 |
Claims  |
|
|
What is claimed is:
1. In an article handling system of the type having transport means for
transporting articles along respectively different transport paths, the
improvement comprising:
(a) means associated with each of said articles bearing data representative
of the intended destination of said article,
(b) a uniquely configured target symbol of predetermined and known
characteristics immediately adjacent said destination data, said target
symbol including a set of geometrically similar parallelograms of
contrasting colors superimposed in mutually concentric relation which,
when intersected by a line through its center, defines a sequence of
contrasting bars and spaces, said bars and spaces having predetermined
widths along said center line,
(c) means responsive to the respective width ratios of adjacent bars and
spaces for initially detecting the presence of said target symbol and
thereafter reading the intended destination data adjacent said detected
symbol, and
(d) control means responsive to the so read intended destination data for
directing each article along a transport path corresponding to said
intended destination.
2. The improvement as defined in claim 1 wherein said data is human
readable.
3. The improvement as defined in claim 1 further comprising electronic
means for determining the width ratio sequence of adjacent bars and spaces
of said detected target symbol and comparing said determined sequence with
a known width ratio sequence of adjacent bars and spaces of a reference
target symbol, thereby to identify said uniquely configured target symbol.
4. An automated baggage handling system, comprising:
(a) transport means for transporting baggage items along different
transport paths corresponding to respectively different intended
destinations of said baggage items,
(b) tag means affixed to each baggage item bearing data representative of
the intended destination of said baggage item, said data being both
machine and human readable,
(c) said tag means further bearing a uniquely configured target symbol
immediately adjacent said intended destination data, said symbol having
predetermined and known characteristics,
(d) camera means including a plurality of video cameras disposed adjacent
said transport means, each video camera being directed to capture the
image of information on tag means passing within its field of view,
(e) target identification means coupled with each video camera for
detecting the presence of said target symbol within said field of view,
(f) means coupled to said target identification means for reading the
intended destination data viewed by only that video camera having the
largest image of said target symbol,
(g) optical character recognition means for reading said intended
destination data, and
(h) control means responsive to the data read by said optical character
recognition means for directing the baggage item along the transport paths
corresponding to the intended destination of that baggage item.
5. The automated baggage handling system as defined by claim 4 further
comprising electronic means for transforming the image of the target
symbol detected by the target identification means to a desired
orientation.
6. The automated baggage handling system as defined by claim 4 wherein the
reading of said intended destination data by said optical character
recognition means is dependent upon the initial detection of the presence
of said target symbol by said target identification means.
7. The automated baggage handling system as defined by claim 4 wherein said
target identification means comprises electronic means for comparing the
characteristics of the image captured by said camera means with the
predetermined and known characteristics of a reference symbol.
8. A method of automatically routing articles to respectively different
destinations, said method comprising:
(a) affixing tags to each of said articles, each tag bearing human readable
data representative of the intended destination of that article as well as
bearing a target symbol of predetermined and known characteristics
adjacent said data,
(b) identifying the data presence and a reference direction of said target
symbol on said tags,
(c) thereafter electronically transforming the image of the so identified
target symbol to an image of desired configuration and orientation,
(d) thereafter machine reading the intended destination data adjacent the
transformed image of the said target symbol, and
(e) routing the baggage to the destination corresponding to the so read
intended destination data.
9. Automated baggage handling system, comprising:
(a) transport means for transporting baggage items along different
transport paths corresponding to respectively different intended
destinations of said baggage items,
(b) tag means affixed to each baggage item bearing human readable
alphanumeric data representative of the intended destination of said
baggage item,
(c) said tag means further bearing a uniquely configured target symbol
immediately adjacent said intended destination data, said symbol having
predetermined and known characteristics,
(d) first camera means disposed adjacent said transport means for capturing
and digitizing the image of scenes within its field of view,
(e) means for comparing the so digitized image with data representative of
a reference target symbol,
(f) means solely responsive to a match between the digitized image and data
representative of the reference target symbol for subsequently processing
the intended destination data adjacent said target symbol, said processing
including the optical character recognition of said alphanumeric data, and
(g) control means responsive to the optical character recognition of said
alphanumeric data for directing the baggage item along a transport path
corresponding to said alphanumeric data.
10. An automated baggage handling system, comprising:
(a) transport means for transporting baggage items along different
transport paths corresponding to respectively different intended
destinations of said baggage items,
(b) tag means affixed to each baggage item bearing data representative of
the intended destination of said baggage item, said data being both
machine and human readable,
(c) said tag means further bearing a uniquely configured target symbol
immediately adjacent said intended destination data, said target symbol
having a plurality of contrasting squares superimposed in mutually
concentric relation which, when intersected by a line through its center,
defines a sequence of contrasting bars and spaces, said bars and spaces
having predetermined widths along said center line,
(d) camera means disposed adjacent said transport means so directed to scan
tag means passing within its field of view,
(e) target identification means coupled with said camera means for
detecting the presence of said target symbol within said field of view,
said target identification means including means for determining the width
sequence of adjacent bars and spaces of said detected target symbol and
comparing said determined sequence with a known width sequence of adjacent
bars and spaces of a reference target symbol, thereby to identify said
uniquely configured target symbol,
(f) means responsive to said determined width sequence of said detected
target symbol for transforming the detected images of the target symbol
and of the intended destination data to a desired orientation for
presentation to optical character recognition means,
(g) optical character recognition means for reading said transformed
intended destination data, and
(h) control means responsive to the data read by said optical character
recognition means for directing the baggage item along the transport paths
corresponding to the intended destination of that baggage item. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
The present invention relates generally to article handling, more
particularly to baggage handling and routing, and even more particularly
to an improved method and system for optically detecting and reading human
readable destination data on tags affixed to the article or baggage.
There are many article handling applications requiring efficient and
automated routing of the articles to respectively different locations. For
example, in airline baggage handling systems, there is an ever increasing
need for reducing the time in which passengers' baggage is processed and
loaded on to the particular flights for the intended passenger
destinations.
In an attempt to meet this need, various systems have been developed for
automating the handling and transport of the baggage, none of the existing
systems, however, being entirely suitable for all conditions of service.
For example, many proposed baggage handling systems contemplate encoded
data representative of the intended destination of the baggage being
affixed or associated therewith, with various types of alternative
detection devices utilized to decode the destination information and
effect selective routing of the baggage to the desired locations. However
existing systems, among other disadvantages, have not been successful in
effectively meeting two major objectives desired by the airlines and/or
users of the systems, namely: (1) that the machine readable encoded
destination data also be human readable for the convenience of both the
airline agents and the passengers and (2) that means be provided to
rapidly identify and read the destination data under circumstances where
the tags or substrates upon which the coded data is printed are randomly
located and oriented.
It is therefore the principal object of the present invention to provide a
new and improved method and system for article, and particularly baggage,
handling.
It is another object of the present invention to provide a new and improved
method and system for the automated transport and routing of articles and
baggage in accordance with coded destination data associated with such
articles and baggage.
It is a still further object of the present invention to provide the
aforementioned automated handling and routing where the encoded data is in
human readable form and capable of being detected and read when in various
positions and orientations.
In accordance with these and other objects, the broad concept of the
present invention contemplates the use of a tag or the like bearing
encoded data, which is both machine and human readable, indicative of the
intended destination of the particular article to which the tag is affixed
or otherwise associated, the tag also bearing a uniquely configuredttarget
symbol positioned adjacent the coded destination data to facilitate the
detection and reading of the destination data by cameras or other
appropriate sensing devices. Specific features of the invention, as well
as additional objects and advantages thereof, will become more readily
understood from the following detailed description taken in conjunction
with the accompanying drawings, in which:
FIG. 1 is a diagrammatic representation of a baggage handling system for
selectively transporting baggage to respectively different locations;
FIG. 2 is a depiction of one form of a baggage tag bearing the human
readable destination data and a uniquely configured target symbol
incorporating the features of the present invention;
FIG. 3 is a simplified block diagram schematic of the overall system
utilizing the concept of the present invention;
FIG. 4 is a block diagram schematic illustrating a preferred embodiment of
the target identification network illustrated in FIG. 3; and
FIGS. 5A and 5B are illustrations depicting the electronic transformation
of the video image from the face of the tag of the desired orientation and
configuration.
The drawings are not necessarily to scale, and in some instances portions
have been exaggerated in order to emphasize the various features of the
invention. Furthermore, the term "destination data", as used throughout
the specification and claims, is to be construed in the broadest sense and
includes any identifying indicia or the like which indicates the desired
routing of the articles such as, in the case of baggage handling, flight
data (airline i.d. and/or flight number), airport destination, etc.
Referring initially to FIG. 1, one type of baggage handling conveyor
assembly is diagrammatically depicted as including a main conveyor belt 1
with branch conveyers 1a, 1b, and 1c intersecting the main conveyor at a
diverter location 2, the main and branch conveyors being motively powered
to move at a constant rate in the direction of the arrows. This conveyor
assembly is effective to transport various types of parcels and luggage,
such as baggage piece 3, to the diverter location 2 where a conventional
diverter (not shown) selectively directs the luggage to one of the branch
conveyors 1a, 1b, or 1c in accordance with the intended destination of the
particular luggage. For instance, in accordance with one illustrative
example where the baggage handling conveyor assembly is used in an airport
operation, each of the branch conveyors 1a, 1b, and 1c would transport the
luggage to respectively different locations or "accumulation points" for
loading on flights headed for various destinations.
It is to be understood that the baggage handling conveyor assembly does not
in and of itself constitute a part of the present invention and can be any
one of a number of types conventionally known in the art. Additionally,
the type of conveyor assembly depicted in FIG. 1 is only illustrative,
with the control of the present invention capable of being utilized with
alternative or modified forms of conveyor assemblies. For example, the
"branch" conveyors 1a, 1b, and 1c, instead of comprising movable belts,
can be inclined chutes. Furthermore, while the particular conveyor
assembly, and inventive control therefor, are described with respect to
the transport of baggage, it will be clear that such can also be utilized
for the transport of various types of packages and articles, for example
in a mail sorting operation in which event the baggage piece 3 would be a
mail sack.
Turning now to the process and control, and in accordance with the overall
concept of the present invention, a luggage tag 4 is attached to the bag
3, the tag bearing a uniquely configured target symbol positioned adjacent
data representative of the intended destination of the particular bag. One
or more cameras 20 are positioned and focused to capture the pertinent
information on the tag as it passes within the cameras' field of view at a
location upstream of the diverter location 2. As subsequently described in
greater detail, the destination information on the tag associated with
each piece of luggage is then processed to actuate the diverter so that
when the luggage reaches the diverter location 2, it is respectively
directed along the correct path (1a, 1b, or 1c), in accordance with its
intended destination.
In accordance with a unique feature of the invention, and with specific
reference now to FIG. 2, each tag 4 has imprinted thereon a symbol or
"target" 10 of predetermined and known characteristics, the target being
positioned immediately adjacent an information field 5 containing the
printed or written "intended destination" data. As hereinafter described
in greater detail, the target 10 enables its (and therefore the tag's)
real time identification/location by the computerized control of the
present invention so that the destination data within the field 5 can be
automatically read and processed.
In accordance with a preferred embodiment shown in FIG. 2, the target 10 is
in the form of a square symbol having concentric squares which, when
intersected by any line through its center, defines a sequence of bars
(11, 11', 11", etc.) and spaces (12, 12", 12", etc.). Each of these bars
and spaces has a specific predetermined width associated therewith so
that, in the example shown, a scan line through the center of the symbol
will intersect the bars and spaces in the respective ratios
2:1:1:1:3:1:1:1:2.
In accordance with the unique advantage of the present invention, and as a
preferred embodiment thereof, the destination data within the field 5 can
take the form of a combination of printed and hand-written alphanumeric
figures, which not only affords the advantage of being human-readable, but
which can be machine-readable by conventional optical character
recognition (OCR) techniques. For example, as depicted in FIG. 2, the
airport destination (DFW) can be preprinted with the specific flight
number (44) hand-written by the agent at the departure point. As indicated
in FIG. 2, lettering blocks 5A can be provided to serve as a constraint
for the hand-written symbols, thus facilitating the OCR readings of such
symbols.
A simplified block diagram of the overall system for carrying out the
unique process of the invention is depicted in FIG. 3 which illustrates,
in this example, the use of three cameras 20, 20', and 20". While various
types of cameras or sensors may be employed for the units 20, 20', and
20", including but not limited to area cameras or line scan cameras, in
accordance with a preferred embodiment, the units 20, 20', and 20"
comprise Model CCD1500R line scan cameras produced by Fairchild Camera and
Instrument Corporation of Palo Alto, Calif.
The video scene data captured by each camera is routed to respectively
associated target identification means (6, 6', 6"), such means being
effective to detect and identify the target 10, and therefore signal the
presence and location of the tag 4 when it passes within the field of view
of the particular camera. The video scene data from each camera is also
routed by way of memory units 7 (7', 7") to control means 8 which, in
addition to the other functions hereinafter described, is effective, in
response to signals from the target identification means 6 and under
circumstances where the target has been identified (and the specific tag
thus located) by more than one camera, to process only that tag data
viewed by the camera having the best and largest image of the tag.
The control means 8 is also effective to electronically transform the image
of the data on the face of the detected tag 4 to the desired configuration
and orientation (for example, to the format like that shown in FIG. 2) and
thereafter, using OCR techniques, to read the destination data contained
within the information field 5. This destination data is then inputted to
a diverter controller 9 which is effective to automatically and
appropriately actuate the diverter of the baggage handling conveyor
assembly (FIG. 1) to direct the luggage 3 along the correct path as
directed by the destination data read from the luggage tag.
In summary, therefore, the basic sequential steps carried out by the
process and control of the present invention are:
(1) Identify the uniquely configured target, thereby identifying the
presence and location of the luggage tag;
(2) Electronically transform the video data corresponding to the
information on the face of the detected tag to the desired orientation and
configuration;
(3) Electronically read the destination data adjacent to the target; and
(4) Direct the baggage carrying the tag along a path determined by the
destination data electronically read from the tag.
TARGET IDENTIFICATION
The target identification process basically involves the comparison of the
video scene data from each of the cameras with data corresponding to the
characteristics of the uniquely configured target symbol 10. Since the
specific disclosed target comprises a sequence of adjacent bars and spaces
of specific and known relative widths, the approach contemplated by the
invention is to initially appropriately binarize the incoming video data
to render the various pixels as a sequence of bars and spaces, to then
electronically "measure" the width of these binarized bars and spaces, and
thereafter compare the relative widths of these bars and spaces with the
bar/space width characteristics of the specific known target 10. As a
particular feature of this process, since the energy density at the
photosites of the camera varies as a function of not only the illuminance
at the tag but also the camera-tag distance, it may be preferable to use
adaptive, rather than fixed threshold comparative, binarization of the
incoming data.
Consequently, in accordance with the process, the incoming video scene data
is adaptively binarized in real time to detect black-to-white and
white-to-black transitions, i.e. in the case of a candidate target, to
identify the bar/space transitions thereo. The next step of the process is
to measure the widths of these detected bars and spaces, followed by
appropriate comparison of such widths with the widths of the bars and
spaces of the reference target 10. A "match" between such data is then
indicative that the target 10 (and hence tag 4) has been detected. Various
techniques, if desired, may then be employed for validating such match.
A block diagram schematic of a preferred embodiment of the system for
implementing the target identification step is depicted in FIG. 4.
Accordingly, video data generated from the line scan camera 20 is inputted
by way of data line 21 to an edge detector network 30. Concurrently, the
video data is also transferred by a local bus to, for storage in, memory
unit 7.
The camera 20 will normally have associated therewith conventional digital,
bias/gain, and scaling circuitry so that the video data appearing on line
21 actually constitutes digitized binary signals, which have been
appropriately adjusted for correct bias and gain as well as appropriate
scaling, these binary signals corresponding to the sequence of gray level
values of the "scene" (pixels) comprising the field of view of the camera.
Under the circumstances where the target 10 becomes part of that scene,
the video data representative of the target would thus constitute binary
signals corresponding to the black bars 11, 11', 11", etc. and spaces 12,
12', 12", etc. (FIG. 2). At this point in time, however, it is to be
understood that the target scanned by the camera 20 is only a candidate
target, not yet having been identified as the known target 10.
The function and purpose of the edge detector network 30 is to locate the
black-to-white and white-to-black transitions, i.e. in the case of the
candidate target, the bar/space interfaces, the output signals from the
network 30 (representing such transitions) then being input to, for
resetting of, a counter 40. The counter 40 is effective to generate
signals at its output representative of the respective widths of the
detected bars and spaces, these signals being inputted (by way of data
line 41) to a target matching network 50. The purpose of the target
matching network 50 is to compare the width data appearing on the data
line 41 with stored data representative of the unique characteristics of
the target 10 and to generate a signal (on the output data line 71) in the
event of a "match", i.e. signifying the detection and identification of
the actual target 10.
The operation of the video camera 20, the edge detector network 30, counter
40, and target matching network 50, as well as coordination of the various
functions discussed hereinafter, are all under the supervision and control
of a central processor unit 25 (which provides the functions of control
means 8). While various types of equipment may be utilized for the CPU 25,
one preferred type is the single board computer iSBC 286/10 produced by
Intel Corporation of Santa Clara, Calif.
The edge detector network 30 includes a derivative filter 31, the function
of which is to generate a signal at its output representative of the
instantaneous rate of change of scene reflectance with respect to its
position in the scan. The filter 31, in a preferred embodiment, comprises
two interconnected digital filter/correlator chips, for example those
presently produced by TRW Inc. of La Jolla, Calif. under the designation
TDC-1028. The positive and negative coefficients are inputted to the
filter 31 by the CPU 25 by way of data line 22.
Inputs to the edge detector network 30 also include the data clock signals
on line 23, the line clock signals on line 24, and the beginning of line
(BOL) address which specifies the address in memory where the beginning of
each new scan line from the camera 20 is stored.
The edge detector network 30 also comprises a counter 32 which is clocked
by the data clock signals on line 23 and reset by the line clock signals
on data line 24. The counter 32 therefore generates, at its output, signal
data corresponding to the offset within a line of the pixel which is being
inputted (by way of data line 21) to the derivative filter 31. In the
instance where the target 10 is detected by the matching network 50, this
"offset" is stored in a section of the memory 56 of the matching network
50 for subsequent input to the central processor unit 25, all as
subsequently described in greater detail.
The output derivative filter 31 is coupled to three sequential delay
registers 33A, 33B, and 33C. The data at the output of delay registers 33A
and 33C are coupled to a conventional subtractor 34, the output signal
ddxnew of which corresponds to the second derivative of the incoming video
data signal. This signal is then applied as one input to the edge
detection logic network 37.
The output of delay registers 33B and 33C are coupled to, and summed
within, a conventional adder network 35, the output of which corresponds
to the first derivative of the video data signal. The output of adder 35
is coupled to a sign detect and thresholding logic network 36 which
generates an output signal dxsgn (constituting the sign of the derivative)
and a flag signal dxdth (which constitutes a binary "1" when the absolute
value of the derivative is greater than a pre-set threshold. The three
signals ddxnew; dxdth, dxsgn; and a fourth signal ddxold (being a delayed
version of ddxnew) are then inputted to edge detector logic 37. Edge
detector logic network 37, in a preferred embodiment, comprises a single
integrated fuse logic (IFL) device having internal feedback capability.
The edge detection logic network 37 is effective to generate a first output
signal tclk which has a binary value "1" when either a black-to-white or
white-to-black transition is detected, as well as a second output flag
signal bw which constitutes a binary "1" if the transition is
white-to-black and "0" if the transition is black-to-white.
In summary, therefore, the overall edge detector network 30, in response to
the incoming video data signals, generates a signal (tclk) indicative of
the presence of a black-to-white or white-to-black transition (bar/space
interface and a signal (bw) indicative of the specific type of transition
(i.e. bar-to-space or space-to-bar).
The counter 40 is effective to actually measure the distances between the
black-to-white and white-to-black transitions which, in the case of a
candidate target, means measuring the widths of the bars and spaces of
that target. Accordingly, the counter 40 is reset each time that the
signal tclk goes "high", indicating the occurrence of a transition, the
duration of the signal at the output of counter 40 (on data line 41) thus
being representative of the distance between transitions, i.e. the width
of the bar or space, as the case may be, most recently detected. The
counter can be of any configuration known to those skilled in the art; and
in the preferred embodiment, utilizes a programmably array logic (PAL)
device.
Since the black-to-white and white-to-black transitions of the video data
will occur at uneven intervals, the clock signal furnished to the various
components of the target matching network 50 will necessarily be
irregular. Accordingly, these clock signals are provided by the clock
generator 38 in response to signals tclk and bw inputted thereto, as shown
in FIG. 4.
As previously discussed, the purpose of the target matching network 50 is
to compare the signal information appearing on the data line 41
(representing the "candidate target") with stored information
representative of the unique characteristic of the reference target 10 in
order to determine whether there is a "match". Specifically, the network
50 compares the sequential bar/space width data (or ratios) appearing at
the output of the counter 40 with the sequential bar/space widths (or
ratios) of the target 10, the network for implementing such now being
described in greater detail.
Accordingly, the target matching network 50 includes a correlation filter
51, the filter 51 having inputted thereto (on data line 52) a binary
signal (from CPU 25 representative of the bar/space widths of the target
10, such signal thus constituting the reference pattern of such target. As
previously indicated, the reference pattern (or width ratios) for the
particular target 10 shown in FIG. 2 would be 2,1,1,1,3,1,1,1,2.
Coupled to the output of the correlation filter 51 is a "match" signal
generator 53 which is effective to generate an appropriate signal, for
example a logic "1", at its output when the bar/space width
characteristics of the data stream on data line 41 matches the bar/space
width data of the reference pattern inputted to the correlation filter. As
an additional advantageous feature, if desired, and in order to
accommodate the fact that the perceived width of the respective bars and
spaces may vary as a function of the proximity (distance and angular
orientation) of the camera to the tag 4, suitably designed proximity
adjustment means 54 is provided to make appropriate adjustments for such
variable. Accordingly, the data appearing at data line 41 is routed
through both the correlation filter 51 and the proximity adjustment means
54 so that the signal generator 53 is responsive to the "proximity
adjusted" data to assure the generation of the "match" signal at the
output means 53 when the actual and true target is presented in the field
of view of one (or more) of the cameras 20.
Also included as part of the target matching network 50 is a
first-in/first-out (FIFO) memory means 56 in data communication with the
central processing unit 25 by way of data line 72, as well as in data
communication (by way of data line 73) with control means 55. The control
means 55 is effective, in response to a "match" signal from generator 53,
to generate an "interrupt" signal to the CPU 25, as well as to instruct
memory 56 (by way of data line 73) to store the relevant information in
memory as a consequence of this match. Such relevant information, which is
then read by the CPU 25 by way of data line 72, includes the offset data
then existing from the output of counter 32 and the BOL (beginning of
line) address. Furthermore, in accordance with a particular advantageous
feature of the preferred embodiment, an additional item of relevant
information inputted to the memory 56 (by way of data line 74) is a data
signal from proximity adjustment means representative of the relative size
of the detected target. Therefore, since a target identification network
like that depicted in FIG. 4 is associated with each camera (20, 20',
20"), under those circumstances where more than one camera identifies the
target 10, the CPU 25 will select that camera having the largest image of
the target, and thus the tag, for subsequent orientation and reading of
the tag destination information viewed by that camera.
TAG DATA TRANSFORMATION
Since the tags, and consequently the image of the target and tag
destination data (within the field 5) will in most instances be at random
orientations from the horizontal orientation shown in FIG. 2, OCR reading
of such tag destination data will be facilitated by an initial electronic
transformation of the viewed image. For example, the captured image of the
target may appear in the configuration and orientation like that shown in
FIG. 5A (the image of the adjacent destination data being similarly
distorted and askewed); and in accordance with a feature of the invention,
the image of the target is electronically rotated, sized, and configured
to appear as that shown in FIG. 5B, such electronic transformation being
similarly applied so that the image of the adjacent tag destination data
(airport designation "DFW" and flight number "44") appears like that shown
in FIG. 5B.
In accordance with a preferred embodiment, the aforementioned
transformation is carried out by the CPU 25 using software implemented
algorithms which "locate" the corners of the target 10 and compute an
affine transformation to a standardized view. Specifically, and with
refere | | |
