The use of computer graphics rendering software in the analysis of a novel autostereoscopic display design

Computer graphics ‘ray tracing’ software has been used in the design and evaluation of a new autostereoscopic 3D display. This software complements the conventional optical design software and provides a cost-effective method of simulating what is actually seen by a viewer of the display. It may prove a useful tool in similar design problems. 1. Computer graphics ‘ray tracing’ In computer graphics the term ray tracing is used to describe a particular method of image rendering. A description of a three dimensional model is entered into the computer. A centre of projection and a window on an arbitrary view plane are selected. A ray is fired from the centre of projection through each pixel in the window. The colour of the first object intercepted by the ray is assigned to that pixel. Secondary rays may be fired from the intersection point to simulate shadowing, reflection, and refraction; this allows the simulation of transparent object, surface reflections and materials with different optical indices. Multiple rays may be fired through a pixel and the resulting colours averaged to ameliorate aliasing artefacts in the resulting image (Figure 1). Ray tracing, in the computer graphic sense, consists, essentially, of tracking photons backward from the eye to the light source(s). As such, it provides a tractable solution to the problem of determining the illumination of all surfaces in an image. It has limitations. It cannot, for example, handle diffuse interreflections between surfaces. It is, nevertheless, widely used in the special effects industry and it has proven useful in the design of a new autostereoscopic display device. 2. The autostereoscopic display The display is a device which allows a viewer to see a different image with each eye, providing stereoscopic perception. Full details of the display may be found in Moore et al and Dodgson et al. It is an unusual display in that the viewer observes a CRT through an optical system, rather than directly (as in a conventional TV) or projected on a screen (as in a video projector). The display consists of two superimposed optical systems (Figure 2). One can be thought of as a compound lens casting an image of the CRT to a plane in space. The other consists of an active shuttering element as close as possible to the front principal plane of the compound lens, and a further lens at the position of the CRT image. The second system can be thought to cast an image of the shutter into the space in front of the display. This image of the shutter is called the eye box. Consult references 4, 7 and 8 for more details. In practice, no image of either shutter or CRT is actually cast onto any surface. The viewer observes the CRT face plate through the entire optical system, but has the illusion that the viewed image is displayed on the front lens element. The combination of a fast CRT with the active shutter enables the display to provide a different image to each eye, giving stereoscopic perception. The original design for the display used a Fresnel lens as the front element. During development of the 50" version of the display it became desirable to replace this lens with a mirror (Figure 3). One of the authors (JRM) experimented with a large (1m radius) spherical Figure 1: an example ray traced image. One sphere is transparent and refractive. The rear plane is semi-reflective. Figure 2: the basic display. CRT on left, compound lens in centre, active shutter to its right, front lens (Fresnel) at right.