Vol.2 No.4 2010

5/68

Research paper : Development of a real-time all-in-focus microscope (K. Ohba)−235−Synthesiology - English edition Vol.2 No.4 (2010) only in retrospect can we say certain places were the forks in the road. While I was groping desperately, honestly speaking, I cannot say I did any strategic decision-making even in retrospect. Even if a theoretical order was given for the twists and turns, the situation differed greatly in the next blind wandering, and there were very few instances when I felt the previous experiences helped at all. Therefore, our history of twists and turns (Fig. 2) will be presented as an article that documents a process of product realization, not in the format of an academic thesis, and I shall explain the phases from a technological standpoint in the following sections.(a) Idea phaseWhen a person looks at an object with the naked eye, the objects both close and afar seem to be in focus. This is accomplished by the focal adjustment of the eyes. In contrast, when one sees things through the lens such as of a camera, it is necessary to bring the lens into focus. The automatic focus camera sets the focus automatically, but it can do so at certain distance only.In a microscopic environment, the depth of field can be increased by narrowing the aperture in the low-power single-lens reflex cameras and stereomicroscopes, but it is impossible to cover all areas as magnification increases due to the optical principle. In microscopic objects, the depth of the objects cannot be discerned through images with deep depth of field. Therefore, we thought of increasing the operability in the microenvironment by making use of the shallow depth of field. To increase the operability in the microenvironment, we decided to create a system that fulfills the following two conditions simultaneously, and conducted theoretical considerations and devised processing algorithms:1. Dynamic observation (30 frames/sec.) of real image with depth of field ideally raised to infinity2. Real time measurement (30 frames/sec.) of three-dimensional shape of the object At this time, we were vaguely thinking that it may be interesting to have this function in the digital camera or eyeglasses, and had no idea about any specific device application.The above conditions 1) and 2) were not stated as initial target specifications, but as we were looking through papers for methods to obtain the image of 1) using the image processing technology, we realized that the 3D position information of objects were not used in the course of processing. Therefore, we thought why not use this information? The details will be discussed later.Fig. 2 Our struggle (Twists and turns) for synthesis.LogicVarifocal lensFocal distance movement mechanism (commercially available)High-speed image capture deviceFPGALVDSVision chipParallel processingSequential processingfocusing applicationfocusing applicationIdea phaseProduct phaseFirst FS phaseSecond FS phaseThird FS phaseMeet PhotronMeet Denso Co. at the Micromachine exhibitMeet Delft High Tech and Kawatetsu Techno Research・Digital camera・Consider usage in eyeglasses・Specialize in microscopes・Industrial usage・Biological usage【Speed-up】・Search for partner company to develop vision chips・Vision chip→FPGA+LVDS【Precision increase】・Search for actuator for microscope【Business】・Build sales channels, etc. (Photron Co.)First obstacleSecond obstacleThird obstacle19971998200020012003ManufactureType 1 Basic ResearchType 2 Basic ResearchInformative configurationSystem configurationThe output systemsHigh-speed variable focus mechanismHigh-speed image capture mechanism / high-speed image processing arithmetic circuitHigh-speed communicationPrototype 1Prototype 2Microscope systemProduct system

※このページを正しく表示するにはFlashPlayer9以上が必要です