Vol.2 No.2 2009

28/98

Research paper : A secure and reliable next generation mobility (Y. Satoh et al.)−108−Synthesiology - English edition Vol.2 No.2 (2009) important to transmit the information to society through papers, the press release[3], and exhibitions as an integrated and synthesized system, rather than to develop individual elemental technologies, and we decided to take the following research strategy (diagram of the model is shown in Fig. 1).(1) Evaluate and publicize the prototype by creating it as quickly as possible, through the integration and combination of the necessary elemental technologies.(2) Seek “elemental technologies that must be created or those that must be improved,” based on our evaluation results and the external comments obtained by publicizing.(3) Research developments and the improvements of the elemental technologies.(4) Return to (1).In many cases, our researches went along the path from the creation of elemental technology, through the sophisticating stage, reaching a point concerning its applications, which is an upside-down strategy to that mentioned above. This is an attempt to consecutively create the necessary Type 1 Basic Research from the starting point of Type 2 Basic Research [4] where the elements are integrated and synthesized.In this research, we decided not to be bound by the framework of a conventional “wheelchair” but decided to engage in R&D for a new personal mobility that can be used by any of us. Development of a device for transporting a person while sharing space with pedestrians, unlike an automobile that runs on a road carrying multiple passengers, is important in allowing all people to travel with minimum energy consumption and low pollution. Also, a large market would be developed in the future by targeting wide user range. The larger market will result in higher performance, lower cost, and better infrastructure designed for the passage of new mobility. As a result, the users of current electric wheelchairs will benefit from such developments.For the new mobility to run safely while sharing space with pedestrians both indoors and outdoors, it is important to have a technology to sense the surrounding environment quickly and accurately and a technology to accurately detect the risks from the information obtained. In this research, by installing a stereo omni-directional camera and by advanced computer vision technologies mounted on an electric wheelchair, we attempted to realize mobility with functions that prevent collision with pedestrians or obstacles and prevent falling or capsizing from level differences.3 Selection of elementsWe shall describe the elemental technologies selected to realize the research objectives. This includes the elemental technologies that were added when we found they were necessary after the integration and synthesis of elements which is explained in chapter 4 later.Figure 2 shows the selected elemental technologies. The function for actively sensing and recognizing the surrounding environment is realized by mounting the following onboard an electric wheelchair: (1) “sensing technology” to obtain information of the surrounding environment, (2) “cognitive technology” to detect danger in the environment from the obtained environmental information, and (3) “interface” that presents the information to the user. The individual elemental technologies are described in the following sections.3.1 Stereo omni-directional cameraThe stereo omni-directional system (SOS)[5]-[7] was developed by Satoh (one of the authors) et al., in the Human and Object Interaction Processing (HOIP) Project, Collaboration of Regional Entities for the Advancement of Technological Excellence in Gifu Prefecture, Japan Science and Technology Agency. It is an innovative camera system with a function to obtain omni-directional color image and 3D information simultaneously and in real time by assembling 36 cameras in a spherical form.To obtain wide-range images, a camera system using a wide-angle lens and/or a parabolic mirror[8]-[10] was used conventionally. However, since it shot a wide range using just a single lens and imaging device, the optical resolution particularly of the lens was inadequate, and the resolution of the image obtained was insufficient. In contrast, the stereo omni-directional camera shoots in all directions using a large number of cameras, and high resolution can be maintained overall even if low-cost general-use lenses are used. Also, 3D information can be obtained easily by using the parallax between the cameras.Fig. 1 Model of the research strategy.Evaluation and presentation by completing the prototype as fast as possibleSeek “elemental technologies that must be created and elemental technologies that must be improved” based on the evaluation results and external opinionsConduct research to improve elemental technologyType 1 Basic ResearchType 2 Basic ResearchUpward spiralStarting from Type 2 Basic Research, the necessaryType 1 Basic Research is created in successionFig. 2 Selection of elements.RecognitiontechnologyDanger detectionprocessingGesture recognitionprocessingInformation displayinterfaceSensing technologyStereo omni-directionalcameraStereo imageprocessingImage integrationprocessingEstimation ofcamerahead positionInterface

※このページを正しく表示するにはFlashPlayer9以上が必要です