Emergency drills and drills against terrorism have been held at nuclear facilities. Shortcomings of these drills are that the numbers of emergency scenarios and roles are limited because many workers must participate in them simultaneously. The times during which all employees can participate simultaneously are limited. As a means of solving these problems, augmented reality (AR) training has been increasingly adopted in recent years, but difficulties of creating contents and the inadequate degrees of freedom in the created contents are becoming problematic. For this study, we developed an AR training environment construction system for users to create scenarios easily by simply assembling and placing blocks representing scenario components in a real-world environment. To realize the construction system, we also developed the “Program Block,” which enables users to create scenarios in a puzzle-like style. The system can create and simulate various scenarios involving interaction between trainees and a real-world environment. To evaluate the ease of system usage and flexibility for scenario creation, a subjective experiment was conducted at a nuclear facility. Results indicate that users can construct an AR training environment easily.

Hidden background cameras used to realize Di-minished Reality (DR), which allows the observer to see through obstacles, are installed behind the obstacles. In order to realize DR, it is necessary to measure the positions of these cameras with respect to the observer's position, but this is difficult because the cameras cannot be seen directly from the observer's position. In this study, two calibration methods, the intersection method and the P4P method, which use the directions of radio waves from radio sources mounted on each camera to measure the positions of the hidden background cameras, are proposed. The intersection method measures the direction from several different points to the hidden background cameras and estimate the 3D position of the hidden background camera as the intersection of the measured directions. The P4P method measures the relative pose between multiple hidden background cameras in advance and uses it as a constraint to estimate the 3D poses of the hidden background cameras. The results showed that the P4P method was more accurate than the intersection method, and able to calibrate the camera with a position error of less than 100 mm.

Free viewpoint video technology, which provides a high sense of presence and immersion by allowing users to view live video from any viewpoint, is increasingly demanded mainly in the entertainment field, as an application that provides new experiences for users. Although there have been many studies on free viewpoint video generation for large-scale spaces, they have focused on generating video for the entire wide-area space. For this reason, challenges still remain in terms of reproduction of object details and real-time video performance of image generation. In this study, for the purpose of developing a system which enables immersion into a moving miniature world, a method to generate high-definition free viewpoint video for small-scale spaces in real time was realized and evaluated. It can generate high-definition free viewpoint images by combining free viewpoint depth images applying noise reduction processing and the corresponding color images considering occlusion and the normal direction of the object's surface. The free viewpoint images generated by the proposed system was evaluated. The results confirmed that consideration of occlusion and the normal direction of the object's surface improved the quality of the generated images. In addition, the realized methods were able to generate clear free viewpoint images even when the free viewpoint approached the object.

With the advent of the information society in recent years, ideas are becoming more and more valuable, and divergent thinking is becoming more and more important. Although there are methods to present visual stimuli to improve divergent thinking, these methods mainly present visual stimuli to the central visual field. The purpose of this study is to test hypotheses about changes in the number and variety of divergent thinking ideas under conditions in which visual stimuli, called geometric Figure animations, are presented in the peripheral visual field. The evaluation experiment was conducted with the participation of 32 university students. We asked them to perform the Alternative Uses Test (AUT) under two conditions: presenting and non-presenting geometric Figure animations in the peripheral visual field. The performance of divergent thinking was evaluated by measuring the number of responses, fluency, and flexibility of the AUT. The scores were standardized because the AUT differed in difficulty depending on each AUT topic. The results showed that all scores were higher in the presenting condition than in the non-presenting condition, and the effect size of the mean difference in flexibility was moderate. On the other hand, according to the results of the questionnaire, most of the participants did not feel that the geometric Figure animations provided any cues for divergent thinking. This suggests that the flexibility of divergent thinking might be improved unconsciously by geometric Figure animations presented in the peripheral visual field.

3D reconstruction from images obtained by RGB-D camera provides a 3D model that reflects the current conditions of nuclear power plants (NPPs), and it is useful for field work support in NPPs. However, machinery equipment and pipes are intricately connected in NPPs and a lot of areas are not photographed because they are occluded by peripheral machines or pipes. It is feasible, but still challenging photographing those areas with RGB-D camera-mounted drones, due to the battery capability and the weight of drones.In this study, 3D reconstruction in cooperation with a lightweight mirror-mounted autonomous drone has been proposed and a novel method for determining drone’s flight path to scan those areas with a mirror has been developed. In this method, suitable locations for scanning the occluded areas with a mirror are detected and connected using multiple paths that are verified as safe to fly.The developed method was evaluated in the simulation environment. The results suggested that the developed method could obtain more points and photograph more occluded areas than the method of repeating random movements and the method to locally search for the optimal direction.

Recently, digital signages are widely utilized to provide information for tourists in plural languages, but it takes longer time to offer information in each language separately. Although simultaneous multilingual display is expected to reduce the time for presentation, there is no guideline for simultaneous multilingual display to keep its legibility. In order to estimate the effect of layout on legibility, the authors prepared contents which have different layout factors and conducted evaluated the legibility of contents with Thurstone’s pairwise comparison. Since the number of combinations in layout factors increases exponentially, the number of stimuli was reduced using an orthogonal table. As a result, significant differences were found in some layout factors and it was found that horizontal arrangement of images and sentences, left alignment in sentences, middle margin from screen edge and medium-sized images made contents more legible.

Pairwise comparison has been widely employed as an easy method for subjective evaluation of several candidates to develop new products or to improve service. In this paper, a method using response time was proposed to improve the accuracy of pairwise comparison. Firstly, a pairwise comparison experiment was conducted to investigate the relationship between response time and difficulty of judgement. Then, a method to improve the accuracy of calculated scales of the objects was proposed where the answer of the pairwise comparison was modified by using the response time and correction function. In this study, three correction functions were compared to find the difference of the accuracy improvement, and two of them showed significant improvement.

Up to this day, many types of 3D displays have been developed. However, there remain some challenges in the displays, for example, restriction of view points and vergence-accommodation conflict. A 3D display was proposed in the past using light diffusion within 3-dimensionally positioned micro voids processed in a crystal glass cuboid. It is able to present 3D objects to any number of simultaneous viewers at any viewpoint, without wearing any kind of special glass-like equipments. However, in order to successfully present 3D objects, the projection must be adjusted precisely for each individual micro void by investigating the correspondence between each projector pixel and micro voids. This process not only requires a great amount of user's effort but also consumes a lot of time. Hence, for this research, pixel-to-void matching methods which can automatically find out correspondence between each projector pixel and micro voids by using a camera were developed. These methods were verified in the experimental systems to be able to perform such matching procedure automatically and accurately. In addition, using multiple projectors, brightness and resolution of the display was improved. Two types of the displays were developed, one aimed to improve especially its brightness and the other aimed to improve especially its resolution. While the latter one resulted in higher resolution but in subjective evaluation experiment no improvement of visibilities was verified, the former one resulted in higher brightness and it was verified to have improved visibilities in subjective evaluation experiment.

As the number of international tourist has been rapidly increasing recently, they have more opportunities to experience different cultures. With the upcoming Olympic game in Tokyo, Japan 2020, digital signage is expected to be utilized for an effective information presentation device for them. This study focuses on the information presentation with four-frame comic for foreign tourists to improve cross-cultural understanding. A subject experiment was conducted to compare four information presentation methods including the four-frame comic from the viewpoint of attractiveness and easy memorization. Participants from various countries watched Japanese food tradition information with four information presentation methods (four-frame comic, photograph, illustration and video) and evaluated their impressions. As a result, it was found that four-frame comic was easy to memorize although its attractiveness was different depending on the participants. However, as the experiment was conducted in a controlled laboratory, it is necessary to conduct a field study using a digital signage in the city to investigate the effective information method.

The energy consumption of commercial and business sector has rapidly increased and saving energy in office building has been encouraged. However, it may not only affect comfort of office workers but also give negative effect on their intellectual productivity. So it is important to evaluate and keep or improve intellectual productivity and authors have been developed the quantitative and objective index of intellectual productivity. In this study, a new working state model including the idea of concentration depth and new index MCTR and CDI were developed from the proposed model. These new index were confirmed by the experiment of varied motivation and data analysis of the past experiment.

For optimizing the balance between intellectual productivity and energy use in office buildings, we should reveal the mechanism of intellectual productivity variation of office workers. For this purpose, the authors had proposed an intellectual productivity model. The model is a state transit model based on short pauses while working. This model can explain the productivity variation of most of subject experiment results. In some cases, however, the explanation is insufficient with previous 2-state transit model. In this study, the authors have proposed an improved model assuming 3 states. A subject experiment was conducted where illuminance on the desk and work motivation were controlled to vary their productivity. The experimental result was emulated with 2-state/3-state transition models and their accuracies were compared. As a result it was confirmed that the 3-state model simulated more accurately. It means the 3-state transition model can explain productivity variation better than the 2-state transition model.

Nuclear power plants (NPP) must be maintained periodically. The maintenance efficiency must be improved and human error must be reduced simultaneously to improve NPPs' competitive capability in electricity markets. Although Augmented Reality (AR) offers great possibilities to support NPP maintenance work, some difficulties exist for application of AR to actual work support because current AR systems cannot be implemented in NPP environments without technical improvement. Problems exist such as recognition distance, tracking accuracy, and a complex working environment when applying AR to NPP field work support. Considerable extension of tracking distance and improvement of accuracy are particularly desired because NPPs are large-scale indoor environments. This study designed a linecode marker, a new type of paper-based marker, along with recognition and tracking algorithms for it to resolve these problems. In contrast to conventional paper-based markers, such as square markers and circle markers, the linecode marker is not merely easier to set up in complex industrial environments: it also enables the use of AR in industrial plants because of its considerable tracking-performance improvement. To evaluate tracking accuracy, the trackable distance, and the tracking speed of the proposed tracking method, an evaluation experiment was conducted in a large room. The experiment results show that the tracking distance is extended extremely over that of the traditional marker-based tracking method: tracking accuracy improved to 20 cm in 10 m distance. The running speed can be as fast as 15 frames per second with a laptop.

A real-time cognitive process estimation system has been developed, by employing machine speech recognition to estimate operators' thinking processes. The developed system consists of a user-interface, speech recognition subsystem, real-time cognitive process estimator, and nuclear power plant (NPP) simulator. The major point of the real-time estimator is that it equipped with the causal relation diagram (CRD) which presents the physical relations between the NPP parameters on man-machine interface (MMI). By using CRD and two types of information, one from data of operational sequence to MMI and the other from verbal information taken by speech recognition subsystem, this estimator computes degree of operators' confidence in judging primary cause of plant anomaly for each plant parameters of CRD. From a experiment, it was confirmed that the developed system would be usable for real-time monitoring operators' thinking process in diagnosing the plant anomaly at MMI.

Head-attached interface device (HIDE) has been developed as a new type of human interface device in an actual work environment. The HIDE has flexible integrated functions which consist of speech recognition and view direction detection as hand-free input channels, direct presentation of audio/visual information as output channels, and mobile system configuration. In this paper, the design concept of the HIDE was first introduced, and individual functions to configure the design concept were developed by their separate testing. An integrated prototype system HIDE was then produced, and its functional evaluation test was made to examine the individual functions. According to the evaluation result, we proposed a design guideline of application software, and developed an example software.