the largest university in Japan. We have 100,000 students including high school students and elementary school students. And we have 14 colleges and more than 20 campuses spreaded

mainly around Tokyo. And our courage is the courage of science and technology. Just a single college but has 10,000 students. And this is just a photograph of our campus. And

here is a list of my research projects. I'm working on the Landcure 3D portrait system and I'm mainly working on the digital holography. Computer generated hologram, fringe printer or CGH. Fringe printer is today's topic. And I'm also working on holographic

video system and the simulation of the image reconstruction. And of course I'm working on the optical holography but still the main topic is digital holography but I'm enjoying work with optical holography. And sometimes I support artists to make their

holograms. Okay, here I have three color lasers so I can make full color hologram but I know it's very difficult, hard work. And I also have the pulse laser. Actually it's Giora's camera so even I can make a portrait. And here I would talk about the

fringe printer. Here is the optical hologram. As you might know that the hologram record intensity and the direction of light employing interference. And here is how to compute

the Fresnel hologram. It's just simulating physical phenomena in computer. And in our case, we assume that the object is the collection of self-illuminated points and

the, okay, so we can easily calculate the wavefront from single point. Then we sum up all points, then we get hologram. Okay, digital holography. It's very simple,

just replace the film to the electronic device but the requirements are very high. We need ultra-high resolution as same as the wavelengths of light. It means that we need a device with less than one micron pitch. And of course we need to handle

a huge number of sampling points. Okay, let's assume our pixel is one micron and our hologram is 100 millimeter by 100 millimeter. Then it means that we need 10 gigapixel. So it's kind of challenging. We still need a lot of work. And here

I want to basically briefly talk about the holographic video display. As you know that the Professor Benton's group at MIT proposed a holographic video system in

1989. And at that time I was a visiting researcher at his group and I was in charge of the computation of the holographic range with a supercomputer. But today I'm

using, okay, this is our holographic video system. It's very simple. We just buy a video projector and remove projection lens and the lamp. Then we press the lamp

with the white LED. Then I can get a hologram image with the video projector. Actually, as you can see, here is a liquid crystal. So you can see the

holographic image here. So this is specs of liquid crystal. It has a resolution of the SXGA and 10.4 micron pitch. Then I put white light LED. And we just

use prisms and dichroic mirrors to separate RGB and also to combine the three programs. So here you can see the full color image. This is the CG model.

And this is the photograph of reconstructed image. So this is hologram. On screen it's 2D, but it's 3D. And, okay, I will show you the video. I'm not sure. Maybe

the codec is different, but actually it doesn't work. Okay. This is the displayed

image on the holographic television. Actually, he is dancing. And this hologram is calculated and displayed in real time. I'm afraid I need a strange codec. Press the file to the desktop.

Desktop? Okay. He's dancing like this. And I will show you another

animation. Okay, this is another animation. This is the sequence of the animation. So this is loads and blooming. This is also calculated in real time

and displayed in real time and 3D. But our liquid crystal is very small. And also the viewing angle is very narrow. So we cannot see the image with both sides, only the molecular observation. But at least we can get real time

calculation and display of the hologram. So here is a conclusion of the

holographic video. So full color and full parallax holographic video display can be technically realized. But with just one liquid crystal or three panels of full color, image size and viewing angle is very small.

But if we can buy, say, 100 or 1000 of the liquid crystal, then we can make image larger. But the cost is matter. Yeah. And so I think some breakthrough on spatial light modulator is required. So now I will talk about

fringe printer. So this is the photograph of fringe printer. It's very simple. It has laser, projector and some optics. It's very simple. And when calculating jigga pixels, CGH becomes easy, even with your personal

computer. You can calculate the huge pixel hologram. And for mass production, electron beam printer gives excellent result, but very expensive. So good for mass production, but not suitable for the small

business or personal use. So I think compact and inexpensive fringe printer is desirable for small business and personal use. Okay. And here is the related works. As I'll show you there, the holographic

hologram printer made by many companies already get good quality. But our method is something different. So the other works are electron beam printer. And here is some works of holographic fringe printer. So this

is the image, the constructed image of the electron beam printer, electron beam lighting. So the pixel pitch is something like a 0.1 micron. And this is very interesting result. So the professor Sakamoto at the

Hokkaido University of Japan modified the CDR lighter to light the computer generated hologram. So this is the photograph of the hologram. Then this is the constructed image of hologram. Okay. Since the

hologram recorded in the CD, so the hologram looks like not square, but something like a fan. Okay. And so Professor Matsushima at the Kansai University developed fringe printer. Actually, they print point

by point. So it is very time consuming. So he modified. Next, he used a rotating drum to light the dot pattern. And this is our system. We use a liquid crystal. And here is a laser. So we illuminate the

liquid crystal. Then here is two lenses to demagnetize. Okay. This is liquid crystal. Then we demagnetize the image of the liquid crystal. Then expose on the holographic plate. Okay. Here is how

it works. Okay. So here is a fringe and demagnetize on the plate. Then expose and move the stage. Then next to exposure, next, next. Then we can make a large hologram with small elementary holograms.

And here is a history of development. Okay. In my mind, I started this project in 1989, but actually we get good results within five years. So just last month, we have developed Mark III type.

So the, okay, here is the pixel pitch is now 0.61 micron. The Mark II is 87. The Mark I is 1.30. The main difference is

liquid crystal. Okay. It's SGA and SXGA. And it's liquid crystal for high vision, HDTV. So here is some results. This is hologram, printed hologram. And this is the constructed image. And this is

10 gigapixel. And here is some difference between Mark II and, okay, this is Mark II and Mark III. So the liquid crystal is different and the folding wavelength is different and also the pitch of the liquid crystal is different. Here is some result

of Mark III. Okay. The pixel number is here. It is roughly 22 gigapixel. And the size is here. Actually, we use a 4x5 plate,

so now the size is limited, the plate size. And this is printed hologram before bleach. And here is a reconstructed image. It's a laser reconstruction of a hologram. Very simple object.

But the big difference is the printing time. Okay. Here shows the printing time to print megapixel. Okay. So Mark II takes 1.7 seconds to print 1 megapixel. But now we can print 1 megapixel at 0.772 seconds. And here is a comparison

with our system and the CDL and single spot and drum. Okay. Our printing speed is most or fastest and the pitch is smallest and the maximum pixel is largest. But we still

need some modification with the quality, image quality. Here I want to show you the hologram. Okay. Fringe printer is different from stereogram printer. As the stereogram printer, you can only record stereogram, but our

fringe printer can record any type of hologram. Here is some sample, Fresnel hologram or H-Lung, rainbow holographic stereogram, cylindrical hologram and disc hologram. I want to show you some reconstructed image. Here again the reconstructed image of the rainbow hologram. But this

is rainbow hologram, so okay. This is view from center, left hand light, then you can see the different path selected. But this is rainbow, so if you move upside and down, then color changes. This is original data, computer graphics data. And we also record the

cylindrical Fresnel hologram, not the multiplex hologram. So this cylindrical Fresnel hologram proposed by Tom Jones in 67, and the difference, big difference is full parallax. So here is optical cylindrical hologram. This is object, and this is here is a

hologram and reconstructed image. Recording is very simple. Just put laser beam, then you can record the Fresnel hologram. And we calculate this hologram with computer, I mean that we made CGH of the

cylindrical hologram. Here is some numbers, okay. Pixel number is very big. Actually we make four holograms to make single cylinder. So it must be more than a strategic pixel. And here is object,

and okay, I forget to bring video with me, so this is just single shot of the image. It is side face of the skull, but it has both horizontal parallax

and even you can get vertical parallax. So that one is single color, but we also make full color cylindrical hologram, combined with rainbow

hologram technology. Okay. And if you use a single point as illumination, the image size is limited, so we use the collimated illumination, then image or hologram getting weaker because of the

maximum deflection angle. Here and here are the same, but the incident angle is smaller with the collimated illumination. So here is some parameters, and this is object, and this is

photograph of reconstructed image. This is, okay, the difference of the size is the same as the actual reconstruction size. So here is a point source reconstruction, and with the collimated reconstruction, you can get much, much bigger

reconstructed image. And here, I will show you some result from this program. This is concept of this program. Okay, here is some parameters. Actually, we just make H1 for this program.

So making this H2, we just put the holographic plate, then transfer optically. Here is a deconstructed image from H1 for laser reconstruction, and this is the H2 with merge color. I made two masters and transfer

with green and red lasers. I also have a movie. Actually, the image is very small,

like roughly 10 millimeters diameter.

Actually, the size, as I mentioned, the image is this big, very small, but multiple small

images. It doesn't hurt. Also, I made a

stereogram. It's very small stereogram, and actually, I presented this result at the meeting, but we can still record stereogram. And this is the merge color

dimension. So we print H1, and this is data. This is the construction from H2. So we optically transfer H1 to H2. Here is a conclusion of my talk. So our

system, fringe printer system, is fastest, finest, and the largest fringe printer, and we still need some luck to improve our image quality. And at now, we can record 22 gigapixel with 0.61 micron pitch. And the printed program

is suitable for both the master program and direct billing with FightLite. And this is just a small advertisement. I wrote a chapter with this digital holography and three- dimensional display, and the chapter title is Computer-Generated Holograms for FightLite Econ Session. So that's

all. Thank you for your attention.