Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
1
system and its apparatuses for irtlage reproduction and recording
with the methods for posixioning, processing and controlling
CROSS-REFERENCE T4 RELATED APPLICATION
The application claims the priority from provisional U.S. Patent Application
No. 60/402,233,
filed on August 12, 2002 entitled "System and its apparatuses for image
reproduction and
recording with the methods for positioning, processing and controlling".
to FIELD OF THE INVENTION
The present invention relates to an image reproduction and recording system
with a flexible
operation (hand, robot, vehicle) of head carrier, and the corresponding
apparatuses and methods
for positioning, processing, and controlling. The motivation is to build a
flexible operation (i.e.
without precise mechanical-apparatus for positioning) for image reproduction
and recording
t 5 system, instead of present conventional image reproduction and recording
systems in plurality
of uses. The conventianal mechanical-apparatus based systems are complex,
costly, and not
flexible, especially for very large printing area. Due to the flexibility of
this invention in
operation, the size of image that will be reproduced or will be recorded can
be as large as the
wall of a building and golf course, or can be as small as any size as long as
it still makes sense.
2Q Therefore, it can be used for plurality of applications, such as images and
patterns on building
wall or cliff, golf courses, basketball courts, football/soccer fields,
billboards, posters, portraits
and paintings, industry design blue prints, industry decorations, decoration
arts (such as
depositing a pattern on china arts), home painting and wall decorations,
archaeological
image/pattern taking and museum image/pattern backup, sculptures, etc. It can
be used for
z5 applications either on any flat surface, or on any curved surface. The
invention includes the
constitutions and designs of the system and its apparatus, including the
motion detectors,
operation modules, communication units, head carrier, operation unit, sprayer
/ sprayer-array
and image reader /reader-array. The invention also relates to the concepts,
ideas, theories, and
methods for positioning, processing, and controlling of the image reproduction
and recording
3o system, and relates to hardware signal processing and software data
processing.
BACKGROUND OF THE INVENTION
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
2
The conventional image reproduction and image recording aystems, such as the
printing
devices and scanning devices sold in the electronics store and those described
in US patents
5968271, 5273059, 5203923, 4839666, 5707689, 6369906, 5642948, 5272543 h~l etc
are based
on the track-guided positioning systems. The spraying head or reading
(recording) head is driven
by electric motors and is limited on two tracks in two directions through the
precise mechanical-
apparatus for positioning. Therefore, they have limitation in size and service
objectives, and they
have no flexibility far plurality of applications, such as image on
hardboards, on the walls, with
huge size or on a curved surface, etc. Also the conventional system is
mechanical-apparatus
based and so is complex and costly. So the motivation of this invention is to
build the flexible
l0 hand-operated, or robot-operated or vehicle carried systems for image
reproduction and
recording, instead of present conventional image reproduction and recording
systems. Due to the
flexibility of operation, the image that will be reproduced or will be
recorded can be as large as
the wall of a high building, or can be as small as any size as long as it
still makes sense.
Therefore, it can be used for plurality of applications mentioned above. It
can be used for either
any flat surface, or any curved surface. The invention includes the
constitutions and designs of
the system and its apparatus: motion detectors, communication units, head
carrier, operation
unit, operation modules, sprayer / sprayer-array and image reader /reader-
array.
SUMMARY OF THE INVENTION
The key spirits of present invention is the image reproduction and recording
system with a
flexible hand-operated or robot-operated or vehicle carried head carrier, and
the corresponding
apparatuses and methods for positioning, processing, and controlling. The
system is flexible,
(hand-operational, or robot-operational or vehicle carried), easy and very
convenient to use for a
plurality of users from industries, offices and home, home decorations,
entertainment and arts,
a5 etc., instead of the complex and costly precise mechanical-apparatus based
systems in present
conventional image reproduction systems.
A further object of the present invention is to provide system constitutions
and apparatuses
for head positioning, data processing, and head controlling.
To achieve the above objects, in the first aspect of the invention, the image
reproduction
(sub)system reproduces the image on any surface based on image data stored in
computer, by
causally moving the flexible-operation (hand, robot, vehicle) apparatus, i.e.
head carrier, on the
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
3
surface. The methods for image reproduction systems are classified into two
catalogs: the
wave-based method and relative-motion-based method. Both systems comprise
these
apparatuses: head carrier, sprayer / sprayer array, and a computer. Besides
these apparatuses, the
relative-motion-based method includes two relative motion detectors (MD) and
an operation
module (OM). The wave-based method includes the communication units (CL)) and
an operation
snit (0U).
In the relative-motion-based method, the system operation procedures include;
OM
executes the commands from computer to read the motion information of head
from MD, and
organizes this information as time-sequences. Then OM sends these time-
sequences to computer
l0 by mufti paths (in parallel). Computer processes the information for
locator positioning and
determining the coordinates of the head in the head array. The OM executes the
commander
from computer to control the action (spraying or reading) of the head in head
array. For
recording system, the OM takes the image information at each image pixel on
sensor array, and
organizes this information as time-sequences and sends them to computer. Also,
as the
alternates, any computer-mouse techniques can be employed as MD.
In the wave-based method, the system open anon p~ ocedures include: operation
unit (0U)
produces and sends the signal current to the transmitting CU. The transmitting
CU radiates and
the receiving CU receives the radio frequency (RFC, electromagnetic wave,
light or ultrasonic
signals that carry the information of the phase differences or the time
differences. The
information is sent back to the OU from the receiving CU. The OU processes and
converts the
information into the data of phase differences or time di$'erences, and sends
the data to
computer. Another alternate uses Doppler effect to detect the velocity of the
receiving CU, and
computer calculates the moving distance by integrating the velocity.
Computer processes these data and inverses the position coordinates of the
sprayerlsprayer
array by using the claimed positioning methods in this invention. According to
the position
coordinates, computer searches for the nearest pixel to this position in the
image data file stored
in disk of the computer and takes the color data of this pixel, and sends the
data to OU or OM.
Then OU or OM sends commands and power to the spray head to execute the jobs.
Computer
then records the history of the image-reproducing process. Any pixel on the
computer screen, of
which the corresponding image has been reproduced on the image surface, will
be marked by the
computer and will not be reproduced again if the sprayer moves back to the
same position later.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
4
The CU in the wave-based system or ML) in relative-motioned-based system is
also called
head locator. Usually there are two of them. With the first one, the second CU
or MD is used for
determining the sprayer array direction, so that the position of each
sprayer/reader in the
sprayer/reader array is determined.
In the second aspect of the invention, the image recording (sub)system takes
the image
digital data from any image surface to computer for storing and reproducing,
and also by
causally moving the hand-operation or robot-operation or velucle carried
apparatus, on the
surface. All apparatuses and procedures in the systems are same as that in the
image
reproduction systems, but use image reader/reader array instead of
sprayer/sprayer array.
Tagged by a trigger clock, the coordinate information and color data are taken
from the image
surface at the triggered moment and are sent back to the computer. The
computer processes the
information and data promptly or stores them into a file for processing
lately. The computer
inverses the coordinate information into coordinates. The coordinates at the
triggered moment
may not be just at a pixel on the pre-formatted pixel grids. So then the
computer calculates the
color values at all pixels on the pre-formatted pixel grids based on the
obtained coordinates and
color data, by using interpolation method.
In the third aspect of the invention, the apparatuses, i.e. motion detectors
(called head
locators) and operation module (OIVi], determine the relative position and
direction of head array
for the relative-motion-based system.
In the fourth aspect of the invention, the apparatuses, i.e. communication
units (CU), send
and receive the signal needed for determining the distance between
transmitting CU and
receiving CU for the wave-based system. The CU's installed on head carrier are
called head
locators.
In the fifth aspect of the invention, the apparatuses, i.e. operation unit
(0U) process the
signals from CU, convert the signal into distance-related data and provide the
data to computer,
and send the color data and action commands to the head, and also piovides
power supply for the
head. In the sixth aspect of the invention, the apparatus, such as the head
carrier or a hand hold
brush like body, provides a flexible operation to move the head with a
guaranteed constant fly-
height.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
In the seventh aspect of the invention, the computer software det,,~.nines the
position of
head or heads in head array according to the information from OU or OM, and
sends back to OU
or OM the commands and color data of the pixels corresponding to this
position.
The final aspect of the invention provides the theories, concepts, ideas, and
methods
5 corresponding to each structure, embodiment, apparatus, and procedure, for
positioning,
processing and controlling the .image reproduction and recording, including
hardware signal
processing and software data processing.
BRIEF DESCRIPTION OF THE DRA'1~JINGS
to A better understanding of the invention will be obtained by reading the
detail description of
the invention below, with reference to the following drawings, in which:
FiG. 1 is a view showing the constitution of one of the preferred embodiments
for the image
reproduction and recording system according to the invention, with the CU
(comnnunication unit)
on the corners, and the color material tanks on the head carrier or in the
cartridge that are build
together with the head.
FIG. 2 is a view showing the constitution of other preferred embodiments for
the system
according to the invention: (a) the color material tanks on the ground, (bj
three CU an the
corners, (c) four GU on the middle edges, (d) two CU on the bottom corners.
FIG. 3 is the schematic chart of one of the preferred embodiments for the head
carrier with
2U single head according to the invention.
FIG. 4 is the schematic chart of one of the preferred embodiments for the head
carrier with
head array according to the invention.
FIG. 5 is the schematic chart of one of the preferred embodiments for the head
carrier with
sprayer array on ink jet cartridge according to the invention.
. FIG. 6 is the schematic chart of the preferred embodiments for the
transmitting CU's: (a)
Radio frequency (RF) antenna, (b) single-light-source transmitter, (c) four-
light-source
transmitter, (d) ultrasonic transmitter.
FIG. 7 is the schematic chart of the preferred embodiments for receiving CU's:
(a) RF
antenna, (b) single-photo-detector receiver, (c) two-photo-detector receiver,
(d) four-photo-
detector receiver, (e) corner single-photo-detector, (fj corner single-photo-
detector with curved
substrate, (g) ultrasonic receiver.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
g
FIG. 8 is tree schematic chart of one of the preferred embodiments for
relation motion
detector (MD).
FIG. 9 is a schematic block diagram of the control and processing for one of
the preferred
RF-based systems according to the invention.
FIG. 10 is a schematic block diagram of the control and processing of another
of the
preferred RF-based system according to the invention.
FIG. 11 is a schematic block diagram of phase processing for the RF-based
systems.
FIG. 12 is a schematic block diagram of the control and processing of one of
the preferred
modulation-based systems according to the invention, with FOUR
wavelengthslfrequencies.
l0 FIG. 13 is a schematic block diagram of the control and processing of
another of the
preferred modulation-based systems, with TWO wavelengths/frequencies.
FIG. 14 is a schematic block diagram of the control and processiwg of another
of the
preferred modulation-based systems, with four waveiengths/frequencies.
FIG. 15 is a schematic block diagram of the control and processing of another
of the
preferred modulation-based systems, with two wavelengths/frequencies.
FIG. 1 fi is a schematic block diagram of the control and processing of one of
the preferred
time-based systems with an ultrasonic approach,
FIG. 17 is a schematic block diagram of the control and processing of one of
the preferred
time-based systems with another ultrasonic approach
FIG. 18 is a schematic chart of the contour curves fox constant phase
differences (hyperbola),
and constant phase sum (ellipse).
FIG. 19 is a flow chart of the position data processing and control for a
single head,
FIG. 20 is a flow chart of the position data processing and control for the
head array.
FIG. 21 is a schematic chart of the wrapping of current-phase-relation of in a
digital phase
detector (DPD) and the wrapped region in the 2-D phase space.
FIG. 22 is a schematic chart of data correlation processing for relative-
motion-based system:
image correlation conception and simple motion.
FIG. 23 is a schematic chart of data correlation processing for relative-
motion-based system:
complex motion.
DETAILED DESCRIPTION OF THE TNVENTION
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
7
The presc .t invention is to provide the image reproduction and recording
system= wiih
the flexibility, easiness, and convenience to use for a plurality of users
from industries, offices
and homes, and home decorations. The systems are flexible and consist of an
easy hand-
operation or robot-operation or vehicle earned apparatus, instead of the
complex and costly
mechanical apparatus-based systems in present conventional image reproduction
and recording
systems in plurality of uses.
< DICTIONARY >
For convenient in reading this invention, building a 'dictionary' for
definitions of some terms
is necessary. In this invention,
(1) In "Flexible operations": "hand-operation" means operation by hand of a
human being;
both "robot-operation" (such as the 'spiderrnan'-like) and "vehicle carried
operation" means the
powered-apparatus-aided operation, but without precise mechanical-apparatus
(such as track
guide for guiding the printing head or scanning head in the conventional
printer, or scanner) for
positioning, if the operation needs a power that exceeds the power of the
human being, or if the
environment of operation is not accessible for human being;
(2) The term "image generation" means reproducing (printing, painting,
spraying, and
deposition) or recording (scanning, and reading) image or pattern an or/and
from any surface.
(3) The term "image" in phases "image reproduction or image recording" has
dual meanings:
(a) any predetermined pattern or deposition to be reproduced, or any pattern
or deposition to be
recorded, which has already existed and was resulted from human's arts or
natural's arts; (b) the
image stored in computer, which could be recorded by scanner, or taken by
digital camera,
digital camcorder, etc.
(4) The term "head" in this invention means either sprayer for image
reproduction or reader
for image recording. Some time the "head" also means the part on which the
head is installed;
(5) The term "sprayer" in this invention means the ink jet, paint sprayer, or
any other devices
for material deposition. "Spray" or "spraying" means any action for material
deposition;
(6) The term "reader" in this invention means any device that takes the image
information
from a predetermined pattern or deposition, such as the image sensor in an
image scanner or in a
camera "read" or "reading" means any action of the reader;
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
(7) The "element" of an array is a general term referring to an element in one-
dimensional
array in positioning method description and claims. However, in image
reproduction or
recording system, it refers to a head in head array.
(8) The CU or MD built on head earner is called head "locator" in claims
"image
reproduction and recording system" '
(9) However 'positioning locator" in the claims of positioning methods is a
general term and
is not necessary only for "image reproduction or recording system";
(10) "Light" or '~rhoto" means visible or invisible, coherent or non-coherent
electromagnetic radiation from T-ray to X-ray;
( 11 ) 'electromagnetic waves (EMS" means all "Light" waves and all
electromagnetic
radiations from 530 kHz to I THz;
(I2) "Wave" mean means all EMW and ultrasonic waves;
(I3) 'information carrier" means RF wave or ultrasonic wave on which the
information is
ridding; while "carrier wave" means the light wave or millimeter microwave on
which the RF is
ridding (i.e. RF modulation);
( 14} The term "in a space" or "in image space" means on 2D flat surface or 2D
cwved
sur&ce, or in our real space (3D). It is a well-known knowledge that ID, is a
line, 2D space is
2D plane and 3D space is our real space;
( 15) The term "computer" means a programmable device (i.e. a generalized
computer) for
system and embodiment controlling.
(16) 'phase detector' means a mixer or a digital phase detector;
(i 7) "hand stick" means a device which provides the power to head-carrier for
making head-
carrier moving, it could be either hand-hold apparatus or powered-apparatus,
<SyS~M CONSTITUTION >
paragraph IJ Herein below are described the constitution and operation of the
system,
apparatuses, and the methods for positioning, processing and controlling, in
detail with references
to the accompanying drawings.
paragraph 2) FIG. 1 is used here to show the constitution of one of the
preferred
embodiments for the wave-based image reproduction and recording systems
according to the
invention. The system reproduces the image on the image area 10 of a surface
based on image
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
9
data stored in computer 900, or record the image data from image area 10 into
computer 900, by
causally pushing and pulling the "hand stick" 102 of head carrier 100 (or any
hand hold brush
like body) , on the surface. The surface can be any surface, such as curved
surface, sphere
surface, etc. The head Garner can be a hand-operational apparatus with a "hand
stick" 102, or can
be a powered-apparatus-aided apparatus for huge applications, or can be robot
operation, or
vehicle carned operation , if the environment of operation is not accessible
for human being.
paragraph 3J For image reproduction, four communication units (CU) 201 ~ 204,
used as
the transmitters / receivers with marks (Al, A2, B1, and B2), are set at the
four corners. The CU
(details in FIGS. 3 ~ 7 later) set on the head holder 300 are used as the
receivers / transmitters,
respectively. The information carrier can be either radio frequency (RF), or
light from T-ray to X
ray, or ultrasonic wave. however, if RF is directly (i.e. not modulation) used
as information
Garner, the CU must be set at corners or edges and must be fairly far away
from the boundaries of
image area 10, due to the nonlinearity of phase dependence of the near-field.
(paragraph 4J For convenience, the case of using the CU 201 ~ 204 as the
transmitters and
using the single CU (head locator) on head holder 300 as the receivei is
described here first. The
operation unit (0U) 400 produces signals and sends signal to CU 201 ~ 204,
through cables
51,52, 61,62. The cables 51 and 52 are split from one source, and have the
same length from the
sputter 50 to Al 201 and A2 202, so that they have the same time delay. The
same is applied for
cables 61, and 62; they have the same length from the sputter 60 to B1 203 and
B2 204. T'he CU
20I ~ 204 transmits the waves out. The receivers receive the waves with phase
or time
information and send the message back to the OU 400 through cable 20. The
hardware in
operation unit 400 processes the message and converts the message into phase
difference or time
difference, and sends, these data to computer 900 through cable 40. From these
phase data,
computer 900 inverses the coordinates of the position of the head locator
(details in FIGS. 3,4,5)
on head holder 300 by using positioning theories and formulas of this
invention. According to the
head position coordinates, computer 900 searches the nearest pixel to this
position in image data
file and takes the color data of this pixel, and sends the data to OU 400
through cable 40. Then
OU 400 sends action commands and power to spray head on head holder 300
through cable 30.
Any pixel on screen of computer 900, of which the corresponding image has been
reproduced on
3o the image area 10, will be marked by computer 900 and will not be
reproduced again if the head
on holder 300 moves back to the same position later.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
paragraph SJ For image recording, an image reader or reader array is installed
on the
head holder 300. The positioning procedures are the same as that for image
reproduction,
described above. Triggered by the trigger clock, the coordinate information
and color data are
taken from the image area 10 at the triggered moment and are sent back to
computer 900 through
5 OU 400. Computer 900 processes the information and data promptly or stores
them into a file for
overall processing lately. Computer 900 inverses the signal that carries the
coordinate information
into coordinates. The coordinates at the triggered moment may not be just at a
pixel in the pre-
fortrratted pixel grids. So computer 900 then calculates the color values at
all pixels in pre-
formatted pixel grids from the obtained coordinates and color data, by using
interpolation
to method.
(paragraph 6J The tru~mitter and receiver can be swapped. The CU 201 ~ 204,
Al, A2, B1,
B2, can also be used as receivers (serve as receiving CU), while the CU on the
head holder 300
can be used as transmitters (serve as transmitting CU). The details will be
descn'bed in sections
below.
(paragraph 7J The procedures described above are applicable for the all
preferred and
alternative constitutions described below.
paragraph 8J FIG. 2 shows the another preferred constitutions for 2-
dimensional (2-D)
applications according to the invention. The color material tanks are
necessary for large images
and are placed on the head carrier 100 (details in FIGS. 3, 4 and 5). However,
for huge images,
2o the color tanks 140,142 and 144 are placed on the ground. The color
materials are transported to
sprayers on the head holder 300, through tubes 130,132 and 134, as shown in
FIG. 2 (a). In FIG.
2 (b) is shown an option to use only three GU at three corners, with CU A1 201
and B 1 203
merged together. FIG. 2 (c) is an option to use four CU 201 ~ 204 on the
middle edges, which
provides the simplest positioning theories and formulas. For the time-based
positioning, the
embodiment shown in FIG. 2(d) is used, here only two CU A1 201 and A2 202 on
bottom corners
are used.
paragraph 9J For 3-dimensional (3-D) applications, another one or two CU's
need be
installed at any points (except too close to the image surface) on z-axis of
all the cases descn'bed
in FIG. 1 and FIG. 2. The z-axis vertical to the 2-D franne plan (image
suWce), and it could be
one edge of the 3-D frame. For all the Gases descn'bed above, CU can be either
fully or partially at
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
I1
either the middle edges or the corners of the frame, and the color tanks can
be either on the
head Garner 100, or on the ground.
(paragraph 10] The cables used for transmitting the phase-doesn't-matter
signal, color data,
and operation commands between operation unit 400 and head 300 can be replaced
by wireless
communication.
jparagraph I1J For the relative-motion-based system, with the optical-image
approach or
mouse-technique approaches, there are no CU (201 204) and OU 400, and the
cables between
them. Instead of CU and OU, MD and OM are installed together with head on head
holder 300
and causally moving on the image surface. The OM (not shown in FIGs.) is
directly connected
i0 with computer 900 through a mufti-path cable. Computer 900 periodically
sends the commands to
OM. OM executes the commands to read the motion information of the locators
from MD, and
organizes this information as tire-sequences. Then OM sends these time-
sequences to computer
900 by mufti paths in parallel through the cable. Computer processes the
information for locator
positioning and determining the coordinates of the head in the head array. The
OM executes the
commander from computer to control the action (spraying or reading) of the
head in head array.
One of the preferred MD's comprises a two-dimensional array of camera-image
sensors (M by N
pixels), two lenses, and one laser. For recording system, OM reads out the
image information at
each image pixel on sensor array, and organizes this information as time-
sequences and sends
them to computer 900, then computer stores this image information on disk.
Also, any computer
2o mouse techniques can be employed as MD.
< APPARATUS CONSTITUTIONS AND OPERATIONS >
Head carrier
(paragraph 12J FIG. 3 shows one of the preferred embodiments for the head
carrier with
single head according to this invention. The head carrier 100 is composed of a
frame I10 (main
body of head carrier, any shape), one front wheel 112, two rear wheels 114,
"hand stick" 102,
head arm 106, and head holder 300. The wheels (112, I 14) enable the caixier
100 moving on the
image area 10 freely, and guarantee a constant ffy height 301 for the head 382
(sprayer or image
reader) over surface 10. The "hand stick" 102 is connected with the head
carrier 100 by a joint
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
12
104, and the stick 102 can freely rotate about joint 104. The head arm 106 is
connected with
the head carrier 100 and can rotate about the axle 105 by hand-operation, for
flexible application
in different situations. The CU 381 and the head 382 are installed on the head
holder 300. Head
holder 300 is supported by head arm 106 at one end of the arm. For small image
applications, the
color materials are stored in the container built-in with the sprayer or color
cartridges. For large
image applications, three (or four if an additional black tank is needed for
color quality) color
tanks 120 (cyan), 122 (magenta), and 124 (yellow) are installed on the head
carrier 100, moving
together with the head earner. The color materials are provided to the head
from the tanks
(120,122,124) through color tubes 130,132 and 134. For huge image
applications, the color
1 0 ~teriaLs are provided to the head from ground tanks 140,142,144 (FIG. 2
(a)) through color
tubes t 30,132 and 134.
paragraph 13J FIG. 4 is used to show one of the preferred embodiments for the
head carrier
with head array according to this invention. The differences of this head
carrier from the one
described in FIG. 3 are in the head holder 300 and head cartridge 385 (instead
of single head). A
i$ number of heads are built on the head cartridge 385 arid form a head
(sprayer or reader) array
386. The image resolution (IR) is determined by head density in head array,
which is detern~ined
by the number of heads in the array and the array length Ll (391). Two CU
(383, 384) (i.e. two
head locators) are installed on the head holder 300. The holder extension 303
is needed to hold
one of the CU, 384, so as to extend the distance L2 (392) between two
locators, 383 and 384. The
2o purpose of this extension is to increase the accuracy in position
determination of each head in the
head array 386. The extension 303 can be added to either side of the head
holder 300, depends on
the convenience. The head holder 300 can rotate about the axle 302 by hand-
operation, by 360°,
for different situations of application.
(paragraph 1 QJ FIG. 5 is used to show another preferred embodiment for the
head carrier
25 yri~ sprayer array built-in an ink jet cartridge according to the
invention. The only difference
from the one descn'bed in FIG. 4 is that a color ink jet cartridge 389 with
sprayer array 390 is
now used.
Communication Units
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
13
paragraph 1 SJ The preferred options for transmitting CU (trar~...~itters)
according to this
invention are shown in FIG, 6, including (a) Radio frequency (RF) antenna 610,
(b) single light
source (Laser or LED) 630, (c) four light source 640, and (d) ultrasonic
transmitter 620.
paragraph 16J The RF antenna 610 is used as the transmitter for RF-based
system design.
The wavelength of RF that it radiates equals the dimension of image area 10.
Here is an
example: dimensions 100 meters, 30 meters, 3 meters, 10 centimeters and 1
centimeter are
corresponding to RF frequency 3 MHz, 10 MHz, lUOMHz, 3 GHz, and 30 GHz,
respectively. If
the technique for current-phase wrapping processing is used, the frequency can
be higher.
paragraph 17J For applications with larger image area, the lower frequency is
used.
to Therefore, the RF can be carned on (i.e. modulates) some extremely high
frequencies -
millimeter microwave, where the frequency allocation is empty and the use of
frequency is
unlicensed (such as those at peak absorption of atmosphere), so as to, avoid
to be interrupted with
public communication and military frequencies. In these cases, the same
procedures as that used
in light-based system descn-bed below are applicable, except the generator,
transmitter, and
receiver of carrier wave.
paragraph 18J For the light-based systems, the RF is carried on the light wave
by
amplitude modulation or frequency modulation. The light is emitted from the
emitter 632, called
single-light transmitter. For 2D application, by a cylindrical lens 634,
rather than by a spherical
lens, the light is uniformly divergent to the region with an angle 636 (any
angle between 90° and
150° is applicable, but 110° is preferred). The design of lens
and light direction makes the light
divergent as less as possible in the direction vertical to the paper plan. The
single-light
transmitter 636 is used for the system of which the transmitters are installed
at the corners of the
image plan. Four-light transmitter 640 is built by four of single emitter 630,
and is used for the
systems of which the transmitter is installed on the head holder 300. The
ultrasonic transmitter
620 is employed for the time-based systems. For 3D application, the lens is
spherical and the
six-light transmitter is used.
(paragraph 19J FIG. 7 is used here to show the preferred embodiments for
receiving CU
(receivers) according to this invention: (a) RF antenna 710, (b) single-photo-
detector 720, (b)
two-photo-detector receivers 730, (d) four-photo-detector receiver 740, (e)
corner single-photo-
3o detector 750, {e) corner single-photo-detector with a curved substrate 760,
and (g) ultrasonic
receiver 770. Due to the reciprocal principle of electromagnetic theory, those
described in R,F
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
14
transmitters a~ve are applied for RF receivers 710. The RF that ii carried on
an extremely
high frequency (millimeter microwave) is demodulated by heterodyne or homodyne
techniques.
paragraph 20J For the systems of which the receiver is on the head holder 300,
the two-
photo-detector receiver 730 (three-photo-detector for 3D), or the four-photo-
detector receiver
740 (and six-photo-detector for 3D), is used. They are built from a single-
photo-detector 720.
The latter is made up of photo sensor (photo detecting material) 728, light
wavelength-selection
filter 726, and cone mirror 724. The cone mirror 724 reffeets the light 722
from all directions to
the filter 726 and photo sensor 728. The current signal is generated fi-om the
sensor and is sent to
the operation unit 400. Inside the sensor, a pre-amplifier may already be
built in.
l0 paragraph 21J For the systems of which the receiver is at the corner of the
image plan, the
single corner photo-detector 750, or the one with a curved substrate 760 is
used. The Light 752
from different directions is focused on the photo-sensing material 728 by the
lens 754, so as to
increase the sensitivity, as shown in FIG. 7 (e) .and (~. Before sensor, there
is also an optical
wavelength-selection filter.
1 5 jparagraph 22J The ultrasonic receiver 770 is employed if the ultrasonic
transmitter 620 is
used in the system
Motion Detector and Operation Module
rparagraph 23J For the relative-motion-based system, the head includes a
motion detector
20 (MD), an operation module (OM), and a sprayer orland a reader. The
preferred apparatus for the
MD is the detector of optical image motion (340), as shown in FIG. 8. The MD
is built together
with the sprayer head 350 orland recording head (not shown in the figures).
The container 359 in
sprayer head 350 is a buffer for ink or paint material, which provides the ink
or paint noaterial for
the sprayers in sprayer array 352. The optical image motion detector 340
comprises laser 341,
25 lenses 342, 344, and camera pixel sensor array 346. The Laser 341 is
installed at a focus of the
lens 342, so the light is converted into parallel light beams and projects
onto the surface of the
path of head locator on image area 10. By lens 344, the optical image of the
object (a 'micro'
texture) 343 (any patterns, roughness distn'bution on the surface) appears on
the surface 345 of
the camera pixel sensor array 346. The light paths 348 for the image system
are shown on the
30 right side. The distance between the object 343 and the center of lens 344
is beyond two focus
length of Iens 344, while image 345 of the object is in between one and two of
the focus length.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
The OM witi~ a small volume (not shown in the figures) is installed together
w:~:i the
sprayerlreader and MD. OM executes the commands from the computer to read the
motion
information from MD, and organizes this information into time-sequences. Then
OM sends these
time-sequences data to the computer by mufti-paths in parallel. OM also
executes the commands
5 from the computer, after the computer finishes the processing, to control
the action of the head.
paragraph 24J For the recording system, the constitution is the same; the
sprayer-array is
replaced by the reader-array.
< SYSTEM OPERATIONS>
to (paragraph 25J The procedures of controlling and processing for one of the
RF-based
system according to the invention is shown in FIG. 9. The RF is directly used
as the information
carnet. The functions of operation unit (OLI) 400, of computer 900, and of
head 300 are shown
in the frarr~es of the left dish-line 401, the right dish-line, and the top
dish-line, respectively.
Before the system goes to work, the distributing signal and noise detector 411
searches the low
15 noise RF channels. According to the channel selection 412, the frequency ev
(higher) and
eev (lower) is determined (using these two frequencies, the frequencies ru, ,
r,~z , rv3 , a~,, for four
RF channels are generated). The oscillators 413 and 414 are tuned to these two
frequencies, and
amplified by amplifiers 415 and 416. The higher frequency is split into three
by sputter 417.
Two of them are sent to mixers 419 and 420 and one of them is sent to
frequency doublet 422
2o and then to a switch 423 (optional). The lower frequency is also split into
three by sputter 418.
One of them is sent to mixer 420 directly and the second is sent to mixer 419
after frequency
doublet 421. The third one is sent to a switch 423, which -is connected to
phase processor 430.
The two mixers provide the sum and differences of the two inputted
frequencies. With filters
424, four frequencies ( wl , ~2 , w3 , w4 ) are separated and are sent to four
transmitting antennas
211 ~ 214 at Al, A2, B1 and B2 shown in the previous figures. All four RF
channels are
amplified by amplifiers, 425. The receiver 311 receives the four signals from
the four
transmitters (411-414). After the band amplifier 426, the amplified four
signals are split into four
paths by sputter 427. The band pass filters, 428, allow only one frequency
pass through each one
of them. Phase processor 430 decodes the phase differences between A1 and A2,
and the phase
d~erences between B1 and B2, if the switch is turned to down side. Or, phase
processor 430
decodes the phase sums of A1 and A2, and the phase sums of B 1 and B2, if the
switch is turned
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
16
to up side. More details at~out the phase processor are described later with
FiG. 11.
Phase calibration can be done by either the software in computer 900, or bythe
phase calibrator
431 before signal goes into computer 900. The same procedures for signal
processing are applied
for the receiver on the second locator shown in FIG. 4 (384) or in FIG. 5
(388). Computer 900
receives two groups of the phase messages for the positions of the two head
locators (i.e. the
antenna receivers), (432,433) and (444,445).
paragraph 2GJ Computer 900 processes the phase data by inverting the
coordinates of the
positions of the two locators from the phase data, which is based on the
positioning theories and
formulas of this invention. According to the coordinates of the two locators,
computer 900
to calculates the coordinates of each of the head in head array (386, in FIG.
4) by using interpolation
method. According to the position of each head, the computer 900 searches the
nearest pixel its
the image data file to this position and takes the color data of this pixel,
and sends the data to
control unit 429. Then the control unit 429 sends the action commands and
power to head 308
through color cables 306 and power cable 307.
paragraph 27J FIG. 10 shows the procedures of controlling and processing for
another RF-
based system according to the invention. The difference here is that the
transmitter and receiver
are swapped from the system described in FIG. 9. The four RF channels are
combined together
by combiner, 434, before being sent to the transmitting antenna 321. The four
receiving antennas
receive the signals and send the signals to four band-pass filters, 435, which
allow only one
frequency to pass through each one of them. The four channels are then sent to
phase processor
430 after being amplified by amplifiers, 436.
paragraph 28J One procedures of phase processing for the RF-based systems are
shown in
FIG. 11 (a), the first two frequencies are conducted to mixer 4301, which
produce another two
frequencies, the sum and difference of inputted frequencies. The band pass
filters 4303 filter out
the sum frequency. At this point, the signal with the difference frequency
carries the phase
difference between A1 and A2. The digital phase detector (DPD) or mixer 430$
decodes the
phase difference by homodyning with the signal from 423. The phase difference
4315 (A2-Al) is
sent to the computer. The same is applied for the other two ,frequencies. The
output phase
difference 4314 (B2-B 1 ) is sent to the computer.
paragraph 29J Another phase processing procedure for the RF-based systems is
shown in
FIG. 11 (b). The largest and the smallest frequencies are conducted to mixer
4307, which also
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
17
produces two frequencies, the sum and difference. But the band pass filters
4309 filter
out the difference frequency, rather than sum frequency. At this point, the
signal with the sum
frequency carries the phase sum of A1 and A2. The digital phase detector (DPD)
or mixer 4311
decodes the phase sum by homodyning with the signal from 423. Then the phase
sum 4317 (A2-
Al) is sent to the computer. The same is applied for the two middle
frequencies. The phase sum
4316 (B2-B 1 ) is sent to the computer.
paragraph 30J FIG. 12 shows the procedures of control and processing for one
of the
modulation-based systems according to this invention. In this system, 1(F is
used as modulation.
The carrier wave of this RF wave is light or miuirneter microwave. For the
miuimeter microwave
to carrier, the frequency with peak absorption (6(7~70 GHz, 120~130GH~, and
170180 GHz, for
example) is preferred but not limited, where the frequency allocation is empty
and the use of
frequency is unlicensed, so as to avoid to be interrupted with public
communication and military
frequencies. Here, the laser, as carrier-wave, is used for illustrations. The
laser driver 437
provides four currents to four lasers (231234) to emit four wavelengths or
frequencies
~1,~2,~3,~4. The lights from all lasers are modulated by the same RF signal
with frequency
w . This RF signal is generated by the RF oscillator 413 and amplified by
amplifier 415. The RF
sputter 438 splits the RF signal into four paths and sends the RF to each
laser (231 234), so that
the light power or light frequency is modulated. The four-detector receiver
331 converts the light
power into RF currents (either coherent or non-coherent detection is used, but
here using non-
coherent as example). Each of the detectors has a different optical filer (726
in FIG. 7) to allow
only one of the four frequencies S2, , S22 , S~3 , S24 to Pass through. The
currents are sent back to the
four RF band pass filters 439 to pass RF frequency c~ . After amplified by
amplifiers 440, the
phase differences of first two signals and the last two signals, 433 and 432,
are recovered by DPD
441 and 442, respectively, and are sent to computer 900. If the mixer is used
at 441 and 442, the
hers 443 are needed, before the signals are sent to computer 900, for
filtering out high frequency
if the phase difference is used, or for filtering out the low frequency if the
phase sum is used.
paragraph 31 J' In the cases of using miljimeter microwave as the carrier
wave, the same
procedures for controlling and processing in the light-based systems
descix'bed above and below
are applicable, except for the generator, transmitter and receiver of Garner
wave.
paragraph 32J In the cases of using a mixer at the last step before the
message goes into
computer 900 above and below, the message is not directly the phase difference
or phase sum, but
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
18
is the sinusoidal function of them So the computer softy;are converts the
message into
phase difference or phase sum for these cases.
paragraph 33J The procedures of control and processing for another light-based
system, but
with two wavelengths, are shown in FIG. 13. The transmitters 243, 244 at A1
and A2 emit the
same light wavelength (or frequency S2, ), while transmitters 241, 242 at B1
and B2 emit the
same frequency SZa . One of the two receivers, 341, filters out the second
light frequency and
detects the two signals that are carried by the first frequency S2, (from A1
and A2), and then
sends the two detected RF signals to 0 RF band pass filter 448. The two
signals are internally
homodyned at mixer 452 after the amplifier 450. The output from the mixer 452,
after a low pass
l0 her 45g, is a sinusoidal function of the phase difference, which is sent to
computer 900. The
other receiver, 342, filters out the first frequency and the sends the two
detected RF signals (from
B1 and B2, and carried by the second frequency s22 ) to a RF band pass filter
449. The dish line-
framed part (446, 454, 455, 456, 457) is an option for using the phase sum.
(paragraph 34J The control and processing of another light-based system, with
four
IS wavelengths, is shown in FIG. 14. The difference from the system descn'bed
in FIG. 12 is that the
transmitters and receivers are swapped. The four-light-source transmitter 341
is installed on the
head holder 300. Four corner receivers 24i ~ 244 are used.
paragraph 35J FIG. 15 is a schematic block diagram of the control and
processing of
another light-based system. All the procedures for this system are the same as
that in the system
20 described in FIG. 14, except that only two wavelengths or frequencies are
used.
paragraph 36J The system with its alternatives described in FIGS. 9 to 15 is
based on the
phase measurement approaches, called phase-based system. The system can be
also based on the
time measurement approaches, called time-based system. The information carrier
for the time-
based system is, usually, ultrasonic wave, but it can be also any kind
electromagnetic wave (light,
25 or millimeter microwave) as long as we have fast-enough clocks in the
future or for huge
applications. Here, the system is illustrated by an ultrasonic-based approach
as shown in FIG. 16
and FIG. 17. The clock 475 sends periodic commands (triggers) for the pulse
generator 476,
which generates a pulse-modulated current with an ultrasonic fiequency. After
the current power
is amplified at amplifier 477, the current is sent to transmitter 371. The
ultrasonic pulse is
30 ~~ed out from transmitter 371 and is received by receivers 271 and 272. In
the meantime,
the power amplifier 477 has also an output signal for the start triggez 478 to
trigger the time
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
19
counters 480 and 481, so as to start time- counting at the moment the
ultrasonic wave is
sent out. After the receiver 271 and 272 receive the pulse, the signal is
immediately (speed of
electrical is far greater than the speed of sonic) amplified by the amplifiers
482, and is sent to
triggers 484 and 485 to stop the time-counting. Then the time counters 480,
481 send the time
differences to computer 900. The ultrasonic frequency filters 483 are used to
distinguish the pulse
from the other transmitters (384 in FIG. 4, or 388 in FIG. 5) on head holder
extension 303 in FIG.
4, because the two transmitters are driven by different ultrasonic
frequencies.
jparagraph 37J FIG. 17 is used to show the control and processing of the time-
based system
with another ultrasonic-based approach. The difference from the system
descn'bed in FIG. 16 is
1o that the transmitter and receiver are swapped. IVIore clearly, the
receiving CU 381 is on the head
holder 300 rather than the transmitting CU on the head holder. Two ultrasonic
pulse generators
488 and 489 are used to produce two driving currents with different
frequencies. Therefore, the
tw~ ultrasonic pulses with different frequencies are transmitted from the
transmitters 281 and
281. The mixed signal from receiver 381 after amplified by amplifier 498 is
split into two paths
by sputter 495. Each of the filters, 496, or 947, blocks out the other
frequency and sends the pulse
to triggers 484 and 485 to stop the time-counting.
jparagraph 38J The Doppler effect is an alternative used for relative motion
detection Only
two transmitting CU (transmitters) at the bottorn corners (such as Al, A2 in
FIG. 2(d)) and two
receiving CU (receivers on the head holder) as two locators are used. Instead
of producing pulse-
modulated PULSES ultrasonic wave or electromagnetic wave, the generators 488,
4$9 in FIG. 17
generate an oscillation current with two frequencies a fair away from each
other, and the
transmitters 281, 282 in FIG. 17 radiate CONTINUOS ultrasonic waves or
electromagnetic wave.
Receiver, 38I in FIG.17, is replaced by a Doppler-Frequency-Detector. When
receiver 381 is
moving around in the two wave-fields, the Doppler frequencies, which carries
the information of
two velocity components along two directions, are detected. One direction is
from one
transmitter A1 (281) to the receiver 381; the other direction is from the
other transmitter A2 (282)
to the receiver 38I. So the angles o~ two directions are timely changing while
receiver 381 is
moving. The Doppler fi~equencies are sent to computer 900. Computer 900
converts the two
Doppler frequencies into velocity components and calculates the two
displacement components
of the receiver (i.e. locator) by integrating the velocity components. Then
from the displacement
components, the relative position of the locator is determined.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
jparagraph 39J The other alternative of image reproduction and image recording
system is to use any mouse-technique-based positioning method for determining
the relative
position of the locator.
5 <COMPUTER PROCESSING >
jparagraph 40J INTRODUCTION --- Computer processing procedures are classified
into
two cases: using phase difference, or using phase sum. Also, there are two
kinds of phase
dependencies on the coordinates of the point in the image area 10. For
modulation-based systems
descn'bed above, the phase dependence on the coordinates is linear; while for
the RF-based
10 systems descn'bed above, the dependence is nonlinear due to phase
nonlinearity of the near-field
and the distortion from the boundary conditions. For the case of linear phase
dependence and
using phase difference, the contour curves for constant phase differences are
a class of hyperbola
curves, as show in FIG. 18(a). While, for the case of linear phase dependence
and using phase
sum, the contour curves for constant phase sum are a class of ellipse curves,
as show in FIG.
15 18(b). All the hyperbolas or ellipses have the common foci at four CU's
(Al, A2, B1, B2), this is
the general conclusion whenever the CU's are located at the corners or at the
four middle edges.
This invention provides the general theories, relations, and formulas for all
cases: linear or
nonlinear, phase difference or sum. This invention also provides the general
cafbration method
. for the case of distortion from the boundary conditions, or nonlinearity.
Computer processing is
20 based on these theories and formulas.
jparagraph 41J CALIBRATION and INITIALIZING (1) --- The communication units,
381
in FIG. 3, 383 and 384 in FIG. 4, 387 and 388 in FIG. 5, are also called head
locators as
mentioned before. Usually there are two locators in the image reproduction and
image recording
system of this invention. With the first locator, the second is used for
determining the head array
direction, so that the position of each head in the head array is determined.
The computer
processing procedures, by using phase difference (rather than sum) and linear
(rather than
nonlinear) phase dependence, and for the case of only one locator, are
descn'bed here first for a
convenient understanding of this invention. A schematic block diagram for the
illustration is
plotted in FIG. 19. The procedure starts with initialization: calibration and
initializing image
status of the pixels. First of all, check the status of the calibration --
911. If the calibration is not
done, put the locator at (0,0), the center of the image area 10, and read out
the voltage (or current)
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
21
for phase differences (PD) (between A2 and Al, and between B2 an;: B1) from
the receiver --
912. Usually, at this point, the PD is not zero. The zero-PD cah'bration at
(0,0) can be achieved
either by hardware (phase shifter) adjustments, or by computer processing.
FIG. 9 shows an
example of phase shifter 431. By adjusting the phase shifter, the PD at (0,0)
can be reduced to
g zero. If by computer processing, these two non-zero PD's will be stored for
later use - 913. Next
put the locator at any corner of the image area and read out the PD - 914.
Procedure 915
calculates the PD changes and distance differences (DD) when the locator moves
from the center
to the corner. The DD are defined as the difference of the two distances: the
distance between Al
(or B 1 ) and the head locator, i.e. r~~ (or rB~ ), and the distance between
A2 (or B2) and the head
to locator, i.e. rAa (or r~2 ), -- 915. Then the cafbration coefficient is
determined by the ratio of DD
over PD change -- 916, which is the proportional coefficients between DD and
PD change, and is
used for converting the PD change to DD during the operation. The final step
of initialization is to
initialize image status by setting all P(i) =0 ( f denotes the i-th pixel) --
91?. The computer also
figures out the scale transformation between the image area 10 and the image
source stored in
computex. According to the size of the image area 10, the computer will
produce a frame on the
computer screen according to the scale, and the operator can move the frame on
the screen to the
source area that he will most likely to reproduce. The status of any pixel
outside the fi~ame is
initially set to 1. However, it is initially set to 0 if the pixel is inside
the frame. For arty pixel of
which the corresponding image has been reproduced on the image area 10, the
status P(i) will be
20 changed to f from 0. If the status of a pixel is 1, the image of this pixel
will not be reproduced
again during the head causally moving. However, multiple reading from same
pixel and
overwriting the old reading doesn't matter.
(paragraph 42J CALIBRATION and INITIALIZING ( 1 ) --- If cafbration is done,
then
jump over the calibration block (the left dish-line frame) and wait for the
commands (a trigger)
for taking the phase information -- 920 that is sent from the phase processor,
such as the one 430
in FIG. 9. By using the PD at (0,0) and the cafbration coeffcient, the two
DD's are determined -
- 921.
(paragraph 43J COMMON PROCEDURES OF COMPUTER PROCESSING {1) --
Procedure 922 and hereafter are the common procedures for different cases:
linear or nonlinear
30 p~ dependencies, using phase difference, or phase sum, or time difference.
Procedure 922 is to
solve the roots of an equation that includes the DD data, and gives the
locator position (x, y). The
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
22
equations are sifferent for the different cases listed at the begi:.-~ing of
this paragraph.
Procedure 923 takes all the stored image data 924. Then checks the status of
each pixel - 925. If
the pixel has been sprayed (P (i) =1), the next pixel is checked. If all the
pixels have been sprayed
(all P(i)=1 ), good job is down, and stop - 926. If there is at least one
pixel with status P (i)=0,
then to tell how close this pixel to the locator position (x, y) - 927. If the
distance is less than or
equal to the criteria (1/20 ~ 1l4 of the pixel spacing is preferred), then
procedure 928 takes the
color data of this pixel from the image file 924, and then sends the commands
for spraying - 929.
Meanwhile, procedure 928 sets the status to 0 for this pixel. If the distance
is greater than the
criteria, then check the next pixel with status 0. If there is no such pixel
that satisfies this
1 ~ condition at ali, then it will wait for the next trigger for the next
chance --930 of meeting a pixel
that is spray-able, during head causally moving.
jparagraph 44J COMMON PROCEDURES OF COMPUTER PROCESSING (2) --- Of
course, as shown in the right dish-line frame, there are alternative
procedures (932, 933, 934) for
improving the efficiency with a cost, if there is no such a pixel (i.e. it's
distance from present
is position of the locator is Less than the criteria) at all. Two 'fast-
response micro-motors (not shown
in Figures) are used to slightly adjust the head position. Procedure 933 finds
the pixel in the
image source which is the nearest to the locator position (x, y) at the
moment. The computer
predicts how much the head should be moved to that pixel, by taking in the
account the velocity
of the head motion and the response time of the micro-motors-driven head, and
then move the
2o head to that pixel -- 934.
jparagraph 45J POSITIONING OF EACH INDIVIDUAL HEAD IN HEAD ARRAY ---
For the case with two locators, the calibration is made for each of the two
locators first -- 935 and
936 in FIG. 20. After the two pairs of calibration coe~cients are obtained for
the two locators,
the status of each image pixel in the image source is initialized to 0 --937.
Now the computer
25 ~e~ the phase information from the two locators - 938, then the two pairs
of DD (distance
differences) for the two locators are calculated by using the calibration
coe~cients --939. In the
same way as in the one-locator case, the position coordinates of the two
locators, (x ~i~, y ~~~ and
(x ~2~, y ~2~, are obtained - 940, and the status of each pixel is checked -
94I, 942, 943. If all the
pixels have been sprayed/read (all P (i) =1 ), stop -- 944. Otherwise, the
program uses
3o interpolation method to determine the position coordinates of each head
along the head array --
946: x(.1)=x~')+Dxx(j-1)a Y(j)=Yt'~+Dyx(j-t), DI= (x~2~-x~'~)IN,Dy=(yc2)-
y~'~)lN.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
23
Here IJ is the total number of head along the head array, and j (--1,~, ...,
ll~ denotes every
head. Procedure 947 checks every head (j=1,2, ..., 11~ on the array --- if the
distance between any
head and any pixel is less than the criteria? If yes, the computer takes the
color data from that
pixel and set status to 1 - 948, and then commands that sprayer to spray ~
949. Procedure 951 in
the dish-line frame is an alternate for improving the e~ciency, if there is no
such pixel at all, or if
there is only one such pixel In this case, three fast-respons~°, micro-
motors are used to slightly
adjust the sprayer array position and direction, so that each sprayer on the
array can aim at a
corresponding pixel Two motors are installed at the end of the array [at the
side of locator
383(FIG. 4) or 387(FTG.S)], and the third motor is installed at the other end
of the array [at the
side of locator 384 (FIG. 4), 388 (FIG.S)]. The third motor drives the head
array rotates about
axle at the first sprayer. Similar to procedure 933 in FIG. 19, the computer
find out the pixel in
image source which is the nearest to the position (x~l~, ytl~ of locator 1 at
that moment (this is
also the position of the first sprayer --- this is just a corresponding
relation, not necessarily
physically the same), The computer predicts how much the array should be moved
to aim at that
pixel, by taking in the account the velocity of head motion and the response
time of micro-
motors-driven spray head, and then move the array so that the first sprayer
aims at that pizceL
Meanwhile, the computer predicts how much the array should be rotated to make
each of the
sprayers aim at a corresponding pixel, by taking in the account the moving
trend and inertia. Then
the micro-motor rotates the array to the predicted angle and computer commands
the sprayers to
spray.
paragraph 46J INVERTING THE LOCATOR' S POSITIONS BY SOLVING
EQUATIONS --- For the modulation-based method, the phase has a linear
dependence on the
distance ( r) between the receiver and the transmitter. For the given two
pairs of detected phase
difference (O~OA = phase of A2 - phase of Al, and Ag~B= phase of B2 - phase of
B1), or phase
s~ (~~~ ~d ~~B )' ~e position coordinates (x, y) of the locator are the roots
of the equations
(xcos8, +ysin9,)2 /at - (-xsin8,+ycos9,)a !b; = 1 and -(xcos~2 +ysin6a)2 /b2 +
(-x sin Ba + ycos~2 )Z / a2 = 1. When Al, A2, Bl, B2 are at the corners, 81 is
the angle between
the line Al-A2 and the right direction of the horizontal line and 92 = 90 - 8,
. However, when Al,
A2, BI, B2 are at the middle edges , 81~ and 9a = 0. For the phase difference
approach,
~~ - DA2-A1 (d~tance between A1 and A2, same meaning hereafter), cx = DB2-a,,
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
24
a~ =c~ -kAOq~A, az =cz -kBnrp$. For the phase sum approach, c~ 'O.sDAZ_A1~
cz = O.SDBZ_s, , a, = O.SkRE~p,, , az = O.SkBE~a . For both cases, b; = c2 -
a; ( i =1,2 ). For the
phase difference approach, b; is a pure real number, and the contour curves
for constant phase
differences are a class of hyperbola curves, and the right root-pair (x, y) is
uniquely distinguished
from the four pairs of the roots by checking the signs of the two phase-
differences. Here is an
example: consider the case Al-A2 is vertical to BI-B2, as the cases shown in
FIG. 2(b) and {c).
The phase information (Q~pA < 0 and ArpB < 0) corresponding to the solution
with (x > 0, y > 0);
( A~p,~ < 0 and O~B > 0) E ~ (x > 0, y < 0); ( ~~pa > 0 and Arpg < 0) ~ ~ (x <
0, y > 0); and
t o ( 11~,~ > 0 and Ag~B > 0) E -~ (x < p, y < 0). While, for the phase sum
approach, b; is a pure
image number, and the contour curves for constant phase sum are a class of
ellipse curves, and
the right root-pair (x, y) cannot be distinguished from the four pairs of the
roots by using the
phase information. In this case, the computer program sets the region ID
(identification) for the
four quarter-regions (left bottom=1, right botton~2, left top=3, right top=4).
The operator inputs
the locator region ID from a keyboard when the locator begins to move. The
computer then
changes the region ID whenever the locator moves across the region boundaries.
Therefore, the
right root pair (x, y) is distinguished from the region ID and the moving
trend.
paragraph 47J INVERTING THE LOCATOR' S POSITIONS BY SURFACE FITTING --
The above procedures for modulation-based method are characterized by the
linear dependencies
of the phase. For the RF-based system, radio frequency (RF) is . directly
(i.e. not used as
modulation) used as the inforn~atioa carrier. The phase has a nonlinear
dependence on distance
(r) between the receiver and the transmitter due to the near field:
~p(r)=kr-tan-'[(kzrz-i)l(kr)], k is propagation constant of RF wave. The
coordinates of
locator position can be determined by finding the minimum point of I (x, y) _
[q~(r~z ) -,~(r," )
W Spy ~z + [S~(r~a ) - ~(ra~ ) ' ~ Pa as ~ or I(x, Y) _ [4~(raz ) + ~(ra~ ) '
~ ~A J2 '~' [~P(raz )
- ~ ~Pa ~ z . Here ( ~q~A and A~OB ), or ( E ~pA and ~ spa ), are the detected
phase
differences, or phase sums, respectively. The first minimization starts at the
initial point that is
defined by the roots of the linear equations from the Linear limit (the larger
r) of the phase
3o dependence, and the later minimization starts at the previous position of
the locator. The
boundary condition of the electromagnetic filed may introduce a discrepancy of
the phase
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
dependence from the formula above, which is determined by the environment and
cannot be
predicted ahead. If the discrepancy is significant, a calibration method is
employed. The
cafbration method is to mesh the image area 10. Move the locator at each node
on the mesh. The
computer then records the phase difference and the coordinates of the node.
Then the computer
5 uses the surface functions to fit the coordinates versus the phase
difference by using numerical
methods (such as finite element method). By using *.hese surface functions,
the computer
determines the coordinates from the phase difference when the locator moves to
any position on
the image area 10.
jparagraph 48J PHASE-CURRENT PROCESSING (1) --- For both cases ofusing digital
1o phase detector (DPD) ar mixer, the phase shiftexs built in the operation
units are so adjusted that,
for the zero phase (i.e. phase difference between two inputs of the DPD, or
mixer), the output
current is zero. The DPD outputs a Iir~ear current that is proportional to the
phase (i.e. phase
difference between the two inputs of the DPD) in the region ( - 2~c , 2~ ).
However, the curve is
wrapped out of this region for every 2~tof increase or decrease in phase, as
shown in bottom of
15 FIG. 21. The mixer outputs a current that is proportional to the sine
function of the phase. So the
monotonous region is (- ~r l 2 , ac l 2 ). Out of this monotonous region,
there are the other
monotonous regions for every ~ increase or decrease in phase. So usually, if
the noise level is
low enough, only the middle region is used for both cases. Therefore, the
wavelength of the 1ZF
modulation or RF-carrier should be the maximum dimension of the image area 10
if DPD is used;
2o while it should be 4 times of the maximum dimension of the image area 10 if
mixer is used. So
using the DPD will result in 4 times better signal to noise ratio (SNR) than
using mixer if the
noise level is same, that is, the resolution is 4 times better by using DPD
than by using mixer. The
minimum (or best resolution) is determined by the noise level.
jparagraph 49J PHASE-CURRENT PROCESSING (2) -- For higher resolution
25 applications, if the noise level cannot be reduced, the current-phase
wrapping needs special
treatment by the computer program For the case of using DPD, as shown in the
top of FIG. 21,
the phase space is divided into (2M-1 )2 regions, with M =3 as an example.
This leads to M times
better resolution. Each region is assigned to an identification (ID) number
(ij)(i, j =1,2,..,2M 1),
(f = 1, 2, ...) denotes the number of wrapped regions for the phase difference
between B 1 and B2,
~d ~ = 1, 2, ...) denotes the number of wrapped regions for the phase
difference between A1 and
A2. The computer always changes the ID number if the locator moves across the
boundary and
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
2b
into a new region. Therefore, before head starts moving at center region, the
computer
initializes ID at the center region, that is, set ID = 33, for the case as
shown in FIG. 21. Then the
locator is moved to the position where the operator wants to start the work,
and the computer
follows the regions that the locator passes and promptly changes the ID
numbers. Finally, for
example, the locator moves to the region S I through a path, and the computer
follows the locator
and fnally changes the ID number to 51 starting from 33. Now the button for
active spraying is
switched on and the sprayers start to work. lxt's define phase-current and
detected-phase-current.
The detected-phase-current is the output of the DPD (the solid lines in bottom
of FIG. 21). The
phase-current is the processed current without wrapping and is scaled to phase
(the dish lines m
bottom of FIG. 21), and the phase is the phase-difference used in the next
computer processing as
descn'bed above. For the non-center region, the phase-current should jump a
value from the
detected-phase-current. As shown in FIG. 21, for regions 34 and 35, the phase-
current for phase
difference of A2-A1 should shift to 959 and 960, respectively, from the
detected-phase-current
(955 and 956). Or, in other words, in the regions 34 and 35, the phase which
is directly (i.e.
15 ~~out shift) determined from the detected-phase-current should add 2~ and
4n , respectively.
If using a very large AI, this method can also serve as an alternate for
relative-motion-based
system that will be descn'bed below.
Cparagraph SOJ PHASE-CURRENT PROCESSING (3) --- For the case of using the
mixer,
the procedures are almost the same, except the region size (all ~r x ~c ,
rather than 2n x 2~r ,
20 2~r x 4~c , 4~c x 2~c , and 4~r x 4~c in the DPD case) and the sine
dependence of the detected-phase-
current on the phase (rather than linear dependence). Therefore, the detected-
phase D~pd is
determined by inverting the sine function from the detected-phase-current. The
phase is
transformed from Og~d , such as ~ - ~~pd and 2~r + dg~d for the first right
region and the second
right region from the center region, respectively, for example.
25 jpara~.aph SIJ COMPUTER PROCESSING OF TIME-BASED METHOD --- For the
time-based method (i.e., based on the time measurement), the computer receives
two time
differences, tAl and tAa from the OU 400, which are the times for the pulse
propagation from the
CU at A1 and A2, as shown in FIG. 2(d), to the CU as head locator. Then the
computer solves the
root pair (x, y) from the equations (x - x~, )a + (y - y~, )~ -- (tAw)a and (x
_ xA2 )2 + (y _ yA2 )2
_ (t,~ZV)a. Here v is the speed of the pulse propagation. If the computer so
sets the coordinate
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
27
system that ~y,~~ =0, yAZ =0, and at the vertical central-line of the image
area 10,
x = 0, then the two pairs of roots with negative y coordinates are dropped.
The root (use
negative sign for x if tAl < too, use positive sign for x if tA~ > t,~2) with
the positive y coordinate is
the solutiozt.
[paragraph 52J COMPUTER PROCESSING OF OPTICAL IMAGE-MOTION-
DETECTOR BASED APPROACH --- For the relative-motion-based method, the head
includes
a motion detector (MD) and operation module (OM). The preferred apparatus for
the MD is the
optical image-motion-detector (340), as shown in FIG. $. The camera pixel
sensor array 346
converts the optical image into electrical signals, which are sent to the
computer's memory for
1o digital processing. The head starts moving at the center of the image area
10 after the initial
setting of the reference point of the relative motion at this point. At this
moment, a picture is
taken --- the middle panel that are shown in FIG. 22, (a) represents the
windows of image 964.
The position of the image 965 is defined as the left bottom corner of the
window. At this moment,
the image position is at 964. At the moment of the next trigger, the head is
moving to the position
966 marked by the filled circle. The picture-taken frequency should be high
enough so that
between the two neighboring pictures, the position just changes a distance of
a few pixels, even if
with the fastest moving. Especially, if the head starts from static state, the
position just changes a
distance within one pixel. The computer starts the analysis by determining the
image-correlation
at assumed positions, one of the positions is at 967 for example. The image-
correlation is defined
2o as the averaged summation of squares of differencz (or absolute value of
difference --preferred)
. of light intensity at pixels over the common image area (thin dash line) of
the two pictures. One
of pictures is the previous picture 964, the other is preset picture 96$ but
at assumed position 96?.
So if the image-correlation is smaller, the assumed position is closer to the
actual position 966.
From FIG. 22 (a) and (b), the image-correlation for position 967 (FIG. 22 (a))
is larger than that
for position 969 (FIG. 22(b)).
/paragraph 53J HEAD SPEED UP MOTION --- For the cases that the head starts
moving
or restarts to move after the speed is reduces to zero, the computer will
determine the relative
position of this picture 971 to the previous picture 972 in FIG. 22 (c). I-
Iowever, the computer
does not know along which direction the head is moving. The computer
calculates the image-
3o correlation at five points (i.e. five assumed positions) and then uses a
surface to fit five points of
the correlation. Computer finds out the maximum point on the surface (or the
minimum point of
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
28
the 'negative surface'), which is (or very close to) the actual position 971
of the image at the
present moment. Among the five points, one point is called the surface-fitting
center 972, which
is, at this moment, at the previous position. The other four points are at the
four nearest corners
(open circles in (c)) of the surface-fitting center. Hereafter, the term
"frame" represents the
quadrilateral fi~ame, of which the four corners are at the four outer points
with the center at the
surface-fitting center, for five-point fitting. And it represents the
pentagonal frame, of which the
five corners are at the five outer points with the center at the surface-
fitting center, for six-point
fitting below. It would be lucky if the maximum correlation point is inside
the fi~ame (as shown
in (c)). If the maximum point is inside the flame but too close to the
boundary, one more point
l0 973 on the lower side of the surface is needed, and computer redoes the
surface fitting by six
points, for better accuracy. For each new position at a new trigger moment,
the first surface
fitting is made by always using.f'n'e points. The second-and-after sur&ce
fitting are made by
always using six points.
~parag~~aph 54J HEAD SIIvvIPLE MOTION --- If the head is moving, the computer
stores
the history data of the head positions. From these data, the, head movement
trend (the velocity and
acceleration) can be determined. Therefore, the position of next picture at
the next triggered
moment can be predicted at 974 (by extrapolation), as shown in FIG. 22(d),
although the actual
position is at 975. Then the computer finds the nearest pixel to the predicted
position 974, and
uses this pixel as the surface-fitting center, and repeats the procedures
described in the above
section (with FIG. 22(c)). If the prediction is accurate enough (i.e. head
motion is not complex),
the actual position (that is the maximum point) of the picture at this moment
should be inside the
frame. So the same later procedures described in the above section (with FIG.
22(c)) are applied.
Otherwise, the computer should finish the following procedures.
(paragraph SSJ HEAD COMPLEX MOTION --- The extrapolation-predictable motion is
called simple motion, otherwise it is called complex motion. If head complex
motion, the
prediction is not efficient. Therefore, as shown in FIG. 23, the actual
position (may at A or B) of
the picture at this moment is out off the fi~ame of which the center (surface-
fitting center) is at
977. The center 977 is the closest pixel to the predicted position 976. This
means that there is no
maximum point in the frame center at 977. Therefore need RE-SETTING SURFACE
FITTING
CENTER: the computer needs to compare the values of the correlation at the
four corners, and
picks out the point with lowest correlation value, Y~ (i.e. the point 980 for
the case shown in FIG.
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
29
23(a)), and picks up the values of the two neighbored corners, y ; for y79 and
Y2 for 981.
Then the computer defines the two variables: Rl = min ~Vi - Y~ I l Y~ , ~Yz -
V~ I ~ Y~ ~ and
RZ = (Y, - Ya I l Y~ , and sets the two criteria's CRi (say 0.5) and CRZ (say
0.2) which need
optimization. Here min ~ ~ means taking the minimum value in the list. If R~ >
RC, , RZ > RCZ
and V < YZ , then use 983 (in FIG. 23 (a)) as the next surface-fitting center.
If R, > RC, ,
RZ > RCa and Y, > YZ , then use 985 (in FIG. 23 (b)) as the next surface-
fitting center. If
R, > RC, , RZ 5 RCZ , then use 987 (in FIG. 23 (d)) as the next surface-
fitting center. If R, 5 RC,
and V < YZ , then use 988 (in FIG. 23 (c)) as the next surface-fitting center.
If RI <_ RCS and
y1 > yz ~ ~~ ~ 989 (~ ~G. 23 (c)) as the next surface-fitting center. This
time, we don't need
four points, but three or two points (thin open circles) around the new
surface-fitting center will
be added. The surface-fitting is carried out by six (rather than five) points,
the center and the
newly added points plus same old paints. If using the old points [980 and 979
for the case in (a),
980 and 981 for the case in (b), 977, 980,979 for the case in (c), and 980,
979 or 981 for the case
in (d)~, the point 984 in (a) or 986 in (b) is not necessary. If a maximum
point is~found in the new
fr~e~ the actual position of the head at the present moment is obtained.
Otherwise, the computer
repeats the procedures until the actual position is found.
jparagraph 56J DOPPLER EFFECT METHOD --- The Doppler effect of ultrasonic or
electronnagnetic wave is used for positioning. Here use ultrasonic wave as an
example. For the
system based on ultrasonic Doppler effect, the generators 488, 489 in FIG. I7
generate the
a0 oscillation current with two frequencies a fair ways from each other, and
the transmitters 281, 282
in FIG. 17 radiate continuous ultrasonic waves. Receiver 381is replaced by a
Doppler frequency
detector. When the receiver 381 is moving around in the two ultrasonic fields,
the Doppler
frequencies are detected. The two velocities (v~ and v2) facing the two
ultrasonic wave sources are
inverted from the two Doppler frequencies, respectively, by the computer. Then
the displacement
of the head facing the two sources can be obtained by ~,r, _ ~ v, dt and ~rz =
~ v~ dt ,
respectively. Here DT is the time spacing of the two neighboring triggers. If
the head positions
relative to the two sources at the moment of last trigger are i~,Q and r~ ,
respectively, the head
displacement is ~ = Dr, ~'° + t~rz X20 . Then the head positions
relative to the two sources at
rio rzo
CA 02530977 2006-O1-30
WO 2004/015980 PCT/US2003/025111
the moment W present trigger can be written as r, ~ no + ~r and - z = rzo + ~
, respectively.
Now the computer solves the root pair (x, y) from equations (x - xA, ) z + (y -
yAi ) z -- r~2 ~
(x ' xAZ )z + (y- yaz )z ~ ~'i . ~e next procedures are same as that in the
time-based system
descn'bed above.
s (paragraph 57J JUMP HAPPENS --- In relative motion method, if a jump happens
to the
head carrier during the its moving on the image surface due to some reason,
the head needs to be
put back to the center of the image area 10 for initially resetting the
reference point of the
relative displacement.
(paragraph S~J RECORDING SYSTEM --- For the recording system, the
constitutions and
10 procedures are the same, except that the sprayer array would be replaced
the by reader array.
20
