Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
1
OPTICAL PROJECTION APPARATUS AND METHOD
FIELD OF THE INVENTION
The present invention was made with U.S. Government
support, and the Government has certain rights in the
invention.
This invention relates to the field of optical display
and, more particularly, to an optical system and method for
displaying an image. In one preferred form of the invention,
an image is projected and displayed on a solid panel display
device.
BACKGROUND OF THE INVENTION
In the field of image projection of a rectilinear object
to a proportionately enlarged or reduced rectilinear image (as
represented by conventional photographic enlargers and slide
projectors), the entire image is projected typically upon a
single plane (e. g., in the enlarger, to the photographic
paper; and from the slide projector, to the screen). A more
difficult task arises when an image must be projected into a
display device having two separate image surfaces for the
vertical and horizontal components of the image; each of which
requires independent magnification and focus of the vertical
and of the horizontal image components. The problem is
further complicated when one of the image surfaces is tilted
with respect to the projection axis; the tilt being so
significant that conventional image focus will not be
sustained along the full image surfaces. The two disparate
image surfaces must be illuminated in such a manner that the
vertical and the horizontal image components maintain
independent focus along their respective tilted surfaces.
Further, since projected images generally expand (or enlarge)
CA 02305268 2000-03-30
WO 98/I5164 PCT/IB96J01129
2
over progressively greater projected field distances, tilted
image surfaces are also subject to "keystoning", whereby one
dimension (say, the horizontal "width") is enlarged
progressively more as viewed from the "top" or the "bottom" of
the image.
An example of a device which requires such image handling
is represented in U.S. Patent No. 5,381,502 entitled, "Flat or
Curved Thin Optical Display Panel". Figure 1 illustrates the
type of panel construction described in the '502 Patent. The
panel comprises a stack of thin waveguide-like transparent
lamina 111 each of typical thickness t. When the stack is cut
at an acute angle S, each lamination exhibits a height h at
the display surface such that h = t sec S. Thus, with S
measuring typically about 70°, h is significantly larger than
t. Also, the full display height H is larger than the base
thickness T by the same factor, sec S.
The device of the '502 Patent is called a "polyplanar
optic display" (POD). The rightmost portion of the POD is
represented primarily in Figure 1 as an isometric view. The
full width W is typically wider than its display height H.
The portion which is detailed serves to describe the operation
of the POD and is useful in understanding its relationship to
the present invention. Each lamination (of thickness t) of
the panel is a transparent sheet (glass or plastic) of nominal
optical index of refraction nl, separated by thin coatings of
index of refraction nz, where nl > nz. Light entering the
laminations at the base is separated into sheets and is
confined to its respective sheets by total internal reflection
at the interfaces. Thus, light focused at the base will
retain "vertical" resolution elements of thickness t (in the
"T"-direction) throughout its propagation "upward" to the
display surface, where each thickness t is displayed as a
corresponding resolvable height h. In the width W direction,
however, there is no confinement of the input illumination,
and each sheet propagates its respective slice (in the width
direction) as would a continuous transparent medium. This
requires that the horizontal image components be focused over
CA 02305268 2000-03-30
WO 98/15164 PCTlIB96/01129
3
varying distances corresponding to the tipped viewing surface.
While the vertical component of the projected image must focus
near the base, the horizontal information must focus near the
sloping plane of the display surface; those components at the
"bottom" of the display focusing close to the base, and those
higher focusing at progressively greater distances to
represent image elements approaching the top of the display.
Also, while propagating through the lamina, the horizontal
components expand progressively as an extension to the
expanding illuminating field. Unless corrected, this
generates keystoning, whereby (in this example) the top of the
displayed image becomes wider than that at the bottom.
It is among the objects of the present invention to solve
image handling problems of the type described above and also
to provide image projection that can be used in conjunction
with a POD type of display panel.
SUMMARY OF THE INVENTION
In one form of the invention, an optical system is
disclosed for displaying an image of an object. A display
device is provided and has an input surface and an output
surface. Means are provided for illuminating the object so
that light from the object is directed toward said input
surface. Anamorphic optical means is disposed in the light
path between the object and the input surface, the anamorphic
optical means being operative to focus one directional
component of the image at the input surface and to focus a
different directional component of the image at the output
surface .
In an embodiment of the invention, the display device is
a panel device formed of a solid material and having disparate
imaging surfaces for said different directional components, at
least one of said imaging surfaces being non-perpendicular to
the optical axis of the light. In this embodiment the object
is a planar object tipped with respect to said optical axis by
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
4
an angle that satisfies the Scheimpflug condition for said at
least one of said imaging surfaces, the angle taking into
account the refractive effect of the solid material on said
light.
Also, in this embodiment a telecentric optical component
can be disposed in the path of the light to correct for
keystoning of said image.
Further features and advantages of the invention will
become more readily apparent from the following detailed
description when taken in conjunction with the accompanying
drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is an isometric view, in partially broken away
form, of a prior art POD display panel.
Figure 2 is a diagram of the projected optical field, and
the POD display, in accordance with an embodiment of the
apparatus of the invention and which can be used in practicing
an embodiment of the method of the invention.
Figure 3 shows a cross sectional view of a POD, and is
useful in understanding determination of imaging surfaces and
the determination of tilt.
Figure 4 shows an embodiment of an apparatus and
technique for practicing the invention using a scanning laser
beam to form the image.
Figure 4A illustrates a lens for providing focusing on a
sloping image plane that can be used in the Figure 4
embodiment.
DETAILED DESCRIPTION
IMAGE PROJECTION: To develop relationships between the
object and its projected image, the entire projected field is
represented in Figure 2, with the POD 205 positioned such that
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
the rays (propagating from left to right) remain "unfolded"
until they arrive at the POD base. (If a "folding" reflective
element is interposed in the ray path, the rays may be
directed "upward" such that the POD can be positioned for
typical upright viewing).
The principles hereof relate to the transfer of
information from the object surface to the image surfaces of
the POD. As such, the manner of illuminating the object is
independent of this transfer. The object will be assumed
conventional; either transmissive or reflective; illuminated
with incoherent or coherent light. One exception to this
independence is the case of illumination of the POD by a
scanned laser beam, wherein the "object" may be virtual; that
is, contained in the software which addresses the laser beam
intensity while it is scanned. This case will be discussed
subsequently.
The object can be one of a variety of light valves which
may be static for projection of a still picture (such as a
photographic slide), or dynamic, forming moving images or
changing data (such as by any of the electronically controlled
light valves). Typically, the object is planar and it
exhibits spatial information which is to be projected to a
distant image surface. At the left side of Figure 2 is
illustrated a plane object "light valve" 211, oriented at the
origin of an x-y-z coordinate system as shown, the object is
tipped such that its plane forms an angle f3 with the z-Axis.
This will be discussed subsequently. Assuming a transmissive
object, transmitted rays propagate to the right (in the z-
direction), encountering anamorphic (cylindrical) lenses Ch and
C~. Each lens provides optical power in only the horizontal or
the vertical direction. The focal lengths of Ch and C~ are
selected to satisfy the desired image distances and the
required magnifications (from the object width to the display
width W, and from the object height to the base thickness T of
the POD). In an exemplary case, the required vertical
magnification is m~ = 6.1 and the horizontal magnification is
mh = 18.5. The well-known "thin lens" relationship for the
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
6
focal length f is given by
= av -_ v
a+v m+1
in which a is the object-lens distance, v is the lens-image
distance, and m = v/u; the image/object magnification. In
application for the differing image distances and
magnifications of these systems, these and related equations
are separated into quadrature directions (with subscripts h
and v) to represent the independent horizontal and vertical
image components.
Figure 2 also snows a lens Lt close to the POD, operating
as a telecentric element to rectify keystoning. It is
provided with a long focal length, whereby its focal point is
positioned near the source of diverging rays (in the vicinity
of C~) so that it operates on the arriving diverging ray
bundles to collimate them as they propagate into the POD.
This additional optical power, positioned close to the focal
regions in the POD, also shortens slightly the original focal
lengths, as calculated per Equation 1 for Ch and C~ alone.
Considerations for this and other factors relating to the
development of the focal surfaces in the POD are now
discussed.
The horizontal projection components must accommodate the
tilting of the horizontal image plane in the POD (due to the
differing propagation lengths within the lamina). The
projected image surface of Figure 2 (along the Tilt Axis) is
determined by (subsequently described) successive calculation
of the optical paths within the POD, allowing for appropriate
depth of focus of the horizontal components. This reveals the
effective tilt of the image plane which the incoming light
must match to provide uniform horizontal focus over the entire
image surface.
This is achieved by instituting the optical arrangement
known as the Scheimpflug condition, established among the
object plane, the image plane and the principal plane of the
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
7
horizontal imaging lens Ch; accomplished by orienting these
planes such that they intersect at a single line. That is,
with the Ch plane (thin lens approximation) normal to the
projection axis and the effective image plane oriented as
determined above, the object plane is tipped so that it
intersects the intersection of the other two. This is
represented by the Scheimpflug rule,
tan~3 = m tana { 2 )
where a is the image plane tilt angle, !3 is the object plane
tilt angle {both with respect to the axis) and m is the
image/object magnification. This, too, is separated into
quadrature directions with appropriate subscripts to represent
the individual magnifications and a-tilts.
LIGHT PROPAGATION WITHIN A TYPICAL POD: Figure 3 is a
section view of a generic POD {e. g. 205), showing its outline
in bold solid lines and several (horizontal component) image
surfaces. In this Figure, the width (or horizontal) dimension
appears in-and-out of the plane of the paper. The axial
dimensions are identified in the z-direction, as is the focal
tolerance ez. The POD base thickness T corresponds to that in
Figures 1 and 2. Illumination, propagating from left to
right, traverses the keystone-correcting lens Lt (not shown)
and encounters the sloping base of the POD. This a~ tilt is
determined by application of Equation (2) for the vertical
component, after iterative determination of the tilt of the
horizontal image surface ah and the tilt of the object plane !i.
(The object plane tilt must satisfy Equation (2) for both
vertical and horizontal components; each having differing
magnifications). The locations and effective tilts of the
horizontal image surfaces are established following the
sequence of lines numbered (0) to (4), as follows:
Line (0) is the viewing surface of the POD; as
illustrated in Figures 1 and 2. This surface and the total
axial distance Zt remain the same, independent of the base tilt
a~. (Note that in Figure 1, there is shown no base tilt; i.e.,
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
8
a~ = 9 0 ~ . )
Line (1) is the corresponding focal surface in n = 1
refractive index material (air). In typical n = 1.5 material
(glass, plastic), it is extended by 1.5x to the viewing
surface, line (0). This type of consideration allows the
optical system to be calculated as though the image distances
are completely in air.
Allowing ~ez focal tolerance in air at the ends of line
(1) forms line (2), the design focal surface which will image
effectively on to the ideal line (1). This reduces
significantly the slope of the image plane and the
corresponding Scheimpflug tilt of the object plane. The image
tilt ah is established at line (2).
Line (3) is the focal surface inside the n = 1.5
material, resulting from focusing on line (2) in air. Note
that the oz near the Base (top left) remains in air, while at
the other end (bottom right) it extends to nez inside the
material.
Finally, line (4) is determined analytically as the
surface to which cylinder lens Ch must be focused such that
with the additional keystone correcting lens Lt, the image
distance is shortened slightly to line (2) in air. It is then
(per above) extended inside the higher index POD material to
line (3).
FOCUS AND KEYSTONE CORRECTION: With the horizontal image
tilt angle ah established at line (2) of Figure 3, one can now
determine the object tilt angle !3 by application of Equation
(2), given mh. This provides horizontal component focus over
the entire plane of line (2). [Line (2) is on the surface
identified in bold dashed lines on the Projected Tilt Axis of
Figure 2.] Then, by re-application of Equation (2) for the
known m~, one determines the base tilt angle a~ in Figure (3)
for vertical focus over the entire POD base. Vertical focus
is then transported via the waveguides to the viewing surface,
and joined by the above-described horizontal components as a
fully focused image.
To determine fh (the focal length of Ch) per Equation (1),
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
9
vh is taken to the vertical center of line (4) in Figure 3;
effectively before the addition of telecentric lens L
t
refocuses line (4) to line (2). Also, for calculation of m
before the addition of Lt, the image width is taken as greater
than W by the ratio of vh to the distance from Ch to Lt. When
Lt is added, it is to collimate the principal rays of the
focusing beams to width W. The focal length of Lt would
normally be taken as the distance from Ch to Lt if there were
no I3-tilt of the object. With tilt, however, minor additional
keystoning develops; accommodated by reducing the focal length
of Lt appropriately. Numerically, for an exemplary design, its
focal length is shortened by approximately 12~ to provide a
rectilinear focused image.
PROJECTION OF SCANNED LASER BEAM(S): An alternate method
of illuminating and addressing a POD-type display device, as
expressed earlier, is by scanning a laser beam (or beams) in
typical raster or line segment format, while modulating the
intensity (or intensities) of the beam(s), and projecting the
appropriately focusing array of beams into the display device
to form an image. This is a relatively conventional "laser
projection system" in which a display screen is mounted
typically perpendicular to the projection axis. However, in
this display system, the vertical and horizontal image
surfaces are not only disparate, but may be tilted with
respect to the axis. Also, there is normally no "real" plane
object (as in the above described systems) which may be placed
into the Scheimpflug condition to render the focus uniform.
This "virtual" object exists only in the video signal.
Furthermore, even if the depth of focus (later defined) is
sufficiently great to accommodate the differing image
distances, keystoning would develop, unless compensated.
To resolve this situation, it is first assumed that the
laser beam scanning process is well implemented, following
well known x, y, z, t (t=time) raster or segment scanning and
intensity modulation procedures. The scanned image can be
considered for this application as integrated over time into a
stationary image. An analog to this process is, therefore,
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
that of a photographic slide projector, as viewed from the
principal plane of the projection lens to the screen.
Everything before the lens is replaced by the scanned and
modulated laser beam. Except for the diffraction-limited
characteristics of coherent beam propagation, the flux from
the aperture of the projection lens to the screen is analogous
to the time-integrated flux during each frame of the scanned
and modulated laser(s).
To adapt to laser operation, the above projection lens is
identified as a "scan lens" of laser scanning vernacular. If
the desired image spot size b is so small as to require an f-
number F of the converging beam which is too low for its depth
of focus ez to straddle the disparate horizontal and vertical
image planes, as governed by the relations,
F = b/~, and ~z = ~ FZ~. ( 3a & 3b)
then this lens may be anamorphic. It is implemented with lens
elements similar to Ch and C~ of the earlier discussions,
whereby the horizontal and vertical image components focus
over different distances. With nominal focal distances
established in a manner represented in Figure 3, and with
horizontal and vertical laser beam scan angles into the lens
determining width W and thickness T respectively, the
specification of an appropriate projection lens is
straightforward.
An alternative which maintains a more conventional flat
field projection lens is to make the beam which illuminates
the scanners appropriately astigmatic, such that the
subsequently-scanned horizontal and vertical image components
are projected to their proper disparate image planes. This
is accomplished by placing an anamorphic lens element into the
beam before the scanners.
In either of the above two cases, if the image display
planes are perpendicular to the axis, image geometry and focus
would be satisfied completely. The vertical and horizontal
scan magnitudes can also be adjusted to satisfy differing
CA 02305268 2000-03-30
WO 98/15164 PCTIIB96/OI129
11
requirements for vertical and horizontal image magnifications.
Keystoning can be controlled in a manner discussed
earlier; by adding a telecentric lens Lt per Figure 2, to
collimate the principal ray groups and to focus them at the
center of line (2) of Figure 3. The focal length of Lt is
determined by its distance to the effective nodal source of
the projected beam. Keystoning can also be nulled by
predistorting the scanned function such that it forms a
keystoned image which is complementary to that which would
otherwise appear on the tilted display screen. This may be
done by addressing low inertia laser scanners (e. g., acousto-
optic or galvanometer deflectors) which are well known in the
art to respond to variable rate electronic drive. This leaves
only the correction of defocus (if required) appearing near
the top and the bottom of the display. Figure 4 shows laser
420, intensity modulating components 430, beam scanning
components 440, and lens components C~ and C~.
Unable to invoke the Scheimpflug condition here (since
there is no object plane), a method of providing focus on a
sloping image plane, as illustrated in Figure 4, is to replace
cylindrical lens Ch (e. g. of Figure 2) with a lens of conic
cylindrical shape, C~; one shaped so that it reduces optical
power gradually from "top" to "bottom". It is essentially a
small portion of a (solid glass or plastic) cone which is cut
therefrom such that its radius of curvature increases
gradually from top to bottom, as shown in Figure 4a. This
reduces optical power from top to bottom, gradually increasing
horizontal focal length to match the tipped horizontal image
plane.
Another alternative which allows the laser scanned system
to act more like that described earlier and illustrated in
Figure 2, is to create a synthetic object plane which may be
tilted for Scheimpflug correction. This synthetic plane can
be formed by having the laser scanner develop a real image in
space; located essentially as is the Object in Figure 2. The
imaging process of Figure 2 may be duplicated by converging
and propagating the flux through the lenses.
CA 02305268 2000-03-30
WO 98/15164 PCT/IB96/01129
12
While this disclosure identifies basic optical elements
which in combination satisfy the objectives of the invention,
it is understood that designs may be conducted by one skilled
in the art to establish characteristics which satisfy such
factors as variable (zoom) magnification, aberration
reduction, optical efficiency, and production effectiveness.
It is also understood that variations to the basic disciplines
expressed here, such as the use of folding reflective and/or
compound or cemented optical elements which may have
equivalent Fresnel, reflective, diffractive or hybrid optical
elements, remain within the scope of the principles of this
invention.