Vol.5 No.3 2012

6/94

Research paper : Developing an evaluation system of visually induced motion sickness for safe usage of moving images (H. Ujike)−141−Synthesiology - English edition Vol.5 No.3 (2012) the condition under which VIMS tends to occur. The understanding of VIMS is expected to deepen, if the degree and the temporal variation of VIMS that are generally difficult to understand are made “visible” by repeated investigation of various images using the system.Second, this will be a method for objectively evaluating the image that will be produced or distributed by the image providers. Since there are individual differences in the susceptibility to VIMS, as discussed later, and people are likely to adapt, it is difficult to evaluate by having certain individuals view images. Therefore, a system that allows objective evaluation of images is necessary.Third, this enables the specific measures to be taken for the images created by the people involved in the production. If the degree and the time when VIMS is likely to occur are known, the images can be edited based on such information. Therefore, it provides a method for realizing image safety.In the creation and standardization of the guideline, the VIMS evaluation system can be a powerful tool to verify the effectiveness of the guideline. In the discussions at ISO,[6] the author has been working on the strategy of international standardization of the guideline based on the basic data obtained using the simple visual motion (see subchapter 4.1). This system is essential to show that the guideline based on the basic experiment data can be applied to general images that contain complex visual motion. Also, in the standardization discussion, an unnecessarily strict standard based on ideals may be suggested, but such standards may not be followable in reality. This system eliminates such useless standards and helps to create a guideline that can be followed practically yet fulfills the minimum requirements, by evaluating the images under various conditions. Moreover, this system provides the method for checking the compliance by the image providers themselves with the image guideline after it is issued. By analyzing and evaluating images using this system based on the guideline and “visualizing” the degree of VIMS for individual images, the policy for production and editing will be easier to establish.We had opportunities to communicate the information on the VIMS evaluation system to the people of related industries in various occasions, and we received voices from many people that they wanted to try it. For example, the reduction of VIMS is actually a concern of the image producers, and while they think they understand from experience how to limit certain motions, they would like to check objectively. Therefore, VIMS evaluation system is highly in demand by people who understand the importance of reducing VIMS.3 Scenario of the R&DThe development of the VIMS evaluation system was done as the “Feasibility Study for the Development (FY 2006; realization in FY 2007) of the Visually-Induced Motion Sickness Guideline Verification System” by the Japan Electronics and Information Technology Industries Association (2006~2007), as a subcontract of The Mechanical Social Systems Foundation.[7] In the R&D committee, image media companies and filmmakers who were interested in VIMS as well as the university researchers of biomedical effect gathered, and we were able to obtain cooperation from their respective standpoints through specific collaborations and joint research in the actual execution of the R&D. For example, as will be explained later, there were cooperative efforts in speeding up the visual global motion analysis essential for the system, and the production of images that readily causes sickness by the image producers. Through such cooperative efforts, we were able to develop a highly effective VIMS evaluation system that surpassed the initial goal.The composition of the system is shown in Fig. 2. Of the components, the VIMS evaluation model involves the output based on the results of measured biomedical effects using the simple visual global motion,[8]-[10] and it does not necessarily guarantee an immediately adequate output for general images in which the visual global motion are complexly mixed. Therefore, it was necessary to calibrate the estimated results of VIMS with the actual biomedical effect measurement, and the following procedures were employed for the R&D (Fig. 3).(1)The whole system is configured.(2)The images that readily cause sickness are prepared. These are entered into the system to obtain the evaluation results, and at the same time, the biomedical effects are measured in a viewing experiment using the same images, and the two results are compared.(3)Based on the comparative results, the components of the system are improved to increase the accuracy of the Fig. 2 Configuration of the VIMS evaluation systemMoving imagesOutput of moving image and evaluation resultsVIMS evaluation modelIntegration of outputTime-series outputBasic data of biomedical effectsVisual global motion analysisGlobal motion vector analysisLocal motion vector analysisImage inputInterface

※このページを正しく表示するにはFlashPlayer9以上が必要です