Image analysis – Applications – 3-d or stereo imaging analysis
Reexamination Certificate
2000-04-28
2004-06-01
Patel, Jayanti K. (Department: 2625)
Image analysis
Applications
3-d or stereo imaging analysis
C382S285000, C345S419000, C356S012000, C348S066000
Reexamination Certificate
active
06744914
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to the mapping of objects, and more specifically to creating three-dimensional models of objects.
BACKGROUND OF THE INVENTION
The use of scanning techniques to map surfaces of objects is well known. Prior art
FIG. 1
illustrates an object
100
having visible surfaces
101
-
104
. Generally, the visible surfaces
101
-
103
form a rectangular shape residing on top of a generally planer surface
104
.
Projected onto the object
100
is an image, which includes the line
110
. In operation, the image of line
110
is received by a viewing device, such as a camera, (not shown) and processed in order to determine the shape of that portion of object
100
where the line
110
resides. By moving the line
110
across the object
100
, it is possible to map the entire object
100
. Limitations associated with using an image comprising a single line
110
is that a significant amount of time is needed to scan the object
100
to provide an accurate map, and a fixed reference point is needed at either the scanner or the object.
FIG. 2
illustrates a prior art solution to reduce the amount of time taken to scan an object. Specifically,
FIG. 2
illustrates an image including lines
121
through
125
. By providing multiple lines, it is possible to scan a greater surface area at once, thus allowing for more efficient processing of data associated with the object
100
. Limitations of using patterns such as are illustrated in
FIG. 2
include the need for a fixed reference point, and that the surface resolution capable of being mapped can be reduced because of the potential for improper processing of data due to overlapping of the discrete portions of the image.
In order to better understand the concept of overlapping, it is helpful to understand the scanning process. Prior art
FIG. 3
illustrates the shapes of
FIGS. 1 and 2
from a side view such that only surface
102
is visible. For discussion purposes, the projection device (not illustrated) projects a pattern in a direction perpendicular to the surface
101
which forms the top edge of surface
102
in FIG.
3
. The point from the center of the projection lens to the surface is referred to as the projection axis, the rotational axis of the projection lens, or the centerline of the projection lens. Likewise, an imaginary line from a center point of the viewing device (not shown) is referred to as the view axis, the rotational axis of the view device, or the centerline of the view device, extends in the direction which the viewing device is oriented.
The physical relationship of the projection axis and the view axis with respect to each other is generally known. In the specific illustration of
FIG. 3
, the projection axis and the view axis reside in a common plane. The relationship between the projection system and the view system is physically calibrated, such that the relationship between the projector, and the view device is known. Note the term “point of reference” is to describe the reference from which a third person, such as the reader, is viewing an
FIG. 4
illustrates the object
100
with the image of
FIG. 2
projected upon it where the point of reference is equal to the projection angle. When the point of reference is equal to the projection angle, no discontinuities appear in the projected image. In other words, the lines
121
-
125
appear to be straight lines upon the object
100
. However, where the point of reference is equal to the projection axis, no useful data for mapping objects is obtained, because the lines appear to be undistorted.
FIG. 5
illustrates the object
100
from a point of reference equal to the view angle fleet of FIG.
2
. In
FIG. 5
, the surfaces
104
,
103
and
101
are visible because the view axis is substantially perpendicular to the line formed by surfaces
101
and
103
, and is to the right of the plane formed by surface
102
, see
FIG. 2
, which is therefore not illustrated in FIG.
5
. Because of the angle at which the image is being viewed, or received by the viewing device, the lines
121
and
122
appear to be a single continuous straight line. Likewise, line pairs
122
and
123
, and
123
and
124
, coincide to give the impression that they are single continuous lines. Because line
125
is projected upon a single level surface elevation, surface
104
, line
125
is a continuous single line.
When the pattern of
FIG. 5
is received by a processing device to perform a mapping function, the line pairs
121
and
122
,
122
and
123
, and
123
and
124
, will be improperly interpreted as single lines. As a result, the two-tiered object illustrated in
FIG. 2
may actually be mapped as a single level surface, or otherwise inaccurately displayed because the processing steps can not distinguish between the line pairs.
FIG. 6
illustrates a prior art solution for overcoming the problem described in FIG.
5
. Specifically,
FIG. 6
illustrates the shape
100
having an image projected upon it whereby a plurality of lines having different line widths, or thickness, are used.
FIG. 7
illustrates the pattern of
FIG. 6
from the same point of reference as that of FIG.
5
.
As illustrated in
FIG. 7
, it is now possible for a processing element analyzing the received data to distinguish between the previously indistinguishable line pairs. Referring to
FIG. 7
, line
421
is still lined up with line
422
to form what appears to be a continuous line. However, because line
421
and line
425
have different thickness, it is now possible for an analysis of the image to determine the correct identity of the specific line segments. In other words, the analysis of the received image can now determine that line
422
projected on surface
104
, and line
422
projected on surface
101
are actually a common line. Utilizing this information, the analysis of the received image can determine that a step type feature occurs on the object being scanned, resulting in the incongruity between the two segments of line
422
.
While the use of varying line thickness, as illustrated in
FIG. 7
, assists identifying line segments, objects that have varying features of the type illustrated can still result in errors during the analysis of the received image.
FIG. 8
illustrates from a side point of reference a structure having a surface
710
with sharply varying features. The surface
710
is illustrated to be substantially perpendicular to the point of reference of FIG.
8
. In addition, the object
700
has side surfaces
713
and
715
, and top surfaces
711
and
712
. From the point of reference of
FIG. 8
, the actual surfaces
711
,
712
,
713
and
715
are not viewed, only their edges are represented. The surface
711
is a relatively steep sloped surface, while the surface
712
is a relatively gentle sloped surface.
Further illustrated in
FIG. 8
are three projected lines
721
through
723
having various widths. A first line
721
has a width of four. A second projected line
722
has a width of one. A third projected line
723
has a width of eight.
The line
721
, having a width of four, is projected onto a relatively flat surface
714
. Because of the angle between the projection axis and the view axis, the actual line
721
width viewed at the flat surface
714
is approximately two. If the lines
722
and
723
where also projected upon the relatively flat surface
714
their respected widths would vary by approximately the same proportion amount as that of
721
, such that the thickness can be detected during the analysis steps of mapping the surface. However, because line
722
is projected onto the angled surface
711
, the perspective from the viewing device along the viewing axis is such that the line
722
has a viewed width of two.
Line
722
appears to have a width of two because the steep angle of the surface
710
allows for a greater portion of the projected line
722
to be projected onto a greater area of the surface
711
. It is this greater area of the surface
722
that is viewed to give the percepti
Rubbert Rudger
Sporbert Peer
Weise Thomas
Carter Aaron
McDonnell & Boehnen Hulbert & Berghoff
OraMetrix Inc.
Patel Jayanti K.
LandOfFree
Method and system for generating a three-dimensional object does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and system for generating a three-dimensional object, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and system for generating a three-dimensional object will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3359610