As application field of future digital signage, two major fields have been introduced and investigated. One is the emergency information provision such as emergency and disaster situation, while another is the provision of tourism and cultural information services as extension field of the existing advertising and information services, which also are the introduction of the following four lectures in the session.

View full abstract

This paper introduces metadata technologies and use cases, and discusses to apply to regional services.

View full abstract

Because civil servants are tend to be disaster victims at natural disaster, municipal governments are accelerating alliances with other municipal governments. However, it’s not easy for estimating how much help is required before the disaster since it depends on the level and domain of the disaster. Therefore, this paper focuses on the tourism as one of the solutions in term of mutual benefits, referring the case of Izu peninsula.

View full abstract

Digital signage is one of the promising technologies which is driven by MIC and widely adopted in a variety of business sector, getting its market size greater. A disruptive technology known as blockchain has been studied and developed since Nakamoto’s paper, and now is the time to begin to find some working cases. In this paper, We toss around what happens when these two interesting technologies are mixed based on rough thoughts.

View full abstract

Captioning movies using audio communication technology and smartphones

View full abstract

Satellite laser communications have gained a lot attention in the recent years with increasing number of successful demonstrations by space agencies, academia and private sector for both low-Earth orbit (LEO) and geostationary orbit (GEO) satellites. For further technology development and wider commercialization there are a number of specific problems still to be addressed. Nowadays, the artificial intelligence, specifically the deep-learning techniques show incredible results for multiple complex tasks and problems in a variety of fields. In this paper, some of the main problems for the satellite laser communications technology will be explained and possible solutions using deep-learning will be discussed.

View full abstract

There are many ways to use natural energy, such as solar panel, Speaking of wind power, a large propeller type is famous, but a small, noise-free vertical axis type is explained.

View full abstract

[in Japanese]

View full abstract

In this paper, regarding to evacuation behavior when a disaster occurs during sightseeing behavior, examination results on effectiveness of installation of digital signage are discussed using multi-agent simulations.

View full abstract

In this paper, we applyed time-sharing displaying process to reconstruct the holographic images with the use of blue-violet laser light source. As a result, we found that our method suggests the possibility of good results to two color holographic multiplex image reconstruction.

View full abstract

We evaluate the characteristics of multiplexing that does not require an optical system, using multiplexing using a lens term and separation of a reconstructed image by simulating the lens term. As multiplexing technology, the method by the redundancy of the information of the hologram is well known. However, most of the methods use optical systems for separation, and there is no one that shows the feasibility of holographic memory by simulating lens terms. Therefore, the basic experiment of separation without using optical system is conducted to confirm the problems, and the application to holographic memory is proposed.

View full abstract

In this paper, we introduce our study related to computational illumination with multispectral light stage, in particular, reflection separation, image-based modeling, and material recognition.

View full abstract

In this study, we investigate the possibility to distinguish elastic and collagen fibers from H&E stained images by applying hyperspectral image analysis based on Linear Discriminant Analysis. With experiments, we showed that the hyperspectral images could perform the classification better than RGB images.

View full abstract

[in Japanese]

View full abstract

In this paper, we propose a method of compensating for temporal variation in environmental illumination for whisk-broom hyperspectral imaging. In addition to the 2d scan, we propose to execute an additional 1d scan orthogonal to the scanline as a reference for the compensation. Our method utilizes a low dimensional structure in the spectral property of illumination to robustly compensate for the temporal inconsistent illumination. We quantitatively evaluate the compensation methods through an experiment under a controllable illumination environment and present an experiment conducted in the wild.

View full abstract

In hyperspectral imaging, we sometimes obtain the saturated spectral signals in certain band or bands due to the limitation of the dynamic range of camera sensors. These saturated bands clip the real spectral signature and lead to spectral distortion, thus infect the object detection accuracy or the color reproduction of objects. In this paper, we propose a simple reconstruction method for saturated signals using principal component analysis. Experiments on real hyperspectral data show that the restoration of saturated signal helps increasing the unsupervised object detection and improving the high-fidelity color reproduction.

View full abstract

In this paper, we propose a method to estimate the spectral reflectance of bare skin and skin coated with foundation from the skin image taken with RGB color camera. In general, estimation of spectral reflectance from RGB data requires estimation of high-dimensional information from low dimensions, so some constraint conditions are required. In this study, first, spectral reflectance data of 741 skins of men and women aged 18 to 48, mainly in their 20s, are measured using a spectrophotometer, and spectral reflectances of the skin using singular value decomposition Reduce the dimension of and analyze its characteristics. As a result of examining the cumulative contribution of the basis functions, it was found that the spectral reflectance of the skin can be expressed with almost three basis functions. Based on the characteristics, we estimate spectral reflectance of skin from RGB color image. Then, using the proposed method, experiments were performed to estimate the spectral reflectance of skin from RGB color images using three subjects. As a result, when spectral reflectance of skin was estimated, it turned out that spectral reflectance of skin can be estimated from RGB color image.

View full abstract