A resource hungry technology could be trimmed , says S.Ananthanarayanan.

Instantly representing distant things in true-life, in 3 dimensions, has been an elusive dream. The technology itself has been there, but the resources needed for implementation are prohibitive.

Takashi Nishitsuji, Takashi Kakue, David Blinder, Tomoyoshi Shimobaba and Tomoyoshi Ito, from Tokyo Metropolitan and Chiba Universities, Japan, and Vrije Universiteit in Brussels, write in the journal, Scientific Reports, of developments in the building and transmission of holographic images, which could enable shapes and images to be created and conveyed over distances in a practical way.

Images that are captured in drawing, or in photography, are essentially two dimensional. What they record are the shape and colours of an object, as projected on a plane surface – a canvas, a camera screen, or the retina of the eye. There is an impression of depth, as distant parts of a familiar object are shown smaller than nearer parts. Meaningful appreciation of depth, however, needs two such images, from points that are separated by a distance, like the images on the retinas of a pair of eyes. Images recorded from different points and projected, one to each eye could thus create a life-like image. We may recollect the 3D images in the Viewmaster device. But the images are static, and one cannot move one’s head and ‘see around’ the objects being shown.

The hologram, however, makes this possible. The hologram is not the recording of the image of an object, as received at a sensor or set of sensors, but is created by capturing the wave-front, or the pattern of light waves that emerge from the object. If the same wave-front is created again, it appears to sensors, like a pair of eyes, that what they see is the real object. In normal imaging, however, it is not the wave-front that is captured, it is an image of the object focused on a screen. The hologram, in contrast, does not focus the image, it arises from two sets of light waves, those that come directly from the source laser and the waves reflected from the objects, and captures the interaction.

Waves from the two sources fall on every point on the hologram plate. And at each point, the waves either add, to get stronger, or cancel, to get weaker. The screen is hence covered with a pattern of dark and bright portions, like a bar-code, and this distribution captures the relationship of the illuminating light and the light that has been reflected by the objects illuminated.

Now, if this screen, with the pattern of dark and bright parts, is again illuminated by a beam from the same laser, the wave-front that emerges would have the same intensity distribution as the wave-front which originally played upon the screen. A person who looks at the screen would thus see the same, original wave-front, and perceive the same objects as before. As the pattern on the screen, and hence the wave-front, does not depend on where the viewer is located, views from different parts of the screen would show different images, as if the original objects were physically there

Nor is it necessary that holograms be made from actual objects and laser light. The interference can be simulated and the pattern on the hologram can be generated by software. Computer Generated Holograms (CGH)- also called electro-holography, has become “a very promising technology,” the Scientific Reports paper says. As the pattern recorded, or generated, arises from interference of light waves, we can imagine that the pattern on the hologram would be fine, with dimensions comparable to the wavelength of light. The quantity of data in the CGH is hence very large and the computers to generate a hologram need extensive resources. And with all this, what is created is a static pattern. To show motion, we need to create a hologram every sixteenth of a second, so that the eyes see continuous motion. Even if this were made feasible, there is a problem when the data needs to be transmitted. This would be needed in practical applications of the hologram, like transmitting a surgical operation, for review in real time. Or a display of navigation data as 3D pictures, to a motorist.

There have hence been several approaches to compress the computation load and data in CGH, says the team writing in Scientific Reports. One method is to consider the light from the object to come from a series of point light sources (PSL), which reduces the extent of data to consider. This is combined with data handling approaches, like maintaining a library of precalculated results, from which the program could pick up the required values, in place of carrying out computation. Other methods are to approximate and simplify the computation or to recognize aspects of the image that would be the same for both eyes, or to embed some of the computation processes into the hardware. However, despite all this, the paper says, sophisticated computing facilities are required to achieve practical computing speeds.

In the work so far, the distributed PSLs that simulate the object generate two-dimensional wave fronts, or surfaces, which the computational utilities superimpose, to arrive at the CGH. In the current work, what is considered are the PSLs which are along a line, at the same depth. The result of this method, which the authors call the Computer Graphics – line method, is a one-dimensional wave front. Superimposition of one-dimensional wave fronts is computationally far simpler than doing it with 2-D wave fronts.

The 3-Dimensional image to be projected is hence built up of multiple 2-dimensional images, which consist of outlines. Although not with details, the basic depth relationships are maintained. The dimension of depth can hence be added to figures like symbols and letters, which would make it suitable for “car navigation systems and remote work support systems,” the paper says.

The method has limitations, one being that ‘full’ 3D is not possible, as continuous lines in the direction of depth are not possible. But it is a method that permits interactive 3-D projection, and has the great merit that it can run on an ordinary, consumer computer.

------------------------------------------------------------------------------------------ Do respond to : response@simplescience.in -------------------------------------------