In this paper, a transmitting process of a sequence of holograms describing 3D moving objects over the communicating wireless-network system is presented. A sequence of holograms involves holograms is transformed into a bit stream data, and then it is transmitted over the wireless network. It is shown that applying this technique, holographic data of 3D moving object is transmitted in high quality and a relatively good reconstruction of holographic images is performed.

View full abstract

A computer simulation method of gray-level image reconstruction from Fresnel computer-generated holograms (CGHs) has been developed using a two-dimensional image as an object. But, generally it needs a long calculation time to synthesize interference fringes of a large size CGH. Therefore, it was studied to reduce the calculation time by considering special characteristics of the interference fringe patterns.

View full abstract

[in Japanese]

View full abstract

A device more than 10,000,000 pixels is put to practical use by the high definition of image pick-up device in itself, but the high definition by the tininess of the sensor has the mask that noises increase by the fall of the incident light quantity, and there is a limit in tininess. With a high-resolution monochrome signal and the color signal of the low resolution, examination was done about the technique to generate a high-resolution color signal for a long time in the field of the satellite photo. However, when high-definition image an agenda, the moire by the frequency that sampling frequency and the imaging thing have of the color sensor of the low resolution becomes a problem, but there is not the effect of the moire reduction by the technique suggested conventionally, and it is it with a problem as a picture. As a moire reduction method, I perform filter processing with the moire reduction effect for the color signal of the low resolution and suggest the method that the higher back definition handles. Because I confirmed a moire reduction effect by suggestion technique, I report it about the contents.

View full abstract

An integrated retrieval system for over a hundred databases have been developed by the National Institutes of the Humanities. Dublin Core and 5W1H meta-data have been applied as common meta-data to realize transversal retrieval for various kinds of humanities databases, These databases have temporal and spatial data. Data described by a period name and a place or regional name have been normalized to numerical data. By this way, it can be achieved to search for data-sets with a spatiotemporal condition and to display retrieval results on a chronological table and a geographical map.

View full abstract

In this article, we describe a case of utilization of Web-Learning system in “Regional Information Map” project. The regional information map includes welfare information for elderly person, disabled person, and also has useful information for daily life in Atsugi-shi, Kanagawa, Japan. In this project, we evaluated our e-Learning system “KBook”, which is kind of CMS (Course Management System), and we used this system for students to pre-learn what they do in an experiment of creating “Regional Information Map”. It was suggested that KBook is useful for learning about the field study of the project.

View full abstract

Recently, due to the increment of the number of product functions, the total pages of the manuals of such products are also increasing. It is necessary to understand the relation between product functions, parts and their operations. We propose an electronic operation manual that represents the relation of parts and operation visually using Web3D.

View full abstract

Brief report on DVD-Download specifications

View full abstract

We proposed the user driven network management and control architecture for the integrated management of home network and sensor network, and the network.

View full abstract

As a locomotion interface for walkthrough in virtual spaces, we have proposed the ball array treadmill　(BAT) with high cost and space performances. The walking velocity detection is important to keep the walker at the center of the walking board during the real walking. This paper describes a useful method for estimating the walking velocity from detected foot positions. When the velocity is evaluated from a walking distance at each sampling interval of foot positions, errors of the estimate are large and unstable. When the velocity is estimated at each walking period in consideration of the walking pattern, the velocity can be estimated stably with high accuracy. It is made clear that the method considering the walking pattern is useful for estimating the walking velocity in BAT.

View full abstract

In the locomotion interface for walkthrough in virtual spaces, how to detect foot positions is an important problem for cancelling the walker movement. This paper proposes a new approach for detection of foot positions where many hall-IC are arranged in array two-dimensionally under the walking board. The magnet plate is fixed on the bottom of the shoes or slipper. The foot position is estimated from the output signal distribution of the hall IC array taking the weight depending on the signal strength into consideration. This sensor array is very useful for detecting foot positions accurately and stably in the ball array treadmill.

View full abstract

This paper describes an impression evaluation of the CGV (Consumer Generated Video) still / video thumbnails. Our proposal, Choco-Para, automatically extracts emphasized speech portions to create skimmed content. Since audiences generally access to the original video after watching a still/video thumbnail on CGV sharing sites, his / her impressions of the thumbnail are very important. We conduct subjective evaluations using the SD scaling method and factor analysis. The results show that, compared to the center frame / portion selection approach, the video thumbnails output by Choco-Para create positive impressions.

View full abstract

Hearing-impaired students need complementary technologies, such as remote captioning. We explain remote captioning.A transcriber watchies the lecture as captured by camera and microphone and transcribes what an instructor says using a PC. The transcription is then sent to a display screen in the classroom. In remote captioning, a transcriber is difficult to assign a demonstrative speech phrase with a pointing object because he / she cannot look at the pointing object when the instructor utters the demonstrative speech phrase. Therefore, we are developing a system to assign the demonstrative speech phrase with the pointing object automatically and then display the pointing object when the transcriber input the demonstrative speech parase. In the paper, we describe a extraction method of the pointing gesture by a visual information.

View full abstract

This paper describes a study on prosody based multi-media content search interface.A number of related work which proposed a search interface based on sensitive word or similarity of pictures were reported. In this paer prosody patterns, which play important roles for emotional presentation, are utilized as inputkeys for searching. This approach enables users to search for 'similar emotional scene' such as 'upset-scene', 'sorrowful-scene', and 'angry-scene'. An assesment experiment was conducted with 18 subjects. The result confirmed the proposed method can be utilized for emotion based search interface.

View full abstract

Existing document creation tools, e.g. MS-Word, can provide some user interfaces for selecting typefaces of fonts. Unfortunately, however, a particular expertise on typography is essentially required to make an appropriate selection of typefaces employed in a document. This paper studies the functionalities to be supported by such a sophisticated user interface that a number of users without particular expertise can easily select appropriate typefaces in their documents.

View full abstract

Typeface identification of fonts in interchanging human readable documents is discussed from its structure and contents point of view and an appropriate guideline for editing and re-rendering the documents is provided. An effective use of the semantics of typeface information makes it feasible to recognize and understand the contents of documents.

View full abstract

Some typeface classifications are studied from several points of view. The mapping between those typeface classifications and the ISO/IEC 9541-1 typeface design grouping is discussed. It will be applied to extension to the scope of font information interchange.

View full abstract

This paper proposes a multimedia presentation for comic/cartoon documents. An effective use of electronic presentation media instead of the conventional paper media can make comic/cartoon documents much more impressive and active. The following presentation elements for comic/cartoon documents are proposed and their presentation effects are studied: a) successive display of frames, b) time-variant frame position, c) time-variant frame size, d) time-variant balloon characters (e.g., successive display, running, time-variant font size and typeface, etc.), e) color changing, and f) synchronization with BGM. This paper proposes a multimedia presentation for comic/cartoon documents. An effective use of electronic presentation media instead of the conventional paper media can make comic/cartoon documents much more impressive and active. The following presentation elements for comic/cartoon documents are proposed and their presentation effects are studied: a) successive display of frames, b) time-variant frame position, c) time-variant frame size, d) time-variant balloon characters (e.g., successive display, running, time-variant font size and typeface, etc.), e) color changing, and f) synchronization with BGM.

View full abstract

When multimedia information is rendered and presented, pets feel a strong interest in the information, and their attention will be caught by the multimedia equipment. Therefore, multimedia equipment should be more significantly designed than other equipment, and some design guidelines are required. Responding the requirements, this paper studies some behaviors of dogs to multimedia (audio, video, and etc.) information and proposes a design guideline for pet protection to/from multimedia equipment.

View full abstract

Image mosaicing for aerial photographs is an effective technology for recognition of wide area. Motion vector is needed for correct mosaicing, and it is estimated from aerial images. A conventional realtime image mosaicing function was archived by block matching method. In this report, we proposes a more efficient and fast algorithm of motion vector estimation. Realtime image mosaicing would be actualized by proposed method, for images were taken from straight-flying aircraft.

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

The case of therapy for the balance a deficit facial surface and piece of chin in an accident or a disease, make an artifact of copy area of body called facial prosthesis, often therapy in such a way that balance a deficit. To maintain the facial prosthesis stable need to fix part with a little movement of face.In this announcement propose restoration technique of face shape that changes time seriesto detect part with a little movement of face from an expression change.This technique uses the pattern projection method that used the multi slit that has the same space resolution as the gray code projection method to obtain minute shape of the face. The shape change is tracked by the multi pattern matching method based on the face shape obtained by the gray code projection method.Reduction in the mis-correspondence by the pattern matching method is achieved by the labeling method using the smoothness of the face.

View full abstract

The case of therapy for the balance a deficit facial surface and piece of chin in an accident or a disease, make an artifact of copy area of body called facial prosthesis, often therapy in such a way that balance a deficit. To maintain the facial prosthesis stable need to fix part with a little movement of face.In this announcement propose restoration technique of face shape that changes time seriesto detect part with a little movement of face from an expression change.This technique uses the pattern projection method that used the multi slit that has the same space resolution as the gray code projection method to obtain minute shape of the face. The shape change is tracked by the multi pattern matching method based on the face shape obtained by the gray code projection method.Reduction in the mis-correspondence by the pattern matching method is achieved by the labeling method using the smoothness of the face.

View full abstract

In this report, the parameter of proposed method is determined based on a reliability of detection results.Robustness of the proposed algorithm against disturbances such as image blur and white Gaussian noise is also discussed.

View full abstract

We proposed the method of water level detection of a river from an image using a camera. But practically all image width is not always be used because images depend on some circumstances, for example loss of transparency of lens.Therefore this paper is proposed how to estimate the minimum image width essential for detecting water level, using the statistics

View full abstract

The problem for deciding the optimum camera locations to cover all regions that should be observed with the minimum number of cameras becomes the set covering problem. In this report, we present the optimum allocation method of cameras to observe all the three dimensional small regions considering the visual angle and the visual distance of cameras. In addition, we show a method to reduce number of cameras by limiting observing regions to some essential regions.

View full abstract

We propose a method for producing water splash appearing on the ground or object's surface in a computer generated image.　In order to develop a water splash motion model which allows us to reflect the parameters of the water splash, they are, size of falling water drop, material of the surface, and depth of water on the surface, we had experiments for investigating a water drop colliding with an object under plural combinations of the above parameters and took movies with a high-speed camera. Based on the analysis of the movies, we developed a simple model for expressing water splash consisting of three types of motion patterns.

View full abstract

A conventional three dimensional (3D) rotation transform (RT) includes nine lifting steps since it has three 2D-RTs and each 2D-RT has three lifting steps. Lifting structure has an advantage that there is no loss between input value to the forward transform and output from backward even though multiplier signal value is truncated. This report proposes a three dimensional rotation transform with reduced lifting steps. It has the same transformed output signals and the number of lifting steps is reduced from 9 to 4.

View full abstract

This paper proposes an efficient data representation scheme to improve the performance of a data hiding method [1] in MPEG compressed domain. Reverse zerorun length (RZL) is proposed to efficiently encode message by considering the distance between two excited macroblocks. RZL simultaneously achieves high payload and high embedding efficiency, thus RZL is able to suppress the filesize increases during data embedding in [1]. RZL is compared to matrix encoding, and computer simulations are carried out to verify the basic performance of RZL.

View full abstract

[in Japanese]

View full abstract

The Luminance Contrast Sensitivity Function (LCSF) of human vision has been widely used in the field of print, public signs, and medicine. LCSF tells us how sensitive we are to various spatial frequencies of visual stimuli and measures the lowest contrast detectable at a given spatial frequency. However, relatively less work has been carried out to measure Colour Contrast Sensitivity Function (CCSF). In this study, we measured Contrast Sensitivity Function of complementary colour pairs (red and cyan). An experiment is made by cosine wave patterns created with red and cyan and done by changing their various parameters. We have got appreciable data by changing spatial frequency as well as contrast, and we redefined coloured gratings to the CCSF using the result data derived from the experiment of sight testing.

View full abstract

This paper describes parallel processing methods of background subtraction using a multi-core processor. In the intra-frame parallel processing, an image is divided into plural area and they are processed in parallel. However, there are several parts which cannot be parallelized, such as labeling and ghost distinction. In the inter-frame parallel processing, labeling and ghost distinction can be parallelized. However, the update of background is delayed. We describe the influence when the background changes abruptly and propose a method for reducing it.

View full abstract

Partial discharge and minute emission from sprout are observed under high electric field.

View full abstract

Exhibit contents: Stereoscopic display and exclusive use contents.

View full abstract

As the Stereoscopic television achievement method, Computer generated hologram data is examined to transmit the space in real time by white LED . It experimented on the data of one frame 76.8kbit at transmission speed 2kbps by having already used the SSTV method. The optical transmission experiment is examined by using the USB1.1 communication method to transmit a Computer generated hologram data by the high-speed.

View full abstract

[in Japanese]

View full abstract

Lossless video coding is indispensable in treating high quality video signals. Therefore, several lossless coding algorithms have been developed recently. A motion compensation prediction is one of the inter-frame prediction methods using correlations between frames, and the motion compensation prediction gives the biggest contribution on the compression performance.In this paper, we propose a method to encode a video sequence using multiple motion vectors. In this method, the sum of absolute difference is used to search for approximate blocks and the mean value of searched two blocks is adopted as the prediction value. Moreover, two prediction modes are introduced in our scheme, and the prediction mode is changed depending on the difference between two sums of absolute differences. Simulation results show that our sheme can reduce the average code length.

View full abstract

There have been increasing chances to exchange image data. In order to handle or manage images easily, it is desirable to embed the attribute information such as copyright ID of the image into the image itself. In our preceding study, Sugeno et al. proposed a method to incorporate data into an adaptive lossless coding method. In our method, coding parameters such as prediction functions are controlled by the embedding data and the original image data are never changed. However the method to expand the range of the selective prediction functions was never considered. This thesis addresses a coding method for embedding information, which improve the compression ratio by increasing the nember of candidate prediction functions . The conventional embedding method by block-based adaptive predictive coding uses two prediction functions in embedding information. However, depending on the statistical property of the block, the same compression ratio can be obtained by the other prediction functions. In this case, utilizing these functions can increase embeddable information than in the conventional method. Simulation results show that the method of utilizing these functions has a little improvement in some cases compared with the conventional adaptive predictive coding.

View full abstract

" Structured Model for Museum Object Information" developed by Tokyo National Museum in 2005 aimed to model museum object information in the field of humanities in order to help developing practical systems. Currently Tokyo National Museum has been developing a information system which suppoorts museum buisiness using the model. In this paper the author will introduce outline of the system and refers some issues in the future.

View full abstract

A classic classification passes also in the world of the painting. Though constructing the data base in the image has respect that is more difficult than the data base only in the character text In the world in the sphere of the culture of the Chinese character, moderate knowhow was accumulated. I would like to describe them about the example report(Historical[Classical]Character Database) applied to the retrieval of the image and relations with ontorogy.

View full abstract

We are developing a retrieval system for compiled historical sources. Due to retrieve for historical sources, it is necessity to study concerning two kind of very important things in advance to developing the system: one is to build a best method to digitize concerning sources with any kind of image data, and another more one is a function to automatic extract retrieval terms. And we are now studying to solve of their things using "History of Science of Japan before Meiji Period (The Japan Academy; the 28 series) ". Eventually we will digitalize all of series and offer through a retrieval system their data. In this paper, we propose concerning index term extracting method and digitization techniques for historical sources.

View full abstract

Evaluation at several aspects of a set of teaching materials on International Standardization was provided. The materials were developed and used by authors themselves to improve them as a public course in a series of 16 lecturers.

View full abstract