Faculty Profiles - Nakazawa Atsushi

MISC - Nakazawa Atsushi

Division display 141 - 160 of about 169 ／ All the affair displays >>

Modeling Indoor Scene by Determining its Reflection Parameters

HARADA TAKAAKI, HARA KENJI, NAKAZAWA ATSUSHI, SAITO HIROAKI, IKEUCHI KATSUSHI

IPSJ SIG Notes. CVIM 2004 ( 6 ) 45 - 52 2004.1

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

To realistically model indoor scene is crucial and challenging problem. Indoor scene has a large number of three-dimensional range data and spatial restriction in two-dimensional photographing, as well as inevitable specularities in certain parts. In this paper we intend to deal with such a problem. Principally, our method use three-dimensional (3D) geometrical data and ordinary two-dimensional (2D) color images. The 3D geometrical data is created using parallel alignment and merging of a large number of range data. We texture the 3D data with 2D color images using simultaneous registering technique. Then, we estimate the reflection parameters of the diffuse and specular reflection components from a single image where specularities occur. Finally, using all acquired and estimated data, we can generate synthetic images of real objects realistically.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00052531/

Simultaneous Alignment of a Large Number of Range Images

大石岳史, 佐川立昌, 中沢篤志, 倉爪亮, 池内克史

3D映像 18 ( 4 ) 2004

Matching and blending human motions using temporal scaleable dynamic programming Reviewed

A. Nakazawa, S. Nakaoka, K. Ikeuchi

2004 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 1 287 - 294 2004

Detecting dance motion structure through music analysis Reviewed

T Shiratori, A Nakazawa, K Ikeuchi

SIXTH IEEE INTERNATIONAL CONFERENCE ON AUTOMATIC FACE AND GESTURE RECOGNITION, PROCEEDINGS 857 - 862 2004

Language：English Publisher：IEEE COMPUTER SOC

In these days, many important intangible culturalproperties of the world are being lost because of the lack of successive performers. Digital archiving technology is one of the effective solutions for this issue, and we have started our digital archiving project of cultural properties including these intangible ones. For these human motion archives, the method of automatic motion structure analysis is vital for a variety of purposes. We believe that the dance motion consists of "primitive motions" and that motion analysis is necessary to detect these components. Particularly for dance motions, we think these primitives must be synchronized to the musical rhythm. In this paper, we introduce musical information for motion structure analysis. This method automatically detects the musical rhythm and segments the original motion, and classifies them as to the primitive motions. The experimental results confirm that our motion analysis yielded the primitive motions in accordance to the musical rhythm.

DOI： 10.1109/AFGR.2004.1301641

Web of Science

researchmap

Representing cultural heritage in digital forms for VR systems through computer vision techniques Reviewed

K Ikeuchi, A Nakazawa, K Hasegawa, T Ohishi

PROCEEDINGS OF THE 17TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, VOL 4 4 1 - 6 2004

Language：English Publisher：IEEE COMPUTER SOC

This paper overviews our research on digital preservation of cultural assets and digital restoration of their original appearance. Geometric models are digitally achieved through a pipeline consisting of scanning, registering and merging multiple range images. We have developed a robust simultaneous registration method and an efficient and robust voxel-based integration method. On the geometric models created, we have to align texture images acquired from a color camera. We have developed a texture mapping method to utilize laser reflectance. In an attempt to restore the original appearance of historical heritage objects, we have synthesized several buildings and statues using scanned data and literature survey with advice from experts.

Web of Science

researchmap

スケーラブルDPを用いた人間動作のマッチングと生成

中澤篤志

MIRU2004 1 171 - 176 2004

Symbolic Motion Description for a Dancing Humanoid Robot

NAKAOKA Shinichiro, NAKAZAWA Atsushi, YOKOI Kazuhito, IKEUCHI Katushi

Technical report of IEICE. PRMU 103 ( 390 ) 55 - 60 2003.10

Language：Japanese Publisher：The Institute of Electronics, Information and Communication Engineers

We have developed a system for humanoid robots to imitate whole body motions performed by a human. Our target motions are traditional folk dances. Human motions can be acquired by a motion capturing system. However, those motion data cannot be imported directly into a robot which have body properties different from that of humans. In order to realize the motion which satisfies the essence of the original dance motion under the robot constraints, we propose a motion recognition and generation method using symbolic motion descriptions. This method have realized an actual performance of "Tsugaru Jongara-Bushi", which is one of the Japanese traditional folk dances, on the humanoid robot "HRP-1S".

CiNii Article

CiNii Books

researchmap

Estimating Illumination Position, Color and Surface Reflectance Properties from a Single Image

HARA KENJI, RobbyT.Tan, NISHINO KO, NAKAZAWA ATSUSHI, IKEUCHI KATSUSHI

44 ( 9 ) 94 - 104 2003.7

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

In this paper we propose a new method for estimating a position and color of a light source, as well as reflectance properties of a real object's surface, from a single image. We use the intensity of the diffuse and specular component for estimating the light source position, while the color distribution of the specular region is for estimating the light source color. The flow of this method is basically as follows: first, an initial position of the light source is estimated from a peak location of the specular region and a rough intensity value of the diffuse region. This diffuse-to-specular intensity value is also used to determine the initial values of the object reflectance properties. After having the initial values, using iterative fitting method, the light position and reflectance properties are estimated simultaneously. Finally, the estimation process of the light source color is based on the color distribution of the specular region. By knowing the light source position, color and the object reflectance properties, we can freely generate synthetic images under arbitrary light source conditions.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00018076/

Retrieving and applying human dance motion styles from several motion data

NAKAZAWA Atsushi, NAKAOKA Shinichiro, IKEUCHI Katsushi

IPSJ SIG Notes. CVIM 2003 ( 36 ) 101 - 107 2003.3

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

This paper presents, the method to analyze human motion for the purpose of digital archive of intangible cultural heritages, such as folk dances. In the recent studies, the whole motion sequence is segmented and directly use for computer animation or motion analysis. We proposes the idea that the human dance motion consists of 'Basic Motion' and 'Motion Styles'. The Basic Motion is common motion for any dancers, and Motion Styles represents the uniqueness of the individual dancers. In the experiment, we confirmed proposed method works effectively through different motion data of male and femail dancers.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00052645/

Parallel Alignment of a Large Number of Range Images on PC Cluster

OISHI Takeshi, SAGAWA Ryusuke, NAKAZAWA Atsushi, KURAZUME Ryo, IKEUCHI Katsushi

IPSJ SIG Notes. CVIM 137 ( 36 ) 27 - 34 2003.3

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

This paper describes a parallel alignment method of multiple range images. It is difficult to align a large number of range images simultaneously. Therefore, we developed the parallel method to improve the time performance ana memory performance. The method was tested on 16 processor PC cluster and shown the high extendibility and the performance improvement in time and memory.

CiNii Article

CiNii Books

researchmap

Geometric and Photometric Merging for Large-Scale Objects

SAGAWA Ryusuke, MASUDA Tomohito, OISHI Takeshi, NISHINO Ko, NAKAZAWA Atsushi, KURAZUME Ryo, IKEUCHI Katsushi

IPSJ SIG Notes. CVIM 137 ( 36 ) 9 - 18 2003.3

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

In this paper, we consider the geometric and photometric modeling of large-scale and intricately shaped objects, such as cultural heritage objects. When modeling such objects, new issues occurred during the modeling steps which had not been considered in the previous research conducted on the modeling of small, indoor objects. First, when modeling a large-scale and intricately shaped object, a huge amount of data is required to model the object. We would like to propose two approaches to handling this amount of data. We propose a novel method for searching the nearest neighbor by extending the search algorithm using a k-d tree. We constructed a merged model in adaptive resolution according to the geometric and photometric attributes of range images for efficient use of computational resources. Second, we reconstructed a 3D model with an appearance which successfully discarded outliers due to specular reflection, by taking a consensus of the appearance changes of the target object from multiple range images. The photometric attribute of the model can be used for aligning with 2D color images. The third issue is complementing unobservable surfaces of an object. We propose a novel method to complement such holes or gaps in the surfaces, areas which are not observed by any scans. In this novel method, the surface of an object is scanned using various kinds of sensors. We efficiently complement them by taking a consensus of the signed distances of neighbor voxels.

CiNii Article

CiNii Books

researchmap

Recognition and Generation of Leg Motion for Dance Imitation by a Humanoid Robot

NAKAOKA Shinichiro, NAKAZAWA Atsushi, YOKOI Kazuhito, IKEUCHI Katsushi

IPSJ SIG Notes. CVIM 2003 ( 36 ) 93 - 100 2003.3

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

This study attempts to realize a humanoid robot which can imitate human dances. Human motion can be acquired by a motion capturing system. However, the captured data cannot be directly imported to a robot because of difference of body structure between a human and the robot. Since leg motion must consider balance keeping, constraints of it is too restricted to apply the captured data. This paper proposes a method to recognize human motion and generate robot motion from the recognition result. Primitive motions which is required to express leg motion in dances are defined. A sequence of primitives is extracted from captured motion data. Leg motion of a robot is generated from the primitive sequence to satisfy the leg constraints and the dynamic balance. Generated motion is tested on OpenHRP dynamics simulator. We certified validity of this method.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00052644/

Generating whole body motions for a biped humanoid robot from captured human dances Reviewed

S Nakaoka, A Nakazawa, K Yokoi, H Hirukawa, K Ikeuchi

2003 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION, VOLS 1-3, PROCEEDINGS 3 3905 - 3910 2003

Language：English Publisher：IEEE

The goal of this study is a system for a robot to imitate human dances. This paper describes the process to generate whole body motions which can be performed by an actual biped humanoid robot. Human dance motions are acquired through a motion capturing system. We then extract symbolic representation which is made up of primitive motions: essential postures in arm motions and step primitives in leg motions. A joint angle sequence of the robot is generated according to these primitive motions. Then joint angles are modified to satisfy mechanical constraints of the robot. For balance control, the waist trajectory is moved to acquire dynamics consistency based on desired ZMP. The generated motion is tested on OpenHRP dynamics simulator. In our test, the Japanese folk dance, 'Jongara-bushi' was successfully performed by HPP-1S.

Web of Science

researchmap

Modeling from Reality - Creating virtual reality models through observation Reviewed

K Ikeuchi, A Nakazawa, K Nishino, R Sagawa, T Oishi, H Unten

VIDEOMETRICS VII 5013 117 - 128 2003

Language：English Publisher：SPIE-INT SOC OPTICAL ENGINEERING

In this paper, we present an overview of our project to construct a digital archive of cultural heritages. Among the efforts in our project, we briefly overview our research on geometric and photometric preservation of cultural assets and restoration of their original appearance. Digital geometric modeling is achieved through a pipeline consisting of scanning, registering and merging multiple range images. For these purposes, we have developed a robust simultaneous registration method and an efficient and robust voxel based integration method. On top of the geometrical model, we align texture images acquired at the scanning. Because the geometrical relation between the range sensor and the image sensor are calibrated, we automatically align texture images onto the geometrical models. For photometric modeling, we have developed a surface light field based method, which captures the appearance variation of real world objects under different viewpoints and illumination conditions from a series of images. As an attempt to restore the original appearance of historical heritages, we have reconstructed several buildings and statues that have been lost in the past. In this paper, we overview these techniques and show several results of applying the proposed methods to existing ancestral assets.

DOI： 10.1117/12.473086

Web of Science

researchmap

A Survey of the Cooperative Distributed Visual Tracking Systems

NAKAZAWA ATSUSHI, HIURA SHINSAKU, KATO HIROKAZU

43 ( 11 ) 11 - 20 2002.12

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

This paper shows the introduction and survey of the Cooperative Distributed Vision (CDV) systems that consist of many image processor and cameras connected through a computer network. Nowadays, some systems that is based on the idea of CDV systems have been proposed. We evaluate the difference between each systems according to their image processing hierarchy, communication procedure, control rules, their independency and resulting features. These survey is expected for the designing new CDV based image recognition systems.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00018093/

JST-1 Digital Archive of Cultural Heritage through Observation

IKEUCHI Katsushi, NAKAZAWA Atsushi

2002 1 - 2 2002.9

解説分散協調視覚による広域観測システム--多数のカメラとネットワークによる高度な広域観測

中澤篤志, 浮田宗伯, 加藤丈和

画像ラボ 13 ( 8 ) 6 - 11 2002.8

Digital Archive of Cultural Heritage through Observation

IKEUCHI Katsushi, NAKAZAWA Atsushi

Journal of the Visualization Society of Japan 22 7 - 10 2002.7

Tracking multiple persons using distributed vision systems

NAKAZAWA Atsushi, KATO Hirokazu, HIURA Shinsaku, INOKUCHI Seiji

IPSJ SIG Notes. CVIM 2002 ( 2 ) 37 - 44 2002.1

Language：Japanese Publisher：Information Processing Society of Japan (IPSJ)

This paper shows the multiple targets watching method in wide-area spatial environment using the Distributed Vision Systems (DVS). DVS is constructed with some "watching stations" that consist of a camera and an image processor and a computer network connects them. This system's goal is to track multiple persons in a wide-area that cannot be watched by one visual sensor. We introduce three algorithms to realize this system, realtime human tracking method, the task decision algorithms of watching stations and the object-matching method between stations. Finally, we describe experimental results that show the validity of our approach.

CiNii Article

CiNii Books

researchmap

Other Link： http://id.nii.ac.jp/1001/00052749/

視覚による舞踊動作の保存解析および生成

中澤篤志

画像の認識理解シンポジウム2002 153 2002

　 PREV - NEXT 　

< Nakazawa Atsushi >

MISC - Nakazawa Atsushi