Note: Descriptions are shown in the official language in which they were submitted.
W O 92t03021 1 2 0 8 ~ ~ 7 9 PCT/GB91/01318
IMAGING SYSTEMS
This invention relates to an improved decoder for use in the ~
imaging systems described in our International Patent Application :
No. WO90/13848 and British Patent Application No. 9015843.7 to which
re~erence should first be made. ~ -
In accordance with the invention there is provided a decoder
screen for use in an imaging system in which a decoder screen is .
disposed between a viewer and a composite image containing alternate
portions of two images which together form a stereoscopic pair, the -
decoder screen comprising a surface carrying a pattern of -
alternating clear and opaque areas and a plurality of lenticular
elements which, in use, are in contact with the pattern-bearing
surface and disposed between the viewer and the said surface; the
lenticular elements being so shaped that the apparent pattern of
clear and opaque areas perceived by one of the viewer's eyes is the
inverse of the apparent pattern perceived by the viewer's other eye.
The raster-lenticular hybrid decoder screen of the invention
is capable of delivering two clear sharp images with the stereo
differences contained between them, one to each eye. As a
consequence, the raster-lenticular hybrid can produce a good 3-D
effect, as each eye receives a full colour image, one of a stereo
pair.
Preferred forms of imaging system decoder in accordance with
the invention will now be described in detail, by way of example,
with reference to the drawings, in which:
Figures l(a) and l(b) show line grid and chequerboard types -:of decoder screen patterns;
Figure 2 illustrates a prior art system employing a decoder
screen;
Figure 3 illustrates the decoder screen of the invention as :
perceived by the right and left eyes, respectively;
Figure 4 shows te images as perceived by the right and left .
eyes when using a decoder screen i~ accordance with the invention;
Figure 5 is analogous to Figure 4 but shows a screen of
chequerboard pattern;
' ~ , ' . ' " ' ' ' ' ,. . ' .
.
W O 92/03021 2 0 8 ~ ~17 9 2 PCT/GB91/013]8
Figure 6 shows schematically 'cross-over effect'; ~
Figure 7 shows a composite imacge of the type used in imaging
systems of the kind~to which the invention relates; and
Figure 8 shows a raster-lenticular hybrid decoder in
accordance with the invention.
In an imaging system of the type in which the decoder of the
invention is used, a composite image is formed for display to the
viewer. The composite image includes two sets of areas which, in
general, alternate, each set containing portions of one of two
images separated by colour overlay, line multiplexing or a
combination of the two techniques. A screen or 'decoder' is
interposed between the image and the viewer, the decoder being such
that, when the viewer looks at the image, each eye sees only one of
the two sets of areas and, hence, the portions of only one of the
two images. At the brain, the two images are combined to produce a
3-D effect, as perceived by the viewer.
Two forms of decoder pattern are shown in Figures l(a) and
l(b), the column line grid and the chequerboard. Taking for example
the line grid of Figure l(a), the decoder would be in the form of a
grid of stripes of equal width, alternate stripes being clear and
opaque respectively. This pattern is reflected in the form of the
composite image with which the decoder would be used; the composite
image for use with the line grid decoder of Figure l(a) would
comprise a similar arrangement of stripes, alternate stripes
containing portions of each of the two images which combine to give
the stereoscopic 3-D effect.
The raster-lenticular hybrid screen or decoder of the
present invention combines the action (relative to the encoded image
on the screen) of a lenticular screen with that of a raster screen,
that is to say, a column-line grid nr chequerboard as described in
our above-mentioned applications and shown in Figures l(a) and (b).
Previously, the column line grid or the chequerboard decoder
10 was positioned away from the pixel-phosphor screen 12 of the
encoded image, the degree of displacement varying with the
circumstances. This displacement between the plane of the decoder
screen 10 and the plane of the image pi~el plane 12 allowed
horizontal parallax between the left and right eye, each eye looking
through the clear sections 14 of the decoder screen to a different
W O 92/03021 ~ 7 9 PCT/GB91/01318
position of the encoded image, as illustrated in F'igure 2. ~he
different positions of the encoded images are due to the nature of
the encoding itself,~as specified in our above-mentioned
applications. Each eye sees through the clear sections 14 of the
decoder screen 10 to either the left image of a stereo pair or the
right image of a stereo pair, produced through pseudo stereo or
through two camera-stereo recording-filming, as described in
International application WO90/13848.
In the earlier system referred to, without displacement
between the plane of the decoder screen 10 and the pixel plane 12
there is no horizontal parallax. However the raster-lenticular
hybrid of the present invention employs the optical activity of its
lenticular component to displace the position of the raster
opponent. Consequently, the raster-lenticular hybrid decoder screen
and the image plane can be in direct physical contact between the
two planes with no displacement.
With the RLH (Raster Lenticular Hybrid) screen, the
horizontal parallax between the left and right eye is such that the
optical activity of the lenticular component displaces the raster
component, as shown in Figure 3. Instead of the raster decoder
screen (line grid or chequerboard) remaining in the same position,
and each eye seeing through to a different position of the image
plane behind it, as in Figure 2, it is the raster image which is
itself displaced for each eye, uncovering for each eye a different
section of the image plane directly behind it, as is illustrated by
Figure 4.
Figure 4 shows a monitor 40 with a RLH screen in accordance
with the invention, as well as a pair of stereoscopic images 44 and
46 and the resultant composite image 48 formed from these in the
image encoding system described in International application
WO90/13848 referred to previously. 40(1) and 40(r) show the RLH
screen as perceived by the left and right eyes respectively and
48(1) and 48(r) show the composit~ image porti ns seen by each of
the two eyes.
Figure 5 is analogous to Figure 4 and shows a monitor with
RLH screen, this time, of chequerboard pattern. 50(1) and 50(r)
show the RLH screen as perceived by the left and right e~es
respectively.
:
wo 92/03021 2 0 8 ~ ~ ~ 9 4 PCr/GB91/01318
The RHL decoder inverts to its raster image laO betwee ~ le
two eyes. As the rlH decoder and the image screen are in direct
contact, there can be a perfect alignment between the encoded
composite image and the RLH decoder screen. Cnce this alignment has
been established it will remain unchanged and then, from any
position in the viewing area, the optical properties of the RLH will
take effect by virtue of the horizontal displacement between the
eyes. Each eye will see an aligned left or right image view, with
the opposite eye seeing the opposite view, also aligned.
The different between previous line grid decoder screens and
the RLH decoder screen is that, with the former, the horizontal
displacement of each eye 'unlocks', through the interposition of the
line grid decoder screen, one of two images encoded in the image
plane, whereas with the RLH decoder screen the horizontal
displacement of ecah eye 'unlocks' one of two image positions of the
raster pattern and each one of the raster positions then 'unlocks'
one of the two images encoded in the image plane.
The improved alignment of the RLH screen, (which could be
fitted to monitor or projection screen of the same variety (except
the liquid crystal display) as line-grid decoder screens), means
that if at the beginning of the programme the two images combined to
form the composite encoded image are the words 'left' and 'right'
and the viewer then positions himself to align their eyes
accordingly, it will be possible, because of the high degree of
alignment, to create the crossover vision effect (see Fig.6) which
allows the brain to bring objects clearly out of the plane of the
presentation screen. The RLH screen will enable software encoded
according to the systemdescribed in the International application
referred to above to create the illusion of objects eaving the
screen plane, returning to the screen grid then going through and
beyond into its depths.
The RLH screen of the invention is designed as a standard
lenticular dual image display of the type which are popularly used
on badges and simple displays (see Figure 8). As shown, with the
RLH screen 8~, the image encoded on strips aligned behind the strips
of semi-circular lenses 82, which are mostly of plastics material,
is of the column-line and screen pattern, alternately, black and
clear, that is, black and no-ink. For each eye the pattern switches
~ J
W O 92/03021 5 PCT/GB91/01318
its apparent black and clear pattern.
The RLH screen may in certain circumstances consist of a
series of prisms as opposed to a ienticular array. The net effect,
however, must be to re-align the perceived raster position for each
eye, that is, effectively to invert the raster pattern by one phase
shift, by 180~.
There are two gauge sizes of note in this construction.
Firstly, there is the gauge of the decoder pattern which corresponds
to the dimensions of the encoded image in the pixel plane shown in
Figure 7. Seconaly, there is the gauge of the semi-circular lenses
82 of the lenticular component of the RLH screen. This latter gauge
determines the width of the strips of the decoder pattern image (see
Fig. 8).
The relationship between t~ese two gauges, i.e. their ratio
will vary with the circumstances of the application of the imaging
system for which the RLH is being specified. :
The basic principle underlying the system is believed to be
that the brain does not demand stereo displacement as measured in
the horizontal plane as generated by the eyes, as the only image
difference acceptable for the generation of stereo 3-D images. The
reason for this is that our evolution has involved several stages
where our distant ancestor or precursor was predatory, fleet of foot
(or rather limb) and far more acrobatic than any of use today are
likely to be. As a consequence in the midst of a hective chase, a
hunt for food, a chapter in survival, with boty eyes swivelling, the
need for the brain to generate an accurate 3-D representation of the
rapidly moving world, the precise position of branches, footholds,
nooks, crannies and prey itself, was vital. Under these
circumstances the eyes would be comparing simultaneous left and
right imgaes that had differences between them that represented
rotations, distortions and with eyes blinking asynchronously - time
differentiation, as well as translations in many planes.
As a consequence we have inherited neural algorithms that
are capable of creating a 3-D picture from two images with
differences between them other than simply a horizontal displacement
and attendant perspectives transformation.
The pseudo-stereo provided by the imaging system works as
well as it does because of this, for the brain is invited to resolve
. . . , -
: .
r
.. .. , , ~ - .
.. : , ': ' ~ ' ~'
2 ~ J
wo s2/03n2l
6 PCT/GB91/01318
the depth puzzle and thereby create a credible sensation of 3-D, ~
from pictures that have differences between them that are everything
but a true horizontal displacement with attendant perspectives
transformations. Pseudo-stereo invites the brain to ma~e an
intelligent guess at what the two images and their differences mean,
the outcome of this process is a 3-D image not historically
accurate, but indistinguishable in sensation from the sensations of
3-D reality.
' ' . ".:: ' :
.~
