U.S. patent application number 12/975866 was filed with the patent office on 2011-05-26 for sensing device and method of detecting a three-dimensional spatial shape of a body.
This patent application is currently assigned to TRW Automotive GmbH. Invention is credited to Rene Pfeiffer, Dirk Rutschmann.
Application Number | 20110123099 12/975866 |
Document ID | / |
Family ID | 39790882 |
Filed Date | 2011-05-26 |
United States Patent
Application |
20110123099 |
Kind Code |
A1 |
Pfeiffer; Rene ; et
al. |
May 26, 2011 |
SENSING DEVICE AND METHOD OF DETECTING A THREE-DIMENSIONAL SPATIAL
SHAPE OF A BODY
Abstract
A method for identifying a best fitting shoe includes the steps
of scanning a foot using a photogrammetric 3D foot scanner for
obtaining a digital 3D model of the foot, and providing a database
in which 3D models of shapes of the 5 interiors of available shoes
are stored. The 3D model of the digitized foot of the customer is
compared with the 3D models of available shoes stored in the
database and a shoe of which the 3D model of internal shape is the
most similar to the 3D model of the customer foot is selected. The
steps of comparing and selecting are performed using a computing
unit. A sensing device for detecting a 10 three-dimensional spatial
shape of a body includes a sensing end and a camera. A method of
detecting a three-dimensional interior spatial shape includes
providing the sensing device and scanning the spatial shape.
Inventors: |
Pfeiffer; Rene;
(Markgroningen, DE) ; Rutschmann; Dirk;
(Stuttgart, DE) |
Assignee: |
TRW Automotive GmbH
|
Family ID: |
39790882 |
Appl. No.: |
12/975866 |
Filed: |
December 22, 2010 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12668112 |
May 6, 2010 |
|
|
|
PCT/EP2008/004961 |
Jun 19, 2008 |
|
|
|
12975866 |
|
|
|
|
Current U.S.
Class: |
382/154 ; 348/46;
348/E13.074 |
Current CPC
Class: |
G01B 5/20 20130101; A43D
1/06 20130101; G01B 11/24 20130101 |
Class at
Publication: |
382/154 ; 348/46;
348/E13.074 |
International
Class: |
G06K 9/00 20060101
G06K009/00; H04N 13/02 20060101 H04N013/02 |
Foreign Application Data
Date |
Code |
Application Number |
Jul 11, 2007 |
DE |
10 2007 032 609.4 |
Claims
1. A sensing device for detecting a three-dimensional spatial shape
of a body, comprising a sensing arrangement (28) which sensing
arrangement comprises: a sensing end (32) for scanning a spatial
shape to be scanned of a body; a camera (30); a connecting device
(34) for rigidly connecting the camera (30) with the sensing end
(32), the camera (30) being arranged such that it can detect a
surface (12) which is provided with marks (22) suitable to be
automatically photogrammetrically evaluated and on which the body
to be scanned has been placed, while the sensing end (32) scans
different points of the spatial shape to be scanned of the body;
and the sensing device further comprising: a photogrammetric
evaluation program for a computing unit (18), the computing unit
being configured such that image signals generated by the camera
(30) can be routed to the computing unit (18) and the evaluation
program can calculate the 3D coordinates of the spatial shape to be
scanned from the sequence of recorded and transmitted image
sections using the marks (22) suitable to be automatically
photogrammetrically evaluated.
2. The sensing device according to claim 1, wherein the spatial
shape to be scanned of a body (26) is an inner wall of a hollow
body.
3. The sensing device according to claim 1, wherein the camera
operates in a video mode and records images continuously.
4. The sensing device according to claim 1, further comprising the
surface (12) provided with marks (22) suitable to be automatically
photogrammetrically evaluated, the exact space coordinates of the
marks (22) being known.
5. The sensing device according to claim 4, wherein an origin of
coordinates of the photogrammetrically marked surface (12) has a
trough for accommodating the sensing end (32).
6. The sensing device according to claim 1, comprising at least two
cameras (30), wherein the cameras are arranged such that a
plurality of image sections of the photogrammetrically marked
surface (12) is detected at the same time.
7. The sensing device according to claim 1, wherein the sensing end
(32) exhibits a mechanical and/or optical and/or inductive and/or
acoustic contact mechanism which generates a contact signal upon
contact with the inner wall.
8. The sensing device according to claim 7, wherein image signals
generated by the camera (30) or the at least two cameras are passed
on to the computing unit (18) or are marked before being passed on
to the computing unit only when the contact signal has been
generated during recording.
9. The sensing device according to claim 1, wherein the sensing end
(32) comprises an opto-electronic and/or acoustic distance
measuring device.
10. The sensing device according to claim 1, wherein the sensing
end (32) comprises a mechanically linearly resiliently displaceable
sensing tip having a linear displacement sensor.
11. The sensing device according to claim 1, wherein the sensing
device is a supplement to a photogrammetric foot digitizer (10)
which includes a photogrammetrically marked surface (12) and an
image sensor device (14), the image sensor device (14) being guided
around a body to be digitized, using a holder (16), and wherein the
sensing device comprises a mount which can be connected with the
holder (16) of the image sensor device (14) of the foot digitizer
(10) and in which the sensing arrangement (28) can be removably
mounted, and the sensing arrangement (28) being oriented in the
mount such that the sensing arrangement only insignificantly
conceals the image field of the image sensor device (14) on the
body to be digitized and the photogrammetrically marked surface
(12).
12. A method of detecting a three-dimensional interior spatial
shape of a hollow body, the method comprising the following steps:
fastening the body (26) to be digitized on a surface (12) which, at
known positions, is provided with marks (22) suitable to be
automatically photogrammetrically evaluated; providing a sensing
device according to claim 1; scanning the spatial shape to be
detected by means of the sensing end (32) of the sensing
arrangement (28); recording at least one section of the
photogrammetrically marked surface (12) by the camera (30) while
the sensing end (32) scans the point, a plurality of marks (22)
suitable to be photogrammetrically evaluated being detected;
repeating the steps of scanning and recording for a multitude of
different points of the spatial shape to be detected; evaluating
the recorded images by the evaluation program on a computing unit
(18), the evaluation program determining, by a photogrammetric
evaluation of the image sequences, the respective spatial position
and the respective orientation of the camera (30) from the known
positions of the photogrammetric marks (22) detected by the camera
(30) and deriving the respective spatial position of the sensing
end (32) rigidly connected with the camera (30) from the camera
position and camera orientation, and the 3D model of the body (26),
in particular the interior of a hollow body and/or geometric
measurements of the interior, being established from the spatial
positions of the sensing end (32) established in this manner.
13. The method according to claim 12, wherein the scanning of the
spatial shape is effected in a continuous movement of the sensing
end (32) and wherein the camera (30) makes recordings continuously
in a video mode.
14. The method according to claim 12, wherein the interior spatial
shape of the hollow body, in particular of a shoe, is detected, and
wherein the camera (30) remains outside of the hollow body while
the interior is scanned.
15. The method according to claim 14, further comprising the
following steps: determining a convex envelope (44) of the interior
from an overall generated point cloud of coordinates of the
interior from established positions; comparing the established
positions of the sensing end (32) with the convex envelope of the
interior; deleting points of established positions that do not lie
on the convex envelope.
16. The method according to claim 12, wherein the sensing end (32)
includes a mechanical and/or optical and/or inductive and/or
acoustic contact mechanism which upon contact generates a contact
signal, and wherein only those camera images in which the contact
signal has been generated are passed on to the computing unit (18)
or are marked before being passed on to the computing unit
(18).
17. The method according to claim 12, wherein the sensing end (32)
disposes of an opto-electronic and/or acoustic distance
measurement, and the method further comprises the following steps:
measuring a distance of the sensing end (32) from the spatial shape
in at least one defined direction; transferring the measured
distance to the computing unit (18) and calculating a correction
vector from the measured distance; adding the correction vector to
the space coordinates of the sensing end (32) that have been
photogrammetrically established from the image recordings of the
camera (30).
18. The method according to claim 12, wherein the sensing end (32)
disposes of a mechanically linearly resiliently displaceable
sensing tip having a linear displacement sensor, and the method
further comprises the following steps: transferring a displacement
signal from the linear displacement sensor to the computing unit
(18) upon contact of the sensing tip with the inner wall and
calculating a correction vector from the displacement signal;
adding the correction vector to the space coordinates of the
sensing end (32) that have been photogrammetrically established
from the image recordings of the camera (30).
19. The method according to claim 12, wherein the method further
comprises the following calibration steps: placing the sensing end
(32) on a defined point, preferably an origin of coordinates of the
photogrammetrically marked surface (12); moving the camera (30) on
a spherical surface around this point, the sensing end (32)
remaining in contact with the photogrammetrically marked surface
(12) during the movement; recording a sequence of images during the
movement and passing it on to the computing unit (18); calculating
the relative position of the sensing end (32) in relation to the
origin of coordinates of the image sensor of the camera (30) as
well as the internal parameters of the camera (30) from the
sequence of images recorded by the camera.
20. The method according to claim 12, wherein the scanned spatial
shape is the interior of a body (26) that has been deformed by use
and/or stress, and wherein the calculated 3D model and/or the
calculated geometric measurements of this interior deformed by use
is/are made use of for selecting and/or manufacturing a new
body.
21. The method according to claim 12, wherein the spatial shape to
be scanned pertains to any one of the following bodies: a footwear;
a headgear and/or a headguard; a covering for clothing and/or
protecting the human or animal body; an orthopedic and/or
prosthetic product having an interior that is open to the outside
and presents a spatial surface area to be adjusted to an anatomy of
a patient requiring orthopedic and/or prosthetic care.
22. The method according to claim 12, wherein the sensing
arrangement is mechanically moved by a motor drive such that the
sensing end (32) contacts as many points as possible of the
interior of the body.
23. A method for identifying a best fitting shoe, the method
comprising the steps of: scanning a foot of a customer using a
photogrammetric 3D foot scanner for obtaining a digital 3D model of
the foot; providing a database in which 3D models of shapes of the
interiors of available shoes are stored; comparing the 3D model of
the digitized foot of the customer with the 3D models of available
shoes stored in the database; selecting a shoe of which the 3D
model of internal shape is the most similar to the 3D model of the
customer foot: wherein the comparing and selecting steps are
performed using a Computing unit.
24. The method of claim 23, wherein the database contains digital
3D models of internal shapes of shoes produced by shoe
manufacturers organized in a business alliance.
25. The method of claim 23, wherein the 3D models of shapes of the
interiors of available shoes are obtained by scanning
nondestructively the spatial inner shape by means of a sensing
arrangement, the sensing arrangement comprising a sensing end which
is moved inside the shoe and a camera which is connected rigidly to
the sensing end in such a way that the camera remains outside the
shoe during scanning.
Description
RELATED APPLICATION
[0001] This application is a divisional of co-pending U.S. patent
application Ser. No. 12,668,112, filed May 6, 2010.
FIELD OF THE INVENTION
[0002] The present invention relates to a sensing device and a
method of detecting a three-dimensional spatial shape of a body,
including a sensing end for scanning a surface to be scanned of the
body, and a camera, a multitude of points on the surface to be
scanned being scanned using the sensing end.
[0003] The manufacture of well fitting shoes within the scope of
the so-called mass customization usually requires the 3D
digitization of the spatial shape of the feet of the customer. This
data is used for selecting from a collection of shoe lasts that are
available physically or virtually as a 3D data set, that shoe last
with which the best fitting shoe can be produced (mass
customization of a best fitting shoe). In the more demanding,
individual mass customization business model, an individual shoe
last is fabricated from the 3D data set of the digitized foot and
an individual custom-made shoe is produced hereby.
[0004] It is, to date, still very difficult to establish the 3D
shape of a suitable shoe last automatically from the 3D model of a
foot since a large number of other factors such as shoe style and
fashionable shape, properties of materials, anatomical and
biomechanical secondary conditions of the customer etc. need to be
considered. Shoe lasts have a 3D shape that distinctly differs from
that of the appertaining non-compressed foot. These difficulties
result in that despite the increasing availability of suitable foot
scanners, the business model of shoe mass customization grows only
hesitantly. Since in particular low-cost and calibration-free foot
scanners such as, for example, the "Lightbeam.RTM." foot scanner of
corpus.e AG, of Stuttgart, Germany (see www.corpus-e.com) have also
become available in the meantime, the economic situation of mass
customization of footwear is unsatisfactory because of these
above-mentioned difficulties in the adaptation of the 3D foot shape
to the 3D shoe last shape.
[0005] An alternative approach is to compare the 3D model of a
digitized foot with the internal shape of a real shoe. This may
involve, for example, the internal shapes of the shoes of a product
range offered for sale, or else, the internal shapes, stored in a
database, of the products of a large number of shoe manufacturers
organized in a business alliance. By comparing the spatial shape of
the digitized foot of a customer with a selection of internal
shapes of footwear coming into consideration, the best fitting shoe
can be found much more directly since the internal shape of a shoe
and the 3D foot model are more similar to each other than a shoe
last and a 3D foot model.
[0006] A shoe that has been worn has stored the process of
adaptation to the anatomy of the foot of the wearer of the shoe.
Therefore, in the case of orthopedic shoes, for example, it makes
sense to make use of the internal shape of a shoe that has been
worn for selecting a new shoe by comparing the internal shape of
the worn shoe with the internal shapes of the shoes offered for
sale.
[0007] However, the problem of nondestructive 3D digitization of
the internal shape of footwear has not yet been solved
satisfactorily. It is known to fill shoes with a compound and to
digitize this compound. But to this end, the shoe, as a rule, has
to be cut open and, hence, be destroyed.
[0008] The French institute CTC from Lyon (www.ctcgroupe.com)
describes on pages 6 and 7 of its brochure CTC Enterprise
International--fall 2006 a complex arrangement for digitization of
the inner shoe with the aid of two endoscopic camera systems
mounted in a stereo arrangement.
[0009] In the PCT publication WO 03/087715 A1 (R. Massen), a method
is described in which a sock provided with photogrammetric marks is
tucked into the inner shoe such that it lies against the inner wall
and the marks face inwards. The interior can then be digitized
photogrammetrically using one or more endoscopic camera systems.
But this method is also complex and time-consuming.
BACKGROUND OF THE INVENTION
[0010] In industrial photogrammetry, systems are used in which a
probe provided with bright marks arranged crosswise is manually
placed point by point on a body to be measured, such as, e.g., a
car body. The probe is observed by usually a plurality of
high-resolution photogrammetric cameras from a relatively great
distance. The internal and external parameters of these cameras are
established prior to the measurement proper by means of an involved
calibration. From the images of these cameras, which show the
marked probe in space, the 3D point coordinates at the places,
touched by the probe, of the digitized body can then be measured
(see, e.g., the V-Stars system of the Australian company Geodetic
Systems Inc. www.geodetic.com). When the cameras are operated in a
video mode and record images of the probe continuously, this is
also referred to as "videogrammetry". These systems are very
involved in respect of calibration, space requirements and the very
expensive photogrammetric cameras, and are therefore not suitable
for use in a shoe store.
[0011] Similarly complex photogrammetric systems are known from the
medical field, which measure the position of a medical device such
as, e.g., a scalpel, in space. Here, in a manner similar to the
V-Stars system of Geodetic Systems Inc., bright or well-reflecting
marks are fastened to the scalpel handle and these marks are
monitored continuously by a plurality of cameras mounted in the
operating room. From the photogrammetric evaluation of the image
sequences of all cameras, the respective spatial position of the
scalpel marks and, derived from this, also that of the scalpel tip
situated invisibly in the body, can be ascertained hereby. Such
systems are available, for example, from the German company
BrainLAB AG, of Munich, Germany (www.brainlab.com).
[0012] All of these expensive industrial photogrammetric
[camera/marked probe] systems have in common that the very
precisely marked measuring probe is detected by, as a rule, a
plurality of exactly specified, oriented and calibrated cameras
from a relatively great distance. They can only be operated by
trained specialists or technicians, in particular also because of
the calibration procedures to be repeated many times before the
actual digitization starts.
[0013] There is therefore a great economic and technological
interest in having a low-cost, space-saving, largely
calibration-free method that is simple to operate, for a
nondestructive detection of the 3D shape of the interior of new or
already worn footwear, with the aid of which 3D models of the
interior of footwear can be established and the best fitting shoe
can be determined by comparing the 3D model of the foot of a
customer with the 3D models of the interiors of shoes offered for
sale. In particular, it would be of great advantage if this
nondestructive detection of the 3D shapes of the interiors of shoes
could be performed using the already available components of the 3D
foot scanner which is required for mass customization at any
rate.
SUMMARY OF THE INVENTION
[0014] This is achieved in accordance with the invention by a
sensing device according to claim 1. A sensing end for scanning a
surface to be scanned of a body is rigidly connected with a camera
by means of a connecting device. As the sensing end moves, the
camera is thus always moved along. The camera is arranged such that
it can detect a surface which is provided with marks suitable to be
automatically photogrammetrically evaluated and on which the body
to be scanned has been placed, while the sensing end scans
different points of the surface to be scanned of the body. Since,
therefore, the viewing direction of the camera is substantially
towards the sensing end, the surface provided with marks suitable
to be automatically photogrammetrically evaluated can be part of a
3D foot scanner, and the body to be scanned is put at the place
that is usually intended for the feet. In one embodiment; the
surface may, however, also be part of the sensing device. Marks
suitable to be automatically photogrammetrically evaluated are
marks which can be automatically individually identified, e.g. by a
computing unit with a photogrammetric evaluation program, and allow
an exact determination of the position thereof. To this end, the
marks are encoded in a suitable manner, as is known from the prior
art. When the surface to be scanned is an inner wall of a hollow
body, that is, for example, the internal shape of a shoe, the
connecting device is configured such that the camera remains
outside of the hollow body, i.e. the shoe, while the sensing end
scans points in the interior of the shoe. The sensing device
according to the invention further includes a photogrammetric
evaluation program for a computing unit, the computing unit being
configured such that image signals generated by the camera can be
routed to the computing unit and the evaluation program can
calculate the 3D coordinates of the spatial shape to be scanned
from the sequence of recorded and transmitted image sections using
the marks suitable to be automatically photogrammetrically
evaluated. Owing to the rigid connection between the sensing end
and the camera, the position and orientation of the camera vary
with each movement of the sensing end, and a different image
section is recorded each time, each of which have a plurality of
marks captured thereon that are suitable to be automatically
photogrammetrically evaluated and have 2D coordinates that are
specific to this image section. The camera position and, thus, also
the position of the sensing end can be determined from the sequence
of recorded and transmitted image sections.
[0015] The sensing device allows an, e.g., tactile scanning of,
more particularly, interiors of bodies and, in doing so, allows the
respective spatial positions of the sensing end to be established
by way of the photogrammetric image recordings of a camera that is
mechanically rigidly connected with the sensing end, purely from a
photogrammetrically marked base plate, i.e. the surface provided
with marks suitable to be automatically photogrammetrically
evaluated and on which the body is fixed in place. Any measuring
mechanisms requiring time-consuming calibration are thus made
dispensable. The only measuring aid, which contains absolutely
precise and known distances, is the photogrammetrically marked base
plate; it can be produced in a simple manner in terms of printing
technology and with high absolute precision. This makes the sensing
device according to the invention and the method according to the
invention self-calibrating, allowing them to be employed using
simple cameras and imaging optics.
[0016] In one embodiment, the sensing end includes an
opto-electronic and/or acoustic distance measuring device. In this
embodiment, the sensing end need not be brought into physical
contact with, e.g., the inner wall of a cavity to be digitized. The
sensing end is thus in the form of an optical non-contact probe
which, with a sufficient proximity to the inner wall as prescribed
by the sensing arrangement, triggers an image recording or, upon
reaching a release distance, marks images from a sequence of
continuous images of the camera for photogrammetric evaluation.
[0017] Further preferred embodiments of the sensing device are
apparent from the dependent claims.
[0018] The present invention further provides a method according to
claim 12. In a preferred embodiment, the interior of a hollow body,
in particular of a shoe, is detected, the camera remaining outside
of the hollow body while the interior is scanned. The convex
envelope of the interior is determined from an overall generated
point cloud of coordinates of the interior from established
positions. A comparison of the established positions of the sensing
end with the convex envelope of the interior allows those points of
established positions to be recognized which do not lie on the
convex envelope. These points are deleted. Any errors in
measurement that arise, e.g., in that the person guiding the
sensing arrangement moves away from the inner wall, are thus
corrected automatically by removing the incorrect measuring points.
This allows the sensing arrangement to be configured particularly
simply since it need not have a release contact which releases an
image recording or marks an image recorded by the camera only when
the sensing end is in tactile contact with the inner wall of the
footwear. Rather, the moving camera records images continuously,
for example, and does not require synchronization to the sensing
end or to the movement.
[0019] In addition to the digitization of the internal shapes of
footwear, the arrangement according to the invention and the method
according to the invention are equally suited to detect the
internal shapes of prosthetic funnels and similar orthopedic or
prosthetic hollow bodies. Moreover, interiors of three-dimensional
technical products may also be digitized. It is, of course, also
possible to scan outer walls.
[0020] In addition to embodiments which provide for the sensing
arrangement to be guided manually, the invention also comprises
motor guides for moving the sensing end within the interior in a
predefined or random movement such that numerous points on the
inner wall are touched, while the camera makes recordings of the
photogrammetrically marked surface at the same time. Owing to the
property of self-calibration being maintained, the motor drive may
be of a very simple design.
[0021] In a preferred embodiment, the method further comprises
calibration steps in which the sensing arrangement, i.e. the
relative position of the sensing end in relation to the origin of
coordinates of the image sensor of the first camera as well as the
internal parameters of the first camera are calculated. In the
process, the sensing arrangement is calibrated with the aid of the
same measuring installation including the photogrammetrically
marked surface and the photogrammetric evaluation of the images
supplied by the camera, i.e. the spatial position of the sensing
end is established with respect to the coordinate system of the
image sensor of the camera. For this purpose, the sensing end is
fixed in place at a known position on the surface provided with
marks suitable to be automatically photogrammetrically evaluated
and the camera is moved on a spherical surface in space while image
recordings of the photogrammetrically marked surface are made. The
positions of the camera can then be calculated by an evaluation of
the image recordings with the aid of the photogrammetric marks, and
the exact spatial distance between the origin of coordinates of the
camera sensor and the sensing end can be determined on the basis of
the spherical shape of the movement.
[0022] Preferably, the positioning of the sensing end during this
spherical movement is facilitated by a trough-shaped seat for
receiving the sensing end in the base plate.
BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Further embodiments and advantages of the invention will
become apparent from the description below of a preferred
embodiment. The invention will be described using the example of
the combination, economically of particular interest, of the
photogrammetric Lightbeam.RTM. foot scanner from the prior art with
the method according to the invention and the device according to
the invention for 3D detection of the interior of footwear, the
following illustrations being used:
[0024] FIGS. 1A and 1B schematically show a Lightbeam.RTM. foot
scanner from the prior art;
[0025] FIG. 2 schematically shows the sensing device according to
the invention; and
[0026] FIG. 3 shows a side view of a point cloud of a 3D spatial
shape as established of an inner shoe and a section taken through
the point cloud and thus the associated convex envelope of the
inner spatial shape.
DESCRIPTION OF PREFERRED EMBODIMENTS
[0027] FIG. 1A shows a side view of a photogrammetric
Lightbeam.RTM. foot scanner 10 from the prior art for obtaining the
spatial shapes of feet, which can be supplemented in an
advantageous manner by a sensing device according to the invention.
The foot scanner or foot digitizer 10 includes a base plate 12, an
image sensor device 14 fastened to a holder 16, and a computing
unit 18. A customer pulls socks 20 over his/her feet, the socks
bearing marks suitable to be photogrammetrically evaluated (see DE
10113211, "Markierungssystem fur die automatische
photogrammetrische Digitalisierung von Korpern and Korperteilen").
The customer stands at a predefined position on the base plate 12,
which is illustrated in a top view in FIG. 1B. The base plate 12 is
approximately circular and its surface is a surface provided with
marks 22 suitable to be automatically photogrammetrically
evaluated.
[0028] The motor-driven image sensor device 14 travels around the
customer on a roughly circular path and detects the
photogrammetrically marked foot/leg area 20 and a section of the
likewise photogrammetrically marked base plate 12 at the same time.
Using the image sensor device 14, both the foot or feet and a
respective section of the marked base plate 12 are recorded out of
a multitude of spatial positions here. The image data generated by
the image sensor device 14 is in this way transferred via an
interface 24 to a computing unit 18 for photogrammetric evaluation.
The transfer is effected in any desired manner as known from the
prior art, such as wire-bound, wireless, by means of a data
carrier, etc. In addition to the measurement of the 3D shape of the
foot or feet, this arrangement permits a precise calculation of the
spatial position (i.e. the external parameters) of the image sensor
device and of the internal parameters thereof at the same time, so
that this system has the great advantage of an inherent
self-calibration.
[0029] In accordance with the invention, a similar basic
arrangement is employed for digitization of the internal shape of
footwear. FIG. 2 shows, in a side view, a shoe 26 having an
interior that is to be digitized. The shoe 26 is positioned and
fixed in place on the photogrammetrically marked base plate 12
having marks 22. In an economically favorable way, the base plate
is that of the foot scanner 10 where provision is already made for
fixing devices. But of course any other surface may also be used
that is provided with marks suitable to be automatically
photogrammetrically evaluated.
[0030] FIG. 2 further shows a sensing arrangement 28 having a
camera 30, a sensing end 32, and a connecting device 34 for rigidly
connecting the camera 30 with the sensing end 32. Advantageously,
the connecting device 34 is in the form of a bent rod, so that the
sensing end 32 can be guided to all or approximately all of the
points of the inner wall of the shoe while the camera 30 remains
outside of the shoe at all times. The form of the connecting device
34 may be adjusted to the respective measuring jobs. Rather than
one camera, it is possible to provide two or more cameras which
then view the surface 12 at different angles. A hand 36 guides the
sensing arrangement 28 manually. Provision may, however, also be
made for a motor-driven guidance of the sensing arrangement 28.
Dashed arrows 38 indicate the image field of the camera 30. The
camera 30 or cameras is/are arranged such that it/they essentially
point(s) towards the sensing end. As a result, the camera 30 is
arranged to detect, within the image field, sections of the surface
12 provided with the marks 22 that are suitable to be automatically
photogrammetrically evaluated when the shoe 26 that has been placed
on the photogrammetric surface 12 is scanned. The images recorded
by the camera 30 are transferred as image signals to the computing
unit 18. When the sensing arrangement 28 is used together with the
foot scanner 10, it is possible to use the computing unit of the
foot scanner 10, which has been supplemented by capabilities of
evaluation corresponding to those of the sensing arrangement 28.
But it is, of course, also possible to use a different, separate,
computing unit with an appropriate evaluation program. The image
signals from the camera 30 are transferred to the computing unit 18
in any desired form known from the prior art, wire-bound, wireless,
by way of a data carrier, by radio or by other means.
[0031] For the scanning process, the sensing arrangement 28 is
guided manually or motor-powered such that the sensing end 32 scans
the inner walls of the shoe 26 at numerous points and, from its
spatial positions, the camera 30 continuously records image
sections of the photogrammetrically marked base plate 12 and
transfers them via an interface to the computing unit 18 for
photogrammetric evaluation. The camera 30 preferably operates in a
video mode here. The inner walls need not be scanned on a
predefined path; rather, the scanning may be performed freely on
any desired non-systematic path.
[0032] In a first embodiment, the sensing end 32 mainly contacts
the inner wall of the footwear 26 to scan it mechanically; here,
the sensing end 32 may be fitted with a mechanical and/or optical
and/or inductive and/or acoustical contact mechanism which upon
contact with the inner wall generates a contact signal. Either an
image is not recorded until triggered by the contact signal, or,
preferably, the camera 30 records images continuously in a video
mode and, based on the contact signal, those images are selected
which were recorded at the moments the contact signals were
supplied.
[0033] In a second embodiment, the sensing end 32 disposes of a
mechanically linearly resiliently displaceable sensing tip having a
linear displacement sensor. Upon contact of the sensing tip with
the inner wall of the shoe 26, a displacement signal is transferred
from the linear displacement sensor to the computing unit 18. A
correction vector is calculated from the displacement signal and is
added in the computing unit 18 to the space coordinates,
photogrammetrically established, of the sensing end 32. This allows
the coordinates of the inner wall to be established without the
sensing tip mechanically loading the inner wall with any
appreciable forces. The second embodiment may be combined with the
first embodiment.
[0034] In a third embodiment, the sensing end 32 disposes of an
opto-electronic and/or acoustic distance measuring device for
measuring a distance of the sensing end 32 from the spatial shape
in at least one defined direction. The measured distance is
transferred to the computing unit 18 and a correction vector is
calculated from the measured distance. This correction vector is
added to the space coordinates of the sensing end 32
photogrammetrically established from the images recorded by the
camera 30. In this way, the coordinates of the inner wall are
established without the sensing end 32 mechanically contacting the
inner wall.
[0035] In all of the embodiments, the camera 30 preferably
permanently records images of the photogrammetrically marked base
plate 12 during the scanning of the inner wall and transfers these
images to the computing unit 18, which uses these image sequences
to calculate, applying the methods of photogrammetry known to those
skilled in the art, the respective spatial position of the camera
30 and, derived therefrom, the spatial position of the sensing end
32 rigidly connected with the camera 30. This produces a point
cloud of spatial points, the vast majority of which consists of
space coordinates of the inner wall of the footwear, mixed with few
spatial points that reproduce sensing positions in the interior of
the cavity, especially when no contact signal is used.
[0036] As already set forth above, the sensing device is
self-calibrating since the images recorded always include marks 22
suitable to be photogrammetrically evaluated which are located at
known positions. However, the sensing arrangement 28 itself
requires calibration, i.e. the relative position of the sensing end
32 in relation to the origin of coordinates of the image sensor of
the camera 30 and the internal parameters of the camera 30 need to
be calculated. The distance between the camera 30 and the sensing
end 32 and the orientation of the camera 30 and the sensing end 32
in relation to each other are calculated. This is important in
particular when the shape of the connecting device 34 is adjusted
to the respective measuring job. In doing so, the sensing
arrangement 28 is calibrated with the aid of the same measuring
installation comprising the photogrammetrically marked surface and
the photogrammetric evaluation of the images supplied by the
camera, i.e. the spatial position of the sensing end 32 is
established with respect to the coordinate system of the image
sensor of the camera 30. To this end, the sensing end 32 is fixed
in place at a known position on the surface 12 provided with marks
22 suitable to be automatically photogrammetrically evaluated; in
the preferred embodiment a small trough is provided in the base
plate 12 for this purpose. The camera 30 is then moved manually or
motor-powered on a spherical surface in space, the camera 30 making
image recordings of the photogrammetrically marked surface 12. The
positions of the camera 30 can then be calculated by an evaluation
of the image recordings with the aid of the photogrammetric marks
22, and the exact spatial distance between the origin of
coordinates of the camera sensor and the sensing end 32 can be
determined on the basis of the spherical shape of the movement.
[0037] FIG. 3 shows a side view of a point cloud 40 as is
calculated by way of example in the computing unit 18 for the 3D
spatial shape of the inner wall of the shoe 26. Along a cutting
line 42, a convex envelope 44 of the interior is produced as is
established from the point cloud by the computing unit 18. It can
be seen that points 46 of the point cloud 40 do not lie on the
envelope 44. The evaluation program is configured such that it
recognizes points that do not lie on the convex envelope of the
internal spatial shape and, out of the generated point cloud, thus
automatically identifies such points as not being part of the
internal shape of the footwear and deletes them from the 3D
model.
[0038] The points 46 are such spatial points that were established
based on sensing arrangement positions in which the sensing end 32
had not momentarily contacted the inner wall. They are easy to
identify and to delete since these spatial points lie in the
interior of the 3D point cloud of the inner shoe.
[0039] Accordingly, the method according to the invention, which
uses the sensing device according to the invention, allows a large
number of spatial points of the inner wall of a footwear to be
obtained in a very short time in a very simple manner and also
allows an available photogrammetric foot scanner from the prior art
to be advantageously used at the same time.
[0040] While the invention is not limited to this particularly
economical combination of a photogrammetric 3D foot scanner and a
photogrammetric-tactile inner shoe scanner, i.e. the sensing device
according to the invention, the method and the sensing device
described here by way of example show how small the additional
expenditure is for the digitization of an inner shoe in comparison
with a mere foot digitization. This especially facilitates to
achieve the above-mentioned object to use the 3D model of a foot of
a customer together with a data base of stored 3D models of the
inner shoes of shoe models coming into consideration for selecting
suitable, well-fitting shoes.
[0041] The invention can also assist in determining a fitting
footwear in that the interior of a well-fitting shoe of a customer,
which has already been worn in by wearing it, is digitized and that
this 3D model of the worn-in inner shoe is compared with the
interiors of the shoe models available in a data base, or an
individual well-fitting footwear is manufactured on the basis of
this deformed 3D model of the inner shoe. Especially in the case of
orthopedic footwear, an inner shoe that has been deformed by
wearing the shoe constitutes a more valuable 3D model than the
directly digitized foot of the patient, since it reflects the
history of the deformations of the footwear.
[0042] In a preferred embodiment, the sensing device is a
supplement to a 3D foot scanner 10 and comprises a mount. The mount
can be connected with the holder 16 of the foot scanner 10 such
that the sensing arrangement 28 can be mounted in the mount in such
a way that for the digitization of a foot using the photogrammetric
foot scanner 10, the image sensor device 14 can detect the foot or
feet to be digitized together with sections of the
photogrammetrically marked surface 12, without the sensing
arrangement 28 protruding in a disturbing manner into the
measurement space to be digitized.
[0043] This may be achieved, for example, in that when the sensing
arrangement 28 is hung on the holder 16 of the foot scanner 10, the
image sensor device 14 detects the foot to be digitized via a
deflecting mirror, undisturbed by the sensing arrangement.
[0044] The best fitting shoe is then established by comparing the
3D model of a customer's foot with the 3D models of the interiors
of shoes that are offered. As an alternative, the 3D information of
the shoe lasts associated with the shoes offered is made use of in
addition to the 3D information of the interior of the shoes
offered, for selecting the best fitting shoe (best fit
customization).
* * * * *
References