応付けに基づく従来手法と比較して,位置合せの成功率が41%から75%に向上した.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人日本外科学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人 日本機械学会  

We have been developing an automatic endotracheal intubation system using a high performance stylet. This system is composed of an endotracheal stylet and a camera attached at the point of the stylet. One of fundamental functions of the automatic intubation system is to check whether the endotracheal intubation is completed appropriately or not. For achieving this function, this paper presents a method for classifying trachea and esophagus images taken by the mounted camera. The proposed method utilizes circular patterns of cricoid cartilage which is observed in trachea images only. Experimental results show that the proposed method can extract the cricoid patterns in trachea images efficiently and accurately.

DOI： 10.1299/jsmermd.2008._2A1-C04_1

CiNii Article

CiNii Books

researchmap

J-GLOBAL

researchmap

記述言語：日本語  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：(一社)日本内視鏡外科学会  

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

多数台のカメラによりシーン内に存在する対象物体の全周の幾何情報及び光学情報を取得し,任意視点からの画像を生成する手法として,視体積交差法と多視点ステレオ法が提案されている.しかしこれらの手法は単一物体あるいはオクルージョンの生じない複数物体を対象とした手法であり,シーン内に複数物体が存在し物体間に相互オクルージョンが生じる場合,それぞれの物体形状を同時に復元することは困難であった.この問題に対し,我々はこれまでに高速な境界追跡手法であるFast Level Set Methodを複数ステレオ距離画像に適用し,複数対象物体の三次元形状をオクルージョンに頑強に復元するシステムを構築している.本論文では,これまでに構築したシステムを8台の計算機からなるPCクラスタへ実装し,Fast Level Set Method処理の並列計算により,より高速な三次元形状の復元を実現する.また対象物体が移動する場合,その移動方向を予測し,移動体を処理する計算機の計算負荷を低減することで,移動体の正確な三次元形状を遅れなく復元する手法を提案する.更に,舞踊の測定実験により,対象が高速に移動しても,従来システムと比較してより正確な三次元形状の復元が可能であることを示す.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人情報処理学会  

ある物体形状をあらわす三角メッシュモデルを曲面へと写像するパラメータ化は、メッシュモデリングにおける要素技術の一つであり、多くの CG アプリケーションに利用することができる。本稿では、このパラメータ化を物体メッシュ間で行う手法を提案する。提案法では、競合学習法の一つである自己組織化マップと形状復元法の一つである最小 2乗メッシュをベースとして用いている。これにより、メッシュの構造的な特徴を考慮したパラメータ化を行うことができる。Parameterization which maps a triangular mesh to a certain domain is a key technique in mesh modeling and useful for varius CG applications. In this paper, we propose a new method for mapping one mesh to another. Our method is based on a learning algorithm of Self-organizing Map and Least-squares Mesh. These techniques enable us to make a parameterization taking account of structural features of meshes.

CiNii Article

CiNii Books

researchmap

その他リンク： http://id.nii.ac.jp/1001/00038039/

DOI： 10.1002/ecjc.20347

researchmap

researchmap

記述言語：英語   出版者・発行元：IEEE  

In 3D computer graphics, mesh parameterization is a key technique for digital geometry processings(DGP) such as morphing, shape blending, texture transfer re-meshing and so on. This paper proposes a novel approach for parameterizing a mesh into another one directly. The main idea of your method is to combine a competitive learning and a least-square mesh techniques. It is enough to give some semantic feature correspondences between target meshes, even if they are in different shapes or in different poses. We show the effectiveness of our approach by giving some examples of its applications.

DOI： 10.1109/PG.2007.21

Web of Science

researchmap

記述言語：日本語   出版者・発行元：画像電子学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：画像電子学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

本論文では,自己組織化特徴マップの概念を用いた可変モデル(SDM: Self-organizing Deformable Model)によるメッシュモデルの三次元モーフィング法を提案する.SDMは,変形の目標とする曲面(目標曲面)上の頂点の分布を学習し,自己組織的に変形を繰り返す可変モデルである.SDMは,変形先の形状を限定しない可変モデルであり,様々なモデルを扱うことができる.本手法では,モーフィングのためのモデル間の対応決定にSDMを用いる.また,ユーザの指定により,モデルの構造的特徴を考慮した対応決定が可能である.本手法のモーフィングにおける有効性を,いくつかの実験を通じて検証した.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人情報処理学会  

本論文では，自己組織化特徴マップの概念を用いた可変モデル(SDM:Self-organizing Deformable Model) によるメッシュモデルの三次元モーフィング法を提案する．SDM は，変形の目標とする曲面(目標曲面) 上の頂点の分布を学習し，自己組織的に変形を繰り返す可変モデルである．SDM は，変形先の形状を限定しない可変モデルであり，様々なモデルを扱うことができる．本手法では，モーフィングのためのモデル間の対応決定にSDM を用いる．また，ユーザの指定により，モデルの構造的特徴を考慮した対応決定が可能である．本手法のモーフィングにおける有効性を，いくつかの実験を通じて検証した．This paper presents a new method for 3-D morphing between mesh models using a Self-organizing Deformable Model(SDM). The SDM is a mesh model which deforms by learning vertex coordinates on target shape. Since the SDM doesn't limit target shapes, we can use various mesh models as SDMs. Therefore, we use SDM for determining correspondence be-tween models. Also, SDM enables user to define the correspondence by model features. We demonstrated some examples for morphing.

CiNii Article

CiNii Books

researchmap

その他リンク： http://id.nii.ac.jp/1001/00052219/

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：英語   出版者・発行元：一般社団法人電子情報通信学会  

In this paper, we address an active way of 3-D distance measurement. Generally, the real-world environment is composed of various kinds of objects. The most adequate parameters of a measurement system may be different in respective objects. It needs to adjust the system parameters such as camera position, pan and tilt angles, focal length, zoom and so on. First, we perform 3-dimensional measurement of the whole scene by a parallel stereo vision and choose some regions for targets and estimate their positions from the measurement result of the whole scene. After adjusting viewing direction and resolution to a target region, we perform a fine measurement again.

CiNii Article

CiNii Books

researchmap

researchmap

出版者・発行元：電気・情報関係学会九州支部連合大会委員会  

DOI： 10.11527/jceeek.2006.0.89.0

CiNii Article

researchmap

記述言語：日本語   出版者・発行元：一般社団法人映像情報メディア学会  

This paper proposes a new method for integrating range images with different resolutions to generate an entire 3D model. The modeling method is based on the fact range images need an overlapping area between the images for the registration process. Thus, we integrate data from two range images into one surface model, and represent the overlapping area with triangular patches. Few conventional methods have paid attention to integrating images with different resolutions. Moreover, most conventional methods find corresponding points by using the distance between points. However, it sometimes causes wrong corresponding points in convex and concave surfaces. We propose a new integration method that uses geometric and topological information. Experimental results showed that this method is efficient for integrating multiple range images.

DOI： 10.3169/itej.60.409

CiNii Article

CiNii Books

researchmap

その他リンク： https://jlc.jst.go.jp/DN/JALC/00274269806?from=CiNii

記述言語：日本語   出版者・発行元：日本工業出版  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人情報処理学会  

本論文では，任意の曲面上にある点データを使って，その曲面へ3次元物体メッシュモデルを写像する新たな手法を提案する．この写像は，コンピュータビジョンやコンピュータグラフィックスの要素技術の一つであり，それを具体化した方法として，平面や球面など単純な形状を持つプリミティブな物体面にメッシュモデルを投影する方法や，逆に，プリミティブな物体の三角メッシュモデルを変形させ，そのモデルを対象物の点データに当てはめる方法がある．しかし，前者の場合，従来の手法では特定の写像面のみを対象としており，もし写像面が適切でない場合，メッシュモデルが写像面にうまく投影されない可能性がある．一方，後者の方法には，復元に要する計算コストと，得られたモデルの形状復元精度が，メッシュモデルの初期設定に依存する問題がある．これらの問題を解決するために，本論文では，競合学習の考えを導入した可変モデルSelf-organizing Deformable Model(SDM)を提案する．まず，競合学習に基づいてSDMを変形することで，写像面の大まかな形状モデル生成する．次に，エネルギー最小化によってSDMによる形状復元精度を向上させる．This paper presents a new method for projecting a mesh model of a source object onto a surface of an arbitrary target object. A deformable model, called Self-organizing Deformable Model(SDM),is deformed so that the shape of the model is fitted to points on the target object. Then, we introduce the idea of combining a competitive learning and an energy minimization into the SDM deformation. Our method is a powerful tool in the areas of computer vision and computer graphics. For example, it enables to map mesh models onto various kinds of target surfaces compared with traditional methods for a surface parameterization, which have only focused on specified target surface. Also the SDM can reconstruct shapes of target objects similar with general deformable models.

CiNii Article

CiNii Books

researchmap

その他リンク： http://id.nii.ac.jp/1001/00052328/

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：英語  

Viewpoint planning plays an important role in automatic 3D model generation. We have proposed an optimal viewpoint planning [9] for the purpose of model generation by merging images acquired from multiple viewpoints on a sphere that encircles object. This method uses a discretized approximate sphere which involves problems of sphere tessellation and of sphere representation. One must consider that the scanned surface of an object is projected onto a sphere, and that the computational cost increases directly with the number of images. However, there are no appropriate sphere representation methods available. This paper proposes a new representation method based on a 2D array, and develops a new viewpoint planning algorithm. The proposed system is verified by algorithmic analysis and experiments using a real modeling system. © 2004 Wiley Periodicals, Inc.

DOI： 10.1002/scj.1232

Scopus

researchmap

記述言語：英語  

Viewpoint planning plays an important role in automatic 3D model generation. We have proposed an optimal viewpoint planning [9] for the purpose of model generation by merging images acquired from multiple viewpoints on a sphere that encircles object. This method uses a discretized approximate sphere which involves problems of sphere tessellation and of sphere representation. One must consider that the scanned surface of an object is projected onto a sphere, and that the computational cost increases directly with the number of images. However, there are no appropriate sphere representation methods available. This paper proposes a new representation method based on a 2D array, and develops a new viewpoint planning algorithm. The proposed system is verified by algorithmic analysis and experiments using a real modeling system. © 2004 Wiley Periodicals, Inc.

DOI： 10.1002/scj.1232

Scopus

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

人間社会において人間の顔は重要な情報源である.機械が顔を扱うためには最初に画像から顔領域を検出する必要があるが,顔は非剛体であるため,その変化に対応する顔検出法を構築するのは簡単な問題ではない.そこで,本研究では,有用な高次元特徴の選択および特徴生成により,顔と非顔の識別を安定して行う汎化性能の高い顔検出法を構築する.提案手法は,カーネル特徴をKullback-Leibler情報量に基づいて多数生成し,ブースティングと呼ばれる機械学習アルゴリズムを用いて,顔検出に有用な特徴を選択する.これにより,カーネル特徴を用いた一般的な学習器であるサポートベクターマシーンと同等の性能を持ち,計算量が少ない顔検出器を実現できた.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

本手法では,階層的な判別空間による新たな多クラス識別法を提案する.テクスチャのような多クラスを識別するには2クラス識別に有効なFisher法が用いられている.しかし,この手法を多クラス識別に適用すると,判別空間内でクラス分布に重なりを生じ,識別率低下を招く場合がある.そこで,判別空間内で重複するクラスに対して新たに判別空間を設け識別を行う.このような判別空間を再帰的に構築することで,識別精度の向上を図る.Outexデータベースのテクスチャ画像を用いた識別実験では,従来法に比べ良好な識別結果を得ることが出来た.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

本論文では,高次元特徴の効果的な次元圧縮に基づいた多クラスパターンの認識の為の枠組みを提案する.パターン認識において,各クラスの識別に適した特徴を事前に特定することは一般に困難である為,特徴抽出時には,様々な観点から幅広く特徴を考慮することが重要である.しかし,その結果得られる高次元特徴を直接的に識別に用いると,計算量爆発などの様々な問題を引き起こしてしまう.そこで,識別に有効と考えられる特徴を多く含ませた高次元特徴を構成し,その高次元特徴をニューラルネットを用いて,識別に有効な情報だけを残しつつ,低次元の特徴ベクトルヘ効果的に圧縮することにより高精度かつ高速な認識の実現を図る.

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人情報処理学会  

従来のステレオ視はカメラパラメータを固定し，3次元計測を行う研究が多い．本研究ではシーン全体の距離計測を行った後，シーンの中の物体ごとに計測を行い，シーンと各物体との位置関係や物体間の位置関係を確立することにより実空間に近い３次元情報を獲得することを研究の目的とする．物体ごとの計測を行う際，カメラの向きや解像度を調節しながら必要に応じて注目したい物体を計測することにより能動的に３次元計測を行い，計測時間の短縮とデータ量の削減をはかる．There have been a number of studies, which were performed for 3-dimensional measurement by means of fixed camera parameters. In this paper, we will perform 3-dimensional distance measurement in the whole scene, and then measure distance of each object in the scene. It is our aim to establish the position relation between a scene and each object, and the position relation among the objects. When we perform 3-dimensional measurement, we will measure an object by adjusting focus length and resolution to the property of each object. Our methodology makes it possible to perform 3-dimensional measurement actively and to shorten the measurement time, and to reduce the amount of data.

CiNii Article

CiNii Books

researchmap

その他リンク： http://id.nii.ac.jp/1001/00052528/

記述言語：日本語  

DOI： 10.3169/itej.58.713

Scopus

CiNii Article

researchmap

記述言語：英語   出版者・発行元：IEEE  

It is difficult to give a robot all possible motions beforehand in a certain environment. Therefore, the robot needs to learn how to recognize others' motions and to generate its own motions autonomously for working well. These learning algorithms need an efficient way to make recognition and generation of motions work together, because the), take many computing resources. This paper focuses on a generation-based recognition.
Our system consists of recognition and generation modules. The former and latter are constructed from left-to-right Hidden Markov Models (HMM) and Reinforcement Learning (RL), respectively. When a HMM in recognition module does not work enough, the model parameters of HMM are re-estimated by using a state-value function of RL in generation module. The proposed method enables us to improve the reliability of the HMM.

DOI： 10.1109/ICSMC.2004.1401029

Web of Science

researchmap

記述言語：英語   出版者・発行元：IEEE  

Recent researches for detecting face regions from images have paid attention to high dimensional kernel features(KFs), which are obtained by a non-linear transformation of original features extracted from images. A Support Vector Machine(SVM) is one of the most prominent learning algorithms for KFs. However; SVM is time-consuming because of needing a large number of KFs to improve the accuracy of the classification. This paper proposes a new method that constructs a classifier between face and non-face regions by generating and choosing KFs based on Kullback-Leibler Divergence(KLD). The KLD means a distance between two distributions of face and non-face data under a given KF and some KFs of large KLDs are selected for the face detection. Moreover; the use of KLD enables us to generate new KFs' and to deal with different kinds of KFs concurrently Some experiments show that our method can reduce the number of KFs much more than SVM, and achieve almost equal or better detection rate than that of SWM.

DOI： 10.1109/ICSMC.2004.1400653

Web of Science

researchmap

記述言語：英語   出版者・発行元：IEEE  

Recent researches for detecting face regions from images have paid attention to high dimensional kernel features(KFs), which are obtained by a non-linear transformation of original features extracted from images. A Support Vector Machine(SVM) is one of the most prominent learning algorithms for KFs. However; SVM is time-consuming because of needing a large number of KFs to improve the accuracy of the classification. This paper proposes a new method that constructs a classifier between face and non-face regions by generating and choosing KFs based on Kullback-Leibler Divergence(KLD). The KLD means a distance between two distributions of face and non-face data under a given KF and some KFs of large KLDs are selected for the face detection. Moreover; the use of KLD enables us to generate new KFs' and to deal with different kinds of KFs concurrently Some experiments show that our method can reduce the number of KFs much more than SVM, and achieve almost equal or better detection rate than that of SWM.

DOI： 10.1109/ICSMC.2004.1400653

Web of Science

researchmap

記述言語：英語   出版者・発行元：IEEE  

It is difficult to give a robot all possible motions beforehand in a certain environment. Therefore, the robot needs to learn how to recognize others' motions and to generate its own motions autonomously for working well. These learning algorithms need an efficient way to make recognition and generation of motions work together, because the), take many computing resources. This paper focuses on a generation-based recognition.
Our system consists of recognition and generation modules. The former and latter are constructed from left-to-right Hidden Markov Models (HMM) and Reinforcement Learning (RL), respectively. When a HMM in recognition module does not work enough, the model parameters of HMM are re-estimated by using a state-value function of RL in generation module. The proposed method enables us to improve the reliability of the HMM.

DOI： 10.1109/ICSMC.2004.1401029

Web of Science

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：英語   出版者・発行元：IEEE COMPUTER SOC  

This paper focuses on the polygonal representation of a 3D object model which is composed of a lot of points on the object surface. In the 3D animation, it is sometimes necessary to use multi-resolution representation of a model to cope with various situations, and it is preferable to represent the models with different resolutions in a certain unified data structure. In order to meet theses requirements, this paper presents a new method which generates the approximated model of an original one by using multiple deformable models, called active balloon models (ABMs). We extract the three-dimensional skeleton of the object. The obtained skeleton comprises nodes and edges, and represents the structure of the object. Based on the skeleton, the approximated model is generated by deforming the ABMs. Some experimental works are made to verify the capability of our method.

DOI： 10.1109/IM.2003.1240236

Web of Science

researchmap

記述言語：英語   出版者・発行元：IEEE COMPUTER SOC  

This paper focuses on the polygonal representation of a 3D object model which is composed of a lot of points on the object surface. In the 3D animation, it is sometimes necessary to use multi-resolution representation of a model to cope with various situations, and it is preferable to represent the models with different resolutions in a certain unified data structure. In order to meet theses requirements, this paper presents a new method which generates the approximated model of an original one by using multiple deformable models, called active balloon models (ABMs). We extract the three-dimensional skeleton of the object. The obtained skeleton comprises nodes and edges, and represents the structure of the object. Based on the skeleton, the approximated model is generated by deforming the ABMs. Some experimental works are made to verify the capability of our method.

DOI： 10.1109/IM.2003.1240236

Web of Science

researchmap

記述言語：日本語  

researchmap

記述言語：日本語  

researchmap

記述言語：英語   出版者・発行元：IEEE  

This paper proposes a method for the animation of things in nature based on the observation of natural phenomena and on the synthesis of their behavioral patterns using machine learning methods. The natural phenomenon to be animated is recorded using a video camera, and its characteristic behavior is captured. A data sequence representing the subject behavior is obtained from the captured video. By learning the inherent structure in the feature space of some sample data, the learned model can synthesize a novel data sequence from the existing sequences. The generated sequences of behavioral patterns could differ from every original data sequence but preserve characteristics of the subject behavior. We demonstrate the natural animation synthesis through such behavioral pattern sequences, and produce some realistic animation which depict the subject.

DOI： 10.1109/CIRA.2003.1222177

Web of Science

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

researchmap

担当区分：筆頭著者   記述言語：英語  

CiNii Article

researchmap

担当区分：筆頭著者  

CiNii Article

researchmap

researchmap

担当区分：筆頭著者  

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

CiNii Article

CiNii Books

researchmap

researchmap

担当区分：筆頭著者   記述言語：英語  

researchmap

記述言語：日本語   出版者・発行元：一般社団法人電子情報通信学会  

3次元物体モデル自動生成には, 距離画像統合と画像取得のための視点計画の組合せが適している.我々は, 対象物を囲む球面上にある多視点からの画像の統合によるモデル生成を目的とし, そのための最適視点計画法を提案した.この計画法では, 球面の代わりに離散化された近似球面を用いるので, 球面分割及び計算機内での球面表現が問題となる.その際, 本視点計画法では既に走査した物体表面を球面へ投影した領域を抽出することと, その計算コストは画像枚数に比例し増加することをも考慮する必要がある.しかし, 従来これに適した球面表現法がなかった.本論文では, 2次元配列による新しい球面表現法を提案し, これに基づく視点計画アルゴリズムを開発した.そして, 解析と実験により提案した手法の有効性を検証した.

CiNii Article

CiNii Books

researchmap

researchmap

researchmap

researchmap

担当区分：筆頭著者  

researchmap

researchmap

担当区分：筆頭著者   記述言語：英語   出版者・発行元：IEEE COMPUTER SOC  

The paper presents a method of creating a complete model of a curved object from a sequence of range images acquired by a fixed range finder. To accomplish the modeling fast and accurately in an optimal manner, we propose a new on-line viewpoint planning algorithm to choose the Next Best Viewpoint (NBV) based on the already obtained partial model. The NBV is determined by evaluating factors such as possibility of merging new data, local shape changes, registration accuracy and control point distribution.

Web of Science

researchmap

担当区分：筆頭著者   記述言語：英語  

researchmap

記述言語：英語   出版者・発行元：I E E E, COMPUTER SOC PRESS  

We propose a new method of creating a complete model of a curved object from multiple range images acquired by showing it at different poses. The pose of the object is changed by a manipulator in order to view the object from some specified viewpoints. The pose is planned after each new image as merged into a unified representation. A rating function for the planning is defined to tate into consideration the factors such as possibility of merging new data, registration. accuracy and control point selection.

Web of Science

researchmap

記述言語：英語   出版者・発行元：I E E E, COMPUTER SOC PRESS  

We propose a new method of creating a complete model of a curved object from multiple range images acquired by showing it at different poses. The pose of the object is changed by a manipulator in order to view the object from some specified viewpoints. The pose is planned after each new image as merged into a unified representation. A rating function for the planning is defined to tate into consideration the factors such as possibility of merging new data, registration. accuracy and control point selection.

Web of Science

researchmap

開催年月日： 2022年1月13日

記述言語：日本語  

researchmap

開催年月日： 2021年11月16日

記述言語：日本語  

researchmap

開催年月日： 2021年9月7日

記述言語：日本語  

researchmap

開催年月日： 2021年7月1日 - 2021年7月3日

記述言語：日本語   会議種別：口頭発表（招待・特別）  

researchmap

記述言語：日本語   会議種別：口頭発表（招待・特別）  

researchmap

記述言語：日本語   会議種別：口頭発表（招待・特別）  

researchmap

researchmap

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

記述言語：日本語  

researchmap

記述言語：日本語   会議種別：ポスター発表  

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

会議種別：口頭発表（招待・特別）  

researchmap

記述言語：日本語   会議種別：口頭発表（一般）  

researchmap

researchmap

researchmap

researchmap