3DCG-related technologies such as 3D printer, or head mound display device has been beginning to be applied to medicine for getting an organ’s model, making it easier to understand the structure of an organ, for example. I will introduce the pros and cons of 3DCG and discuss if 3DCG really helps medicine.

View full abstract

We have investigated a method for displaying an image of scenery with a high dynamic range of intensity to a binocular display device to be able to give a viewer an impression of a high contrast environment. We have constructed two types of image dataset by using different procedures of generating low dynamic range images from a high dynamic range panorama image. A set of two images in the two types of dataset is displayed to each eye. Our experiments show the proposed method is useful for perceiving more visual content than just displaying a tone-mapped image in the same procedure to both eyes in looking around a surrounding scenery.

View full abstract

LED lenses used for a backlightunit of a LCD panel, a signboard, etc. have functions of saving energy and thinning the size of a device. Using a genetic algorithm, we optimize the shape of an LED lens modelled by using a Bezier curve to obtain a uniform and bright illumination distribution in a surface to be illuminated.

View full abstract

Road situations such as pavement types and road surface conditions have a great effect on driving a car. The automatic grasping these situations is required to realize the autonomous driving system. In this research, we consider the index of discriminating pavement type from images captured by driving recorder. We propose a method for texture analysis using the di vision of a pavement surface into plural areas considering image characteristics of each pavement and its difference We show the effectiveness of the proposal method by confirming characteristics of an asphalt and interlocking block pavement.

View full abstract

Users of smartphones and/or tablet computers browse and download confidential document files routinely. Therefore, the higher security level is needed for smartphones and tablet computers than conventional mobile phones (feature phones). From this kind of background, Kosugi et al. proposed SWIPASS, an image-based user authentication method for touch screen devices by using images shot by user oneself and gotten from the WEB. SWIPASS has resistance to observation attack and recording attack, the most serious threats for touch screen devices. However, there is an issue such that the fake image, that is always become the swipe object, is comparatively easily recognized by attackers. Then, in this study, we propose a method that prevents identification of the fake image by attackers by blurring images that appear on authentication screen. We moreover conduct several experiments to verify the effectiveness of the proposed method.

View full abstract

Recently, monitoring cameras are widely prevalent . But if we install them outside, sometimes raindrops or mud stick to their lens protecting glass and they appear as a noise in photograph .In this research, we propose a method that can remove noises better than conventional method. In conventional method, we take photographs while turning the camera and detect noises by considering the relationship between angles of photographing and objects in the images. But in this method, if the luminance of the objects and noises are so close, we can t detect noises exactly. So we solve the problem by using color information and clustering.

View full abstract

Recently, professional animation production techniques and its research has been noted. In the workflow of existing commercial animation production, it remains still many hand-drawing by the paper and pencil. In addition, its expression and workflow may have identity and advantage but there are less academic evaluation. In this study, we focused on developing tools to support the animation production for hand-drawing by the paper and pencil. As a first basic data, we have evaluated specification of current scanner products available in the market. As a result, sheet-feed scanners got good results in the pixel size errors, image quality and speed. However, they are not acceptable to evaluate not available from the point of view of ethics in actual workflow because of the possibility of damage to the original picture and paper by rollers, remaining of pencil powder inside of the scanner. The paper should be handle as a fine art papers and it should not be touched in the actual animation production workflow. On the other hand, non-contact type of stand scanner got worst result in image quality but it had been recognized as enough in the workflow.

View full abstract

Cached video content management methods (CVCMM) by using priority estimation have some problems of decreasing the efficiency for the change of the access numbers in video content delivery system with multiple quality. Then, the improvement method of CVCMM is proposed for them In this study, in particular, since the priority is calculated less than the actual value by the access change due to the estimation errors in the case of decreasing in a monotone manner, the caching efficiency decreases. Approximation estimation methods by the quadratic function and the exponential function are used. In addition, the estimation value is compared with the content priority of the lowest rank in the previous period, and that value is determined as the threshold value. By simulation experiments, it is revealed that the proposed method is more effective than the conventional method.

View full abstract

In recent years, the shape analysis of the 3D point cloud has become an important research theme of computer vision.This process is expected applying to the modeling, the object recognition and the localization, however mainly requires huge computation time for amount of data of point cloud. In order to solve this issue, the approach to extract a small number of feature points having repeatability and distinctiveness is effective. On the other hand, the performance of conventional feature point is strongly dependent on the sensor noise and resolution, because its extraction method selects actual points.We proposed a novel method to allocate the keypoints from multiple plane equations to a virtual location without selecting from actual points. In this report, we proposed a novel feature description method that added rotation invariant. Moreover, we confirmed that proposed method is superior than conventional method by performance evaluation of keypoint detection and feature description.

View full abstract

In regression testing, it takes much costs to check whether screens are displayed normally. In the conventional way, Bit map checks are used. Then, we have to check the screens displayed abnormally.We propose a method on extraction of different pixels and show the results of the evaluation.

View full abstract

The purpose of this study is to clarify the basic characteristics of the recognition of peripheral vision under conditions expanding circle image . Experiments measured the time when the subject noticed a change under expanding circle image . Time for recognition was found of be longer in the vertical direction than the horizontal direction . In the vertical direction , the upper region was found to be longer recognition time than the lower region .

View full abstract

This study analyzed the eye-movement during watching TV, for number of occurrences of person, character, cut image and view angle. The content used the TVCM to switch the screen in a short period of time.The research purpose is to determine the creation guideline when creating large-screen content. The experimental results are as following ; According to increase of the number of cut image, the eye-movements were concentrated in the center of the screen. ; The eye-movement tended to move on from the center.; As viewing angles increase, amount of eye-movement becomes more.

View full abstract

This paper proposes a method of example based segmentation using detection of region boundary to segment the regions such as a city, a mountain, and the river in satellite images. These regions are classified as texture regions. Example based Segmentation is performed with characteristic models based on texture analysis by using local fractal dimension . Unfortunately, this method segments the region in the boundary with a low accuracy. To solve this problem, it is necessary to detect the region boundary. The region boundary is detected by using Canny operation and mathematical morphology. The output image obtained by integrating the example based segmentation image with the image in which the region boundary is detected improves accuracy of the segmentation in the region boundary. We show the effectiveness of the proposed method in comparison with the correct image we make with a manual from the input image.

View full abstract

The purpose of this paper is to introduce a Kinect goggle syste m that helps a visually impaired user detect obstacles and recognize objects, such as chairs or elevator doors, from daily indoor environments. Through a keypad, the user can specify which object the system should search for. If the object is found, audio feedback will be provided. Therefore, the output can be restrained only to the chosen object, which means a more efficient way of identifying the desired object. The system was evaluated through object recognition tests in real office environments.

View full abstract

Road vanishing point could be used as clues to driving directions in automatic driving system. Therefore, we proposed a novel data-driven method based on convolutional neural network in order to detect the location of vanishing point of either straight or curved roads on each single given image input. Our mainly contributions are: 1. We bring advantages of deep learning technologies to vanishing point localization problem instead of using traditional manually designed texture filters; 2. a coarse-to-fine approach is proposed through cascaded heat map output CNNs. For evaluation purpose, we manually collect and annotate about 2100 road images. The experiment result on this dataset proves the robustness of our approach on almost any kinds of road type. Next we are going to integrate and test our implementation on an on-borad camera in order to perform precise navigation for robotic system such as unmanned aerial vehicle (UAV).

View full abstract

3D Plane Detection in Point Clouds has many important applications, such as Object Recognition and Robot Navigation. The state-of-the-art approaches are namely RANSAC and Randomized Hough Transform, although, due to their non deterministic nature, the finishing conditions have to be chosen carefully, hence, these approaches tend to not have a defined processing time . Therefore, we propose a method that speeds up the plane detection process in a deterministic way by looking at the centroids of individual patches of the Point Cloud. Then, a refinement of the decision is made by looking at the eigenvalues of each patch. Consequently, Hough voting is performed to the plane calculated from the coplanar patch. The experiment results have demonstrated that this method outperforms state-of-the-art approaches while being deterministic.

View full abstract

As a method for measuring the depth of micro-size objects, “Depth from Focus” (DFF) is known as an effective method. DFF methods evaluate focus from the texture sharpness of object surfaces. However, the sizes of texture pattern change the considering various sizes of texture patterns, filter window sizes have to be appropriately selected. Additionally, reducing bad effects caused by large edges is needed. In this paper , we propose a DFF method that adaptively selects window sizes depending on texture patterns.

View full abstract

A lot of security cameras are installed in the station and the downtown area in recent years. However blind areas where no camera covers that area are left when a broad territory is watched (sparse distributed cameras). So it is necessary to estimate the correspondence of moving objects taken by different cameras .The appearance sometimes changes drastically by the illumination change The proposed method uses the visual feature and spatial temporal information to estimate the correspondence of the moving objects. The optimum correspondence is obtained by solving Mixed Integer Programming ( MIP). It shows that robust correspondence of moving objects can be estimated under the various environmental changes.

View full abstract

In previous studies, we proposed a data structure that can order user interaction into a time series, and an AR(augmented reality), MR(mixed reality) based collaborative modeling system. In this study, we are going to propose an interface system to operate the data of 3D model by using a gesture sensor.

View full abstract

In cartoon line drawing of Non-Photo realistic rendering, our approach uses image processing for line detection and model shape processing for rendering. This technique can detect the lines with high accuracy, can also change the line thickness on real-time rendering. In this paper, we compared G-Buffer texture formats for acceleration of this approach.

View full abstract

This paper proposed a welfare recreation system to allow senior citizens or impaired people (i.e.　users) to enjoy virtual walk in their houses. First, the users or their supports (such as their families) record videos along their walk courses. If the courses are bifurcated, videos are recorded from the bifurcation points. The users can enjoy virtual walk by watching the videos that are played back according to the mouse button presses of the users. If the users are at the virtual bifurcation points in the videos, they can choose the courses with the mouse button presses. Several experimental results are shown.

View full abstract

Diffracted field of the electromagnetic wave in a pyramidal horn antenna is calculated under the condition that Fresnel's approximation is formed. As a result, a complex hologram is obtained by diffracted field as a simulation by a computer. It's thought that this is one of the computer generated holograms. We showed the expression that Fresnel approximation solution by spectral method agrees with analytic solution in this report.

View full abstract

Diffracted field of a electromagnetic wave in a pyramidal horn antenna is obtained by spectral method. The formulation obtained by this report can be applied at the wide region including the near-field. We considered together about the Fresnel approximation of the diffracted field and the Fraunhofer approximate condition by the direct integration method.

View full abstract

We have experimented cat face detection by searching face components such as eye, nose and mouth for cat face candidate regions detected by a classifier constructed with Haar-like feature. In this paper, we describe how to improve the detection accuracy by increasing learning images with their rotation. As the result of the experiments, we have succeeded in detecting cat faces that have rotation, and the precision and the recall were 100% and 88%, respectively.

View full abstract

In this paper, we propose a method to objectively find plagiarism of logo using SIFT Features. SIFT feature is robust to the scale change, but the detection position of key point change by the image resolution. Proposed method makes it possible to allow detection of key points by arranging a plurality for simple logo which is difficult to detection of the key points.

View full abstract

Correction for perspective distortions of Japanese characters on signboards is discussed. The perspective distortions is a technical issue, when we take a photo of the signboards obliquely. Two vanishing point s generated by horizontal and vertical lines on the Japanese characters these vanishing points are computed using the Radon transform. A homography matrix for our correcting is determined , and we correct the perspective distortions. We confirm our correction and its accuracy in our experiments. In a comparison with the conventional methods , our proposed method has the best performance of all.

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

[in Japanese]

View full abstract

Major international standardization activities on AAL (active assisted living) taken by IEC TC 100 during 2015 and 2016 are shown and some future issues are discussed.

View full abstract

[in Japanese]

View full abstract

HTML5 + CSS is the key of the one-source multi-use publishing.

View full abstract

In this paper I described the explanation of IEC International Standard about Texture map for auditory presentation of printed texts. In IEC62665 Ed2.0 the process how to generate the texture map from texts and how to synthesis audio are explained. In IEC 62875 Ed1.0 the printing specification of the texture map for auditory presentation of printed texts are explained from the point of view of the quality of printing of texture map.

View full abstract

The “ barrier -free movie” , and people who suffer from a variety of access barriers in order to watch a movie , that movie create an environment so that both can enjoy the movie . Since 2001 , and tell activities and of City Lights , which has supported the movie of the visually impaired , the current situation , such as know-how and delivery method of “movie voice guide ”, which began to build through its activities , permanent Universal Theater It tells the dream of the establishment .

View full abstract

Introduction of International Standardization activities related to e-book in IEC TC100/TA10.

View full abstract

IEC TA10 published a technical report to introduce various information, including market situation, latest related technologies, regarding digital sheet music market. This presentation material explains background and summary of the technical report.

View full abstract

IEC 62605 is an international standard that defines the interchange format for e-dictionaries. This talk describes the main points of its second edition published in February 2016 and its structure. It also discusses the significance of such formats as well as the growth and outreach of the e-dictionary business.

View full abstract

[in Japanese]

View full abstract

MIDI was formulated in 1982. MIDI is the protocol for connection between musical instruments to the other one. Along with some standardization, the platform that MIDI is able to use is expanded. And types of transmission path are increased. Recently by using Web MID API, a browser application can directly use MIDI. In addition, We will make MIDI world standard to be used in world wide.

View full abstract

Some case study of digital exhibitions of the National Museum of Japanese History are introduced. Relationship between museum exhibition and digital technology to produce academic entertainment is also commented.

View full abstract

This decade saw an unprecedented increase in exhibitions and archival activities on manga, anime and games, an area which used to be outside the traditional scope of museum facilities. Recent approaches and attempts shall be discussed in the presentation.

View full abstract

A lot of Japanese companies think standardization along with products already they have. Thus they are stuck in dilemma that they have not been able to apply their thoughts for standardization to requirements by the development of IoT and ICT. One of the reasons is that they could not connect accordingly their standardization strategy with business one Based on the research of making standardization reality for some companies, this paper discuss some hints to escape this dilemma.

View full abstract

[in Japanese]

View full abstract

I would like to share 1) our leadership development process, mechanism, and practices in IBM, 2) characteristics of leaders in transformational stage, and 3) key insights and takeaways we’ve obtained through our long-term leadership development practices

View full abstract

Introduction of the current situation of international standardization activities in the field of ICT, including hot issues such as IoT/M2M, trend of related organizations and fora, government efforts, and the considerable measures for human resources development for the activities.

View full abstract

IIEEJ SIG on International Standardization Education’s activities from its establishment in 2007 are reviewed in terms of numbers of participants, web page access hits, and paper titles. Also reviews current status of education about standardization along with the problems that have been issued within the management group for SIG. We provide several points for discussion for how to manage the SIG in the coming yearas.

View full abstract

Research topic on smart architecture using ICT is introduced. Monitoring project in battle ship island and development of autonomous time synchronization sensing system by chip scale atomic clock are presented.

View full abstract

[in Japanese]

View full abstract