U.S. patent application number 10/697533 was filed with the patent office on 2008-08-14 for mail image profiling and handwriting matching.
This patent application is currently assigned to Lockheed Martin Corporation. Invention is credited to Alfred T. Rundle, Lennart A. Saaf, Richard C. Vanhall.
Application Number | 20080192978 10/697533 |
Document ID | / |
Family ID | 39685846 |
Filed Date | 2008-08-14 |
United States Patent
Application |
20080192978 |
Kind Code |
A1 |
Rundle; Alfred T. ; et
al. |
August 14, 2008 |
MAIL IMAGE PROFILING AND HANDWRITING MATCHING
Abstract
Method and systems for early detection of potentially hazardous
material associated with the collection of mail or other objects.
In an embodiment of the method of this invention, image
characteristics for an image of an item being examined are
obtained. A predetermined profile is then retrieved from a profile
database. The image characteristics for the image of the item are
compared to the image characteristics present in the predetermined
profile. If the image characteristics present in the predetermined
profile substantially match the image characteristics for the image
of the item, the item is identified for special processing. If the
image characteristics present in the predetermined profile do not
substantially match the image characteristics for the image of the
item, another predetermined profile is retrieved and compared to
the image characteristics for the image of the item. A system that
implements the method of this invention includes a transport
sub-system and a computing sub-system.
Inventors: |
Rundle; Alfred T.; (Endwell,
NY) ; Saaf; Lennart A.; (Endicott, NY) ;
Vanhall; Richard C.; (Owego, NY) |
Correspondence
Address: |
BURNS & LEVINSON, LLP
125 SUMMER STREET
BOSTON
MA
02110
US
|
Assignee: |
Lockheed Martin Corporation
|
Family ID: |
39685846 |
Appl. No.: |
10/697533 |
Filed: |
October 30, 2003 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
60422311 |
Oct 30, 2002 |
|
|
|
Current U.S.
Class: |
382/101 ;
209/584 |
Current CPC
Class: |
G06K 9/222 20130101;
G06K 2209/01 20130101; B07C 1/00 20130101; G06K 9/00422 20130101;
G06K 9/2063 20130101 |
Class at
Publication: |
382/101 ;
209/584 |
International
Class: |
G06K 9/00 20060101
G06K009/00; B07C 5/00 20060101 B07C005/00 |
Claims
1: A method for identifying mail pieces having postage sides for
special processing based upon a plurality of predetermined
profiles, the method comprising the steps of: a) selecting a
combination of values for individual profile image characteristics
to create each one of the plurality of predetermined profiles, the
values being chosen such that the combination is consistent with
threat mail, the individual profile image characteristics being
associated with various locations on the postage side of the mail
piece; b) storing the plurality of predetermined profiles, each of
the predetermined profiles having values for the plurality of
selected profile image characteristics; c) obtaining mail piece
image characteristics associated with an image of a mail piece; d)
retrieving one of the plurality of predetermined profiles that has
not yet been used in a comparison; e) comparing the mail piece
image characteristics to the one of the plurality of predetermined
profiles; f) identifying the mail piece for special processing, if
the one of the plurality of predetermined profiles substantially
matches the mail piece image characteristics; and g) repeating
steps d) through f) for another one of the plurality of
predetermined profiles, if the one of the plurality of
predetermined profiles does not substantially match the mail piece
image characteristics.
2: (canceled)
3: The method of claim 1 wherein the step of comparing the mail
piece image characteristics to the one of the plurality of
predetermined profiles occurs during real-time processing.
4: The method of claim 1 wherein the step of comparing the mail
piece image characteristics to the one of the plurality of
predetermined profiles occurs during offline processing.
5: The method of claim 1 wherein the profile image characteristics
are selected from the group consisting of barcodes, address styles,
address resolution, envelope size, initiating processing and
distribution center, addressee, postage characteristics, markings,
and handwriting characteristics.
6: A system enabling identification of mail pieces based upon
predetermined profiles, the system comprising: a detector and
optical elements configured to obtain an image of a mail piece; at
least one processor configured to obtain mail piece image
characteristics associated with the image of the mail piece; at
least one computer readable memory having: (a) a database storing
data for each one of a plurality of predetermined profiles, the
data comprising: a profile identifier, a plurality of profile image
characteristics and associated values, the values being chosen such
that the plurality of profile image characteristics is consistent
with threat mail, the profile image characteristics being
associated with various locations on the postage side of the mail
piece, and an action identifier configured to specify a plurality
of actions associated with the profile identified by the profile
identifier; and, (b) computer readable code embodied in the at
least one computer readable memory, the computer readable code
configured to cause the at least one processor to: retrieve the
data for one of the plurality of predetermined profiles from the
database, compare the mail piece image characteristics to the one
of the plurality of predetermined profiles from the retrieved data,
identify the mail piece as requiring one of the plurality of
actions identified by the action identifier from the retrieved
data, if the mail piece image characteristics present in the
retrieved data substantially match the one of the plurality of
predetermined profiles from the retrieved data.
7: The system of claim 6 wherein the computer readable code is
further configured to cause the at least one processor to: retrieve
further data corresponding to another predetermined profile from
the plurality of predetermined profiles from the database, compare
the mail piece image characteristics for the image of the mail
piece to the values for one of the plurality of predetermined
profiles from the retrieved further data, and identify the mail
piece as requiring one of the plurality of actions identified by
the action identifier from the retrieved further data, if the
values for one of the plurality of predetermined profiles from the
retrieved data substantially matches the mail piece image
characteristics for the image of the mail piece.
8: The system of claim 6 further comprising: a network interface
and a server configured to communicate with a network.
9: The system of claim 8 further comprising: a remote server
configured to receive and send the data.
10: The system of claim 7 wherein the plurality of predetermined
profiles include selected combinations of the plurality of profile
image characteristics selected from the group consisting of
barcodes, address styles, address resolution, envelope size,
initiating processing and distribution center, addressee, postage
characteristics, markings, and handwriting characteristics.
11: A system enabling identification of mail pieces based upon
predetermined profiles, the system comprising: a detector and
optical elements configured to obtain an image of a mail piece; at
least one processor configured to obtain mail piece image
characteristics associated with the image of the mail piece; a
first computer readable memory for storing data for access by a
process executed by at least one processor, said memory comprising:
a database storing data for each one of a plurality of
predetermined profiles, the data comprising: a profile identifier,
values for a plurality of selected profile image characteristics,
the values being chosen such that the plurality of selected profile
image characteristics is consistent with threat mail, the plurality
of selected profile image characteristics being associated with
various locations on the postage side of the mail piece, and an
action identifier configured to specify a plurality of actions
associated with the profile identified by the profile identifier;
at least one second computer readable memory having computer
readable code embodied therein, the computer readable code
configured to cause the at least one processor to: retrieve the
data for one of the plurality of predetermined profiles from the
database, compare the mail piece image characteristics to the one
of the plurality of predetermined profiles from the retrieved data,
and identify the mail piece as requiring one of the plurality of
actions identified by the action identifier from the retrieved
data, if the mail piece image characteristics present in the
retrieved data substantially match the one of the plurality of
predetermined profiles from the retrieved data.
12: The system of claim 11 wherein the computer readable code is
further configured to cause the at least one processor to: retrieve
further data corresponding to another predetermined profile from
the plurality of predetermined profiles from the database, compare
the mail piece image characteristics for the image of the mail
piece to the values for one of the plurality of predetermined
profiles from the retrieved further data, and identify the mail
piece as requiring one of the plurality of actions identified by
the action identifier from the retrieved further data, if the one
of the plurality of predetermined profiles from the retrieved data
substantially matches the mail piece image characteristics for the
image of the mail piece.
13: The system of claim 11 further comprising: a network interface
and a server configured to communicate with a network.
14: The system of claim 13 further comprising: a remote server
configured to receive and send the data.
15: The system of claim 11 wherein each of the profile image
characteristics is selected from the group consisting of barcodes,
address styles, address resolution, envelope size, initiating
processing and distribution center, addressee, postage
characteristics, markings, and handwriting characteristics.
16: The system of claim 14 wherein the remote server includes: at
least one third memory having computer readable code embodied
therein, the computer readable code configured to cause the at
least one processor to: retrieve the data for one of the plurality
of predetermined profiles from the database, compare the mail piece
image characteristics for the image of the mail piece to the values
for one of the plurality of predetermined profiles from the
retrieved data, and identify the mail piece as requiring one of the
plurality of actions identified by the action identifier from the
retrieved data, if the one of the plurality of predetermined
profiles from the retrieved data substantially matches the mail
piece image characteristics for the image of the mail piece.
17: The system of claim 16 wherein the computer readable code is
further configured to cause the at least one processor to: retrieve
further data corresponding to another predetermined profile from
the plurality of predetermined profiles from the database, compare
the mail piece image characteristics for the image of the mail
piece to the one of the plurality of predetermined profiles from
the retrieved further data, and identify the mail piece as
requiring one of the plurality of actions identified by the action
identifier from the retrieved further data, if the one of the
plurality of predetermined profiles from the retrieved data
substantially matches the mail piece image characteristics for the
image of the mail piece.
18: The system of claim 14 further comprising: another memory for
storing the data for access by a process executed by the remote
server, said memory including a database.
19: The method of claim 1 wherein said step of comparing can be
executed with a subset of the plurality of the predetermined
profiles simultaneously, and wherein said step of identifying the
mail piece for special processing can result from a match from any
one of the plurality of the predetermined profiles.
20: The system of claim 6 wherein said computer readable code is
further configured to compare the mail piece image characteristics
with a subset of the plurality of the predetermined profiles
simultaneously, and wherein said computer readable code is further
configured to identify the mail piece for special processing if
there is a match from any one of the plurality of the predetermined
profiles,
21: The system of claim 11 wherein said computer readable code is
further configured to compare the mail piece image characteristics
with a subset of the plurality of the predetermined profiles
simultaneously, and wherein said computer readable code is further
configured to identify the mail piece for special processing if
there is a match from any one of the plurality of the predetermined
profiles.
22: A system for processing mail items and identifying mail items
for special processing comprising: a transport sub-system
configured to transport a mail item; an imaging sub-system
configured to obtain an image of the mail piece; a database
configured to store a plurality of predetermined profiles, each of
the plurality of predetermined profiles having a plurality of
selected profile image characteristics, the database including a
combination of values for the selected profile image
characteristics, the values being chosen such that the combination
is consistent with threat mail, the selected profile image
characteristics being associated with various locations on the
postage side of the mail piece; a detector and optical elements
configured to obtain mail piece image characteristics associated
with an image of the mail piece; computer readable code embodied in
computer readable memory configured to retrieve the plurality of
predetermined profiles; computer readable code embodied in computer
readable memory configured to successively compare the mail piece
image characteristics to each of the plurality of predetermined
profiles until a match is found, if any; and computer readable code
embodied in computer readable memory configured to identify the
mail piece for special processing, if the one of the plurality
predetermined profiles substantially matches the mail piece image
characteristics for the image of the mail piece.
23: The system of claim 22 further comprising: computer readable
code configured to identify the mail piece as requiring one of a
plurality of actions identified by an action identifier from the
retrieved other of the predetermined profiles, if the one of the
plurality of predetermined profiles from the retrieved data
substantially matches the mail piece image characteristics for the
image of the mail piece.
24: The system of claim 22 wherein the profile image
characteristics are selected from the group consisting of barcodes,
address styles, address resolution, envelope size, initiating
processing and distribution center, addressee, postage
characteristics, markings, and handwriting characteristics.
25: The system of claim 22 further comprising: a network interface
and a server configured to communicate with a network.
26: The system of claim 25 farther comprising: a remote server
configured to receive and send the data.
27: The system of claim 22 wherein said computer readable code is
further configured to compare the mail piece image characteristics
with the combinations of values from more than one of the plurality
of the predetermined profiles simultaneously, and wherein said
computer readable code is further configured to identify the mail
piece for special processing if there is a match from any one of
the plurality of the predetermined profiles.
28: The system of claim 8 wherein said database is downloaded from
the network.
29: The method of claim 1 further comprising the steps of:
modifying a result record associated with the image of the mail
piece if the mail piece is flagged for the special processing;
transmitting the image to a remote site for manual image inspection
according to the result record; and classifying the mail piece
based on the manual image inspection.
30: The method of claim 1 further comprising the steps of: h)
archiving the mail image having archived mail piece image
characteristics; i) updating the plurality of predetermined
profiles by creating at least one new predetermined profile having
a plurality of selected new profile image characteristics; j)
retrieving one of the new predetermined profiles that has not yet
been used in the comparison; k) comparing the archived mail piece
image characteristics to the at least one new predetermined
profile; l) identifying the archived mail piece for the special
processing, if the new predetermined profile substantially matches
the archived mail piece image characteristics; and m) repeating
steps (i) through (l) for another one of the new predetermined
profiles, if the new predetermined profile does not substantially
matches the archived mail piece image characteristics.
Description
BACKGROUND OF THE INVENTION
[0001] This invention relates generally to the detection of
hazardous material and, more particularly, to the application of
image processing to early detection of potentially hazardous
material associated with mail collection or the collection of other
objects.
[0002] Recently there has been increased awareness of the potential
for large-scale introduction of hazardous materials, for example,
biological organisms, to create chaos or to harm an intended set of
victims. One potential delivery method that terrorists or other
criminals utilize to deliver such hazardous materials is through
the mail or other form of a delivery. In so doing, not only is
damage incurred by the intended victims, but also by any set of
potential victims that may be in a position of handling such
objects as the mail during the delivery or distribution
process.
[0003] There is currently technology available to law enforcement
organizations for detecting the presence of chemical and biological
threats. Such test materials generally are sensitive to specific
hazardous materials and are utilized by directly putting them in
contact with To date, however, there is a lack of early detection
of such hazardous material especially during the early phases of
mail handling or processing. Systems in place today do not deal
with detection prior to entering into the formal distribution
process. Thus, throughout the distribution process potential
non-intended victims are being subjected to hazardous material
carried by, for example, letter or package mail.
[0004] While complete inspection would be costly, early detection,
which allows suspected items to be culled from the processing
stream and be fully inspected, is desirable. There is a need for
automated technology to help identify mail pieces with
threat-consistent characteristics for early detection. Once the
candidate mail pieces have been identified as potentially harmful,
the analysis, inspection, and neutralization efforts could be
focused on a manageable subset of the mail.
SUMMARY OF THE INVENTION
[0005] The method and systems of this invention provide the
capability to help identify items based upon predetermined
characteristics. The method and systems of this invention provide
the capability to help identify mail pieces with threat-consistent
characteristics for early detection.
[0006] In an embodiment of the method of this invention, one or
more image characteristics for an image of an item being examined
are obtained. A predetermined profile is then retrieved from a
profile database. The one or more image characteristics for the
image of the item are compared to the one or more image
characteristics present in the predetermined profile. (The term
"image characteristics," as used herein below refers to one or more
image characteristics.) If the image characteristics present in the
predetermined profile substantially match the image characteristics
for the image of the item, the item is identified for special
processing. If the image characteristics present in the
predetermined profile do not substantially match the image
characteristics for the image of the item, another predetermined
profile is retrieved and compared to the image characteristics for
the image of the item.
[0007] In a specific embodiment of the method of this invention,
image characteristics for an image of a mail piece being examined
are obtained from a Remote Computer Reader (RCR). A threat profile
is then retrieved from a profiling configuration file (database).
The threat profile is then compared to the image characteristics in
order to determine whether the image characteristics match the
threat profile. If the threat profile matches the image
characteristics, the mail piece is identified for special
processing. Another threat profile is then retrieved and compared
to the image characteristics. If none of the threat profiles match
the image characteristics, normal mail processing continues. The
method described above can be executed in "Real Time" or can be
executed offline (also referred to as "Archival Processing").
[0008] An embodiment of a system of this invention includes a
transport sub-system and a computing sub-system. Transport
sub-system includes a transport, conveyor or pinch belt, a
radiation source, a detector and any optical elements necessary to
produce a scanned image of mail piece utilizing the detector.
Computing sub-system includes one or more processors, computer
readable media (one or more memories), a network interface and a
database. Computer readable media (one or more memories) has
computer readable code embodied therein which causes the one or
more processors to execute the method of this invention.
[0009] Another embodiment of a system of this invention includes a
transport sub-system, a computing sub-system, a network, a server
and a server database.
[0010] For a better understanding of the present invention,
together with other and further objects thereof, reference is made
to the accompanying drawings and detailed description and its scope
will be pointed out in the appended claims.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1 depicts a flow chart of an embodiment of the method
of this invention;
[0012] FIG. 2 depicts a graphical representation of a mail piece on
which this invention is practiced;
[0013] FIG. 3 depicts a schematic graphical representation of
another embodiment of this invention;
[0014] FIG. 4 depicts another graphical representation of a mail
piece on which this invention is practiced;
[0015] FIG. 5 depicts a flow chart of an embodiment of another
method of this invention;
[0016] FIG. 6 depicts a block diagram representation of an
embodiment of a system of this invention;
[0017] FIG. 7 depicts a block diagram representation of another
embodiment of a system of this invention.
DETAILED DESCRIPTION OF THE EMBODIMENTS
[0018] Method and systems that provide the capability to identify
items with predetermined characteristics are described herein
below. In the detailed embodiment described herein below, the
method and systems described provide the capability to detect, by
image analysis, combinations of characteristics consistent with
threat mail, to the extent that further inspection or processing is
warranted.
[0019] In order to facilitate the disclosure of detailed
embodiments of this invention, the following terms and acronyms are
defined herein below: [0020] facer-canceler--Mail processing
equipment that automatically faces letter-size mail in a uniform
orientation and cancels the postage stamps. (Also referred to as
canceling machine, facing identification mark, postmarking stamp,
roller-canceler, and tagging.) (Called Advanced Facer Canceler
System (AFCS) and Mark II Facer Canceler/Edger Feeder.) [0021]
multiline optical character reader (MLOCR)--An optical character
reader that reads and interprets more than one line of the delivery
address on a mail piece. [0022] Delivery Bar Code Sorter
Input/Output Subsystem (DIOSS), a universal modification kit that
will be installed on existing Delivery Bar Code Sorters (DBCSs) to
improve the capacity, speed and accuracy in which letter mail
processing equipment reads, processes and sorts mail. [0023]
PICS--ID Code Sorting system [0024] RBCS--Remote Bar Coding System
[0025] Postal Numeric Encoding Technique (POSTNET)--The barcode
system used on letter-size and flat-size mailpieces for encoding
the delivery point information and ZIP+4 code information. (Also
see delivery point barcode.) [0026] PLANET--PostaL AlphaNumeric
Encoding Technique [0027] facing identification mark (FIM)--A
series of five or six vertical bars used by automated postal
equipment to identify, orient, and separate reply mail. [0028]
REC--Remote Encoding Center
[0029] A flow chart of an embodiment of the method of this
invention is shown in FIG. 1. Referring to FIG. 1, image
characteristics 20 for a mail piece being examined are obtained
from an Image reading and Processing Center (2, FIG. 3; in one
embodiment, a Remote Computer Reader--RCR). The RCR (Remote
Computer Reader) 2 is a key site-centralized system that processes
letter mail images for the purpose of address resolution--reading
address text and deriving delivery point postal (ZIP) codes. (As a
central processing point for all images lifted by mail transport
and scanning systems, such as AFCS, MLOCR, and DIOSS transports,
RCR is a logical central source for archiving letter mail images,
enabling their later retrieval for investigative purposes.) A
threat profile 40 is then retrieved from a profiling configuration
file or database (step 45, FIG. 1). In one embodiment, the threat
profiles 40 are maintained in a Profiling Configuration File (5,
FIG. 3) stored in the RCR system 2. The Profiling Configuration
File (5, FIG. 3) can be created at and downloaded from an Image
Inspection System Server (17, FIG. 3). (The term "image
characteristics," as used herein below refers to one or more image
characteristics.) The threat profile 40 is then compared to the
image characteristics 20 (step 50, FIG. 1) in order to determine
whether the image characteristics match the threat profile (step
60, FIG. 1). If the threat profile 40 matches the image
characteristics 20, the mail piece is identified for special
processing (step 80, FIG. 1). ("Matching a threat profile" as used
herein below can include taking account the severity of the
profile. Thus, in some embodiments the image characteristics are
compared to one or more profiles.) If the threat profile 40 does
not match the image characteristics 20, another threat profile is
then retrieved and compared to the image characteristics. If all of
the threat profiles 40 do not match the image characteristics 20,
normal mail processing continues (step 70, FIG. 1).
[0030] The method described above can be executed in "Real Time" or
can be executed offline (also referred to as "Archival
Processing"). "Real Time" Mail Image Profiling describes the
capability to detect mail characteristics, determine that they fit
a pre-defined threat profile, and act upon that detection decision
as early as possible in the mail processing operation.
Specifically, Real Time Mail Image Profiling, in one embodiment,
accomplishes the profiling steps while the mail piece is undergoing
processing on image collection equipment (such as MLOCR, DIOSS, or
AFCS/DBCS), or, more broadly, at the originating processing
facility. "Archival" (also referred to herein as offline) Mail
Image Profiling refers to the process of detecting that the
characteristics of a mail piece fit a defined threat profile late
in, or after completion of, the mail processing cycle. A detailed
description of one embodiment is given herein below.
[0031] As shown in FIG. 2, the characteristics of a high profile
target, handwritten destination address, and lack of return address
are discernible by image processing. The method and system of this
invention enable the early detection by image analysis of
combinations of characteristics consistent with threat mail, to the
extent that further inspection or processing is warranted. There
are a large number of characteristics consistent with threat mail
that can be discerned from mail piece images. For many
characteristics, the bi-tonal image of the front (postage side) of
the envelope, received by remote image reading and processing
system 2 (RCR), is adequate. Other mail piece characteristics such
as specific stamp type identification would require grayscale
imagery.
[0032] The presence of a single mail piece characteristic is in
most cases insufficient to classify a mail piece as fitting a
threat profile. (However, in some embodiments, one characteristic
such as, but not limited to, a handwriting characteristic may
suffice.) Specification and detection of combinations of individual
characteristics, however, can be a powerful technique to isolate a
subset of mail that is consistent with a threat profile, and
therefore warrants time-consuming inspection or extensive special
processing. The methods of this invention enable profiles to be
defined and tested against each image processed by the remote image
reading and processing system 2 (RCR). An example profile for
handwritten mail without Return-Addresses could consist of the
following combination of characteristics:
Profile 1:
TABLE-US-00001 [0033] Initiating P&DC: TRENTON, NJ Destination
Address: US SENATE Destination-address style: HANDWRITTEN
Return-address style: NOT PRESENT POSTNET bar code: NOT PRESENT
Envelope length: <7 INCHES REJECT ENTRY CENTER (REC) SITE
REVIEW
[0034] It should be noted that not every characteristic need be
specified in the profile. For example, if a profile is silent
regarding the presence, location or content of a POSTNET bar code,
that characteristic can be ignored for that profile. In the method
of this invention, multiple profiles may be defined and
simultaneously active in the remote image reading and processing
system (RCR) 2, with the set of characteristics for each image
tested against each profile. When an image matches one or more of
the profiles, the image and/or the mail piece are identified for
special processing (step 80, FIG. 1) as discussed herein below.
[0035] In one embodiment, the profiles would be maintained in a
Profiling Configuration File (database) on the remote image reading
and processing (RCR) system 2, which is be created at and
downloaded from an Image Inspection System Server 17. Performing
Real Time Mail Image Profiling opens the possibility of taking
immediate action for mail detected as threat-consistent early in
the mail processing cycle. Several embodiments of actions that may
occur in reaction to detection of a mail piece fitting a threat
profile are described below. Actions range from simply logging the
image (retaining a copy of the image within a file for later
inspection) to affecting sorting before the mail piece leaves the
originating facility (placing the mail piece out of the processing
stream).
[0036] In one embodiment, the remote image reading and processing
(RCR) system 2 can perform "Front-end" and "Back-end" image
processing (see FIG. 3). For "Front-end" processing, the remote
image reading and processing system (RCR) 2 receives and analyzes
an image while the mail piece is in flight on a mail transport
(such as a MLOCR or DIOSS transport). The result message from the
remote image reading and processing system (RCR) 2 is normally used
to indicate the destination postal (ZIP) code result achieved by
address recognition processing of the remote image reading and
processing system (RCR) 2. In this embodiment, this result message
may be modified or augmented to indicate that a mail piece fits a
"Profile" as described earlier. This special result, which could
take the form of a reserved postal (ZIP) code, could be used to
sort the "Profiled" mail to special bins for subsequent special
handling or manual inspection (3, FIG. 3). This approach is
practical only for very accurate high-resolution profiles, such as
detection of a letter from a specific return address. The approach
is impractical for profiles based on characteristics that are
coarse or subject to inaccuracies, as the volume of mail out-sorted
for special handling may be too high. For profiles based on
characteristics that are coarse or subject to inaccuracies, a
preferred alternative involves routing the "Profiled" mail piece to
"Back-end" processing at the remote image reading and processing
system (RCR) 2, as discussed below.
[0037] In this embodiment, the "Back-end" processing at the remote
image reading and processing system (RCR) 2 occurs while mail is in
trays waiting for the next (in one embodiment, RBCS) processing
step. Normally, the result from "Back-end" processing is used to
determine whether video coding of an image must occur to determine
the destination result address. The video coding occurs at remote
(REC) sites. As in the case of "Front-end" processing, the result
record obtained from the remote image reading and processing system
(RCR) may be modified to indicate that a mail piece fits a
"Profile". For "Back-end" processing at the remote image reading
and processing system (RCR), two scenarios for handling "Profiled"
mail pieces may be implemented. Mail pieces fitting very accurate
high-resolution profiles (again, the example of detecting a
specific return address) could be flagged immediately as requiring
manual inspection or neutralization. The mail pieces matching very
accurate high-resolution profiles may be identified, through
barcodes or other means, so that the mail pieces can be out-sorted
for inspection (or neutralization) on its next transport pass (for
example, on an OSS or DIOSS transport).
[0038] For coarse profiles (e.g., mail pieces without a return
address), indication of a profile match at the remote image reading
and processing system (RCR) 2 may cause the image to go to a remote
(REC) site for manual image inspection (4, FIG. 3). Specifically, a
special "Security Desk" at another remote (REC) site 4 could be
created to screen images flagged by the remote image reading and
processing system (RCR) 2 as fitting a profile. The specially
trained individual screening the profiled images would have access
to up-to-the minute investigative information allowing the image to
be further classified as innocuous or suspicious enough to warrant
physical inspection or neutralization of the mail piece. Images
classified as innocuous would require no further address keying if
the remote image reading and processing system (RCR) 2 had resolved
the destination address.
[0039] "Back-end" processing is not subject to the stringent
latency constraints (approximately 3 seconds in one embodiment) of
"Front-end" processing. Mail piece characteristics requiring
significant computational time to detect may be detectable with
high accuracy only during "Back-end" processing. "Front-end"
detection of such characteristics may be limited to a coarse
determination that further processing is needed to screen the mail
piece, causing the mail piece to undergo "Back-end" processing.
[0040] Other possible embodiments can include neutralization as
part of the "Back-end" processing as well as more detailed
analysis.
[0041] "Archival Mail image profiling" can, in one embodiment, be
utilized in circumstances in which the detecting of a match to a
threat profile can occur late in the mail processing cycle. Some of
those circumstances include, but are not limited to, the following:
[0042] Detecting mail fitting a new threat profile. A new threat
profile would be created, for example, when an incident occurs of
delivery of a piece of mail containing a harmful chemical or
biological agent. Archival profiling is appropriate in this
situation to profile mail processed at the same facility (or
facilities) as the specimen mail piece, to identify similar mail
pieces. [0043] Mail characteristics whose detection requires
extensive computation time. Detection of some mail characteristics
may require extensive processing power and time, to the extent that
they are not normally active for Real Time Mail Image Profiling
(that is, in one embodiment, not normally active within RCR). A
possible example would be one or more Handwriting Matching
algorithms (it is possible, for example, that a coarse Handwriting
Matching algorithm could be active for Real Time Mail Image
Profiling, but that a more accurate but time-consuming algorithm is
executed later to reduce the number of false positive errors).
[0044] In an embodiment utilizing Archival Mail Image Profiling,
the profiling activity is decoupled from and possibly distant in
time from the mail delivery cycle.
[0045] The mail piece characteristics applicable to Archival Mail
Image Profiling can be a superset of those characteristics used for
Real Time Mail Image Profiling. That is, the characteristics can
include all of those applicable to Real Time profiling plus others
extracted on demand for purposes of Archival profiling. In another
embodiment, the Archival Mail Image Profiling occurs after the Real
time Image Profiling.
[0046] In that last embodiment, the processing could be delayed for
particular mail pieces while the Archival Mail Image Profiling
occurs.
Image-Detectable Mail Piece Characteristics
[0047] Table 1 lists several, but not limited to, mail piece
characteristics that can be reasonably detected, in one embodiment,
by image analysis, grouping them into phases for deployment based
on development time required to implement them. FIG. 4 depicts a
mail piece and illustrates some of the characteristics detected, in
this embodiment, by image analysis. A number of mail piece
characteristics are detected as by-products of address recognition
processing at the remote image reading and processing system (RCR),
so are available for virtually immediate deployment, including:
[0048] Destination address information: postal (ZIP) code, print
style (Machine, Hand) [0049] Barcode information: presence and
value of several barcodes of predetermined form (POSTNET and PLANET
in one embodiment) [0050] facing identification marks [0051] Return
Address postal code (ZIP) information [0052] Envelope size (See
FIG. 4 for a graphical representation illustrating some of the
characteristics of a mail piece.)
TABLE-US-00002 [0052] TABLE 1 Detectable Image Characteristics List
Detectable Image Charact ristics Destination address Resolution
Barcodes (FIM, PLANET, POSTNET in one embodiment) Return Address
Determination Envelope Size Destination/Return Address Style Clear
Area Infringement Destination Address Infringement Stain Postage
Characteristics Type Excessive Postage Restrictive Markings Return
address Resolution (including misspellings) Envelope
Characteristics Handwriting Similarity
Mail piece characteristics requiring more sophisticated algorithms
for detection are grouped into later phases for deployment.
[0053] Exemplary, but not limited to, image-detectable mail piece
characteristics, addressing degrees of effectiveness and accuracy,
development time, image type, and relative processing power
required, as described herein below.
[0054] Handwriting Matching, the capability to analyze handwriting
or hand-printing on mail images to determine similarity to that of
other (evidence) documents, is potentially a very valuable mail
piece characteristic for Mail Image Profiling, and is treated
separately herein below.
[0055] Several characteristics, but not limited to these, relevant
to classifying a mail piece as threat-consistent from analysis of
mail piece images are described below in terms of the differing
degrees of effectiveness, accuracy, image type, and processing
power required. [0056] 1)Barcodes (90, FIG. 5) and identifying
marks (85, FIG. 5) (in one embodiment, FIM, POSTNET & PLANET).
The remote image reading and processing system (RCR) 2 currently
detects the presence of the four different facing identification
mark 85 (FIM) types. This function is very accurate on most images,
unless the mark (FIM) is obscured by stamps or cancellation marks.
The barcodes 90 (POSTNET and PLANET) are effectively and accurately
detected and decoded at the remote image reading and processing
system (RCR) 2. The identifying marks 85 (FIM) and barcode 90
reading capabilities enable definition of mail profile
characteristics based not only on presence or absence of these
barcode types, but also on exact matches against specific values.
[0057] 2)Destination Address Style. In one embodiment, the remote
image reading and processing system (RCR) 2 detects multiple
candidate address blocks (groupings of text that appear to have the
form of a multi-line text address) on each letter image, and
classifies each block as to the probability that it is a return or
destination address. A byproduct of the optical character
recognition (OCR) process on an address block is a coarse
characterization of print style: [0058] Machine Printed
Styles--Sub-styles for Machine Print are Machine solid, Machine
broken, and Machine dot-matrix. [0059] Handwritten
Styles--Sub-styles for handwritten are hand-printed and cursive.
The categorization of sub-styles is more accurate for machine print
addresses than for handwritten addresses. [0060] Skew--A coarse
indication that pronounced skew of address text (where skew refers
to a pronounced deviation from horizontal, i.e., from parallelism
to the envelope bottom edge) exists may be obtained. [0061] 3)
Destination Address Resolution. In one embodiment, the remote image
reading and processing system (RCR) resolves the mail piece
destination address from the envelope image, determining the
finest-depth 11-digit postal (ZIP) code that corresponds to the
address text. The remote image reading and processing system 2
(RCR) address resolution is performed in conjunction with address
directory files created from previously obtained source data and
refreshed weekly at each the remote image reading and processing
system 2 (RCR) site, so the remote image reading and processing
system 2 (RCR) address information is always up to date. A mail
piece characteristic indicating that the destination address postal
(ZIP) code resolved by the remote image reading and processing
system (RCR) 2 matched any from a list of target postal codes
(ZIPS) can be immediately provided. The matching process may
support a wild card character (*) allowing a truncated comparison
(e.g., for a 5-digit ZIP match). [0062] 4) Return Address Style.
The classification of return address style at the remote image
reading and processing system 2 (RCR) is similar to that described
for destination address resolution. [0063] 5)Return Address
Resolution. Return address resolution (including handwritten return
address resolution) may support the following mail piece
characteristics: [0064] Postal Code (ZIP) Match--against a list of
postal codes (ZIP) in a configuration file, with wild-card support,
allowing truncated comparison (e.g., for a 5-digit ZIP match).
[0065] Invalid Return Address--indicating whether a match can be
found in a postal code (ZIP+4) address database. The text return
address block (as two separate characteristics). This indication is
coarse, as false indications of an invalid address may be asserted
due to incorrect OCR results. [0066] Invalid Delivery
Point--indicating that while the text return address is consistent
with the postal code (ZIP+4) address data, no match can be found in
a more detailed address database. The data in the more detailed
address database specifically identifies individual delivery points
(e.g., each house on a street) rather than just a range of
addresses as is contained in the postal code (ZIP+4) data. A
randomly chosen house number for a valid street has a reasonable
probability of being detected by this comparison. [0067] Non-local
Return Address--Two embodiments of methods for providing a reliable
indication as to whether the return address is not local (that is,
is inconsistent with the collection point) are described below.
These methods may be implemented at the remote image reading and
processing system (RCR). A return address anomaly, such as a mail
piece with a New Jersey return address that was evidently inserted
into the mail stream in Florida, could be identified by one of the
methods discussed below. [0068] In one embodiment the consistency
between the return address and the location of the mail processing
transport that first processed the mail piece is examined. For each
mail piece, the remote image reading and processing system (RCR)
receives ID Tag information containing an identification number
(ISS #, a number that uniquely identifies a specific MLOCR, DIOSS,
or AFCS transport, throughout the nation) of the transport that
first processed the mail piece. At the remote image reading and
processing system (RCR) 2, the ID TAG ISS# is compared to a static
table associating identification numbers (ISS#s) to the Sectional
Center Facilities (identified, in one embodiment, by the first
three digits of a postal (ZIP) code) served by the transports.
[0069] In a second embodiment, the consistency between the Return
Address and the postmark on the mail piece is examined. The ID TAG
comparison described above in the first embodiment utilizes
existing information. Obtaining the information contained in the
image of postmark on the mail piece requires additional image
processing. [0070] 6) Envelope Size & Skew. [0071] Envelope
Size--At the remote image reading and processing system (RCR) 2, a
"cropping" function detects the height of the mail piece in the
image and excludes any overscan areas. Envelope length is detected
directly by the number of scan lines contained in the image
(nominally 212 scan lines per inch). In this embodiment, the
envelope dimensions are used to characterize the envelope size
(e.g., business envelope, personal envelope) or to support specific
tests for envelope height and width. The camera scan height
(approximately 4.5'' for presently utilized cameras) limits the
maximum range of envelope height detection. [0072] Envelope
Skew--Envelopes containing powder or other bulky contents may have
a tendency to skew on the mail processing transports. In one
embodiment, the skew is detected utilizing the bi-tonal image of
the mail piece. In another embodiment, grayscale image processing
may be used to detect envelope skew. [0073] 7)Postage
characteristics. The following embodiments, but not limited to, can
be used to identify postage characteristics: [0074] Postage Type--A
postage type classifier, based on neural net technology, may be
integrated into the remote image reading and processing system
(RCR) to distinguish the following postage types: [0075] Stamp
[0076] Metered [0077] Pre-printed [0078] Embossed [0079]
Effectiveness and accuracy may initially be limited to a coarse
classification, with moderate accuracy, since the mail pieces whose
images are captured on present mail processing transports have
cancellation markings over the postage. Use of grayscale images
would allow higher accuracy of classification. [0080] Excessive
Postage--The following embodiments, but not limited to, can
distinguish instances of probable excessive postage. An assessment
of the extent and shape of the postage area may be obtained from
image processing of the binary (bi-tonal) images. Combined with the
neural net technology described above for determining Postage Type,
this method could assess the probability of presence of multiple
stamps. The result is a probabilistic indication of excessive
postage. A more precise determination of the postage amount on an
envelope requires processing grayscale images in order to recognize
in detail specific stamp images and their amounts. [0081]
8)Restrictive Markings--The possible presence of restrictive text
such as the words "Personal" or "Confidential" may be detected by
means of image character recognition. In some embodiments, the
image processing could be performed at the remote image reading and
processing system (RCR) 2. For handwritten marks, the detection
would require handwriting analysis and may be more appropriately
performed offline. For machine printed mail, optical character
recognition results, usually obtained at the remote image reading
and processing system (RCR) 2, enable searching machine printed
text for keywords. [0082] 9) Clear Zone Infringements--Several
embodiments, but not limited to, of methods for detecting Clear
Zone Infringements are disclosed below. Referring to FIGS. 2 and 5,
the bottom band of envelope images seen by the remote image reading
and processing system (RCR) 2 is normally clear, as this is the
area in which a (POSTNET) barcode 90 is printed after the remote
image reading and processing (and video coding if applicable) is
completed. Different infringements to the clear area could be
detected in both binary and gray images by the methods described
herein below. [0083] Destination Address Infringement--Referring to
FIG. 2, the letter 10 exhibits a handwritten address block that
infringes on the clear zone at the bottom of the envelope and on
the barcode 90. The coordinates of a bounding box for each
candidate address block found in the image are obtained by image
processing. Infringement of the address block on the clear zone can
also be detected by image analysis. [0084] Stain. Stains from
settling of chemical compounds may be visible as noise in the
bi-tonal image of the front of the envelope used for optical
character recognition (OCR) processing. The image of the mail item,
and in particular the bottom of the image (bottom of the envelope),
may be analyzed by means of image processing techniques in order to
classify the area as "Not Clear", possibly indicative of a stain.
More definitive analysis may be performed on gray images. In one
embodiment, the analysis may be performed at the remote image
reading and processing system (RCR) 2. Handwriting
Matching--Handwriting analysis may be utilized to compare
handwriting on images of mail to one or more reference images of
evidence documents. The result of this comparison can produce a
similarity score that could be thresholded to identify mail pieces
that may have been penned by the author of the evidence mail
piece.
TABLE-US-00003 [0084] TABLE 2 Image-Detectable Mail Piece
Characteristics Image Detection Characteristic Categories Type Rate
Accuracy 1) FIM, FIM presence Bi- High High POSTNET, & &
type Tonal PLANET Barcode type Barcodes Barcode field match
(against list) 2) Destination- Handwritten Bi- Moderate Moderate
Address (any form) Tonal Style Hand printed Bi- Moderate Moderate
Hand cursive Tonal Machine Bi- Moderate Moderate printed (any Tonal
form) Machine Bi- Moderate Moderate solid Tonal Machine broken
Machine dot- matrix 3) Destination Zip Match Bi- High High Address
(against a Tonal Resolution list of ZIPS) Mis-spelled Bi- Low
Moderate words (MP) Tonal 4) R turn- Not present Bi- Moderate
Moderate Address Tonal Style Handwritten Bi- Moderate Moderate (any
form) Tonal Hand printed Bi- Moderate Moderate Hand cursive Tonal
Machine Bi- Moderate Moderate printed (any Tonal form) Machine Bi-
Moderate Moderate solid Tonal Machine broken Machine dot- matrix 5)
Return Invalid Bi- Moderate Moderate Address address Tonal
Resolution (ZIP + 4 Directory) Invalid Bi- High High address (DPF
Tonal Directory) Non-local Bi- Low (HW) Moderate return Tonal
Moderate (HW) address (MP) High (MP) Bi- High High Tonal Postal
Code Bi- Low (HW)- High (Zip) Match Tonal 5 digits (against a
Moderate list of (MP) to ZIPS) High -11 digits 6) Envelope Envelope
Bi- High High Size & Skew Length Tonal Envelope Height Envelope
Bi- Moderate Moderate Skew Tonal 7) Postage Postage Type Gray
Moderate Moderate Characteris- (Stamp, tics Metered, Pre-printed,
Embossed) Excessive Bi- Moderate Moderate Postage Tonal Gray High
High 8) Restrictive Configurable Bi- Low (HW) Moderate Markings
list of Tonal Moderate (HW) keywords (MP) High ("Personal", (MP)
"Confidential", etc.) 9) Cl ar Area Destination Gray Moderate Low
Infringement Address Infringement Stain Bi- Low Low Tonal Gray
Moderate Moderate 10) Handwriting Similarity Matching score (degree
of similarity to image of evidence mail) Table Legend: Image Type:
Type of image required to support detection of the image
characteristic: Bi-tonal: Current bi-tonal images of the front of
the envelope are usable Gray: Grayscale images required Color:
Color images required Detection Rate: An estimate of the frequency
of false negatives - instances where a mail piece exhibits the
characteristic but the system fails to detect and indicate it: Low
UP to 25% false negatives Moderate Up to 15% false negatives High
Up to 5% false negatives Very High Approximately 1%-2% false
negatives Accuracy: An estimate of the frequency of false positives
-- instances where the system incorrectly indicates that the
characteristic is exhibited by the mail piece Low: Up to 25% false
positives Moderate: Up to 5% false positives High: Approximately
1%-2% false positives
Handwriting Matching
[0085] The term "Handwriting Matching" is used here to describe a
capability to compare characteristics of handwriting or
hand-printing from mail piece images to characteristics obtained
from evidence documents, to determine (with a quantified confidence
level) whether they were penned by the same author.
[0086] Handwriting Matching is a potentially valuable technology to
apply within the previously described Mail Image Profiling
framework, with possible use for both Real time and Archival
profiling. The confidence output from a Handwriting Matching
algorithm can be assigned as a mail piece characteristic, and
thresholded to determine if the mail piece meets the requirements
of a specific profile.
[0087] While handwriting recognition (as performed in support of
address recognition in the remote image reading and processing
system (RCR) system 2) has different goals than handwriting
matching (determining whether documents come from the same author),
many of the underlying technical principles and image analysis
techniques apply to both. Examples of technology available in
address recognition and Handwriting Matching technology are
described herinbelow.
[0088] Numerous handwriting features and attributes have been
explored for applicability to identification and profiling through
sampling (see, for example, S. N. Srihari and S. Lee, Automatic
Handwriting recognition and Writer Matching on Anthrax-related
Handwritten Mail, Proc. of the 8th Int'l Workshop on Frontiers in
Handwriting Recognition, 2002). Computer enabled methods, which are
able to locate character strings within handwritten documents, have
been developed. These methods could be modified to assist in
handwriting matching (see, for example, Recognizing the Truth About
Character Recognition, Jim Terry, Stacy Kamigaki, Tatyana Vazulina,
May 19, 2000, available at
http://www.parascript.com/products/whitepapers.cfm).
[0089] FIG. 5 indicates an embodiment in which the Handwriting
Matching software is utilized to pre-process samples of a person's
handwriting to extract distinguishing characteristics, then is
utilized to identify other mail pieces penned by the same
individual. Referring to FIG. 5, a handwriting sample 300 from one
individual is processed by means of a handwriting profiling method
310 and the distinguishing handwriting characteristics for that
individual 320 are included in the profile. A hand addressed mail
piece is imaged and the image 305 and the handwriting
characteristics are obtained for that image 315. The handwriting
characteristics are obtained for that image 315 are compared to the
distinguishing handwriting characteristics for that individual 320
and mail piece is identified for special processing 350, if the
distinguishing handwriting characteristics for that individual 320
substantially match the handwriting characteristics are obtained
for that image 315.
System
[0090] FIG. 6 depicts a block diagram representation of an
embodiment of a system of this invention. Referring to FIG. 6,
system 160 includes a transport sub-system 140 (such as in an AFCS,
MLOCR, or DIOSS transport) and a computing sub-system 120. The
system 160 includes both the transport and the remote image reading
and processing system (RCR) 2. Transport sub-system 140 includes a
transport, conveyor or pinch belt 150, a radiation source 100, a
detector 110 and any optical elements necessary to produce a
scanned image of mail piece 10 utilizing detector 110. The detector
110 (including any optical elements necessary to produce a scanned
image of mail piece 10) constitute means for obtaining an image of
the mail piece (item) 10. The detector 110 may be, but is not
limited to, a digital acquisition device, a video camera, or a
detector such as a CMOS or CCD image detector. Computing sub-system
120 includes one or more processors 210, computer readable media
(one or more memories) 220, network interface 230 and database 130.
The computer readable media (one or more memories) 220 has computer
readable code embodied therein which causes the one or more
processors 210 to execute methods that process images of the mail
piece 10 for the purpose of address resolution, reading address
text and deriving delivery point ZIP codes, obtain image
characteristics for the mail piece 10, retrieve a threat
(predetermined) profile, compare the image characteristics to the
threat profile, if the image characteristics match the threat
profile, identify the mail piece (and image) for special
processing. The computer readable code that causes the one or more
processors 210 to execute methods that process images of the mail
piece 10 for the purpose of address resolution, reading address
text and deriving delivery point ZIP codes, obtain image
characteristics for the mail piece 10 constitute means for
obtaining image characteristics for the image of the mail piece 10
(item). The computer readable code also causes the one or more
processors 210 to, if the image characteristics do not match the
threat profile, retrieve another profile and repeat the comparison
step, and, if the image characteristics do not match any one threat
profile, continue normal mail processing. The threat profile can
reside in the computer readable media (one or more memories) 220 or
in database 130. Also, scanned images of mail pieces and their
corresponding image characteristics can be stored in the computer
readable media (one or more memories) 220 or in database 130.
Threat profiles can be obtained via a network 240 through a network
interface 230 from a server 260, where, in one embodiment, the
profiles are stored, (FIG. 7).
[0091] In one embodiment, the database 130 stores data for the
threat profiles after receiving the profiles via the network
240.(In another embodiment, the profiles reside at the database
130.) The data corresponding to one profile includes an identifier
for the profile, data for image characteristics present in the
profile, and an action identifier.
[0092] FIG. 7 depicts a block diagram representation of another
embodiment of a system of this invention. Referring to FIG. 7,
system 290 includes transport sub-system 140, computing sub-system
120, network 140, server 260 and server database 280. Transport
sub-system 140 and computing sub-system 120 operate as described
above. Server 260 and server database 280 can be utilized to store
and provide threat profiles, as also described above. Server 260
and server database 280 can also be utilized in "offline" mail
image profiling.
[0093] In one embodiment of "offline" mail image profiling, the
comparison of the image characteristics to the threat profile
requires extensive computations, as in the case of handwriting
analysis. Server 260 includes one or more processors (not shown),
computer readable media such as one or more memories (also not
shown). The computer readable media (one or more memories) in sever
260 has computer readable code embodied therein which causes the
one or more processors in server 260 to execute methods that obtain
image characteristics for a scanned image of a mail piece, retrieve
a threat profile, perform extensive calculations needed to compare
the image characteristics to the threat profile, compare the image
characteristics to the threat profile, and, if the image
characteristics match the threat profile, identify the image for
special processing. The computer readable code, in server 260, also
causes the one or more processors in server 260 to, if the image
characteristics do not match the threat profile, retrieve another
profile and repeat the comparison step. The threat profiles are, in
one embodiment, stored in the server database 280. In another
embodiment, the threat profiles are stored in the database 130. The
scanned image of the mail piece can reside in the server database
280 or in database 130. If the scanned image of the mail piece
resides in database 130, it can be provided to server 260 via
network 240.
[0094] It should be noted that, in one embodiment, a subset of the
methods for comparing the image characteristics to the threat
profile can be performed utilizing the one or more processors 210
in computing sub-system 120. In that embodiment, the remainder of
the subset including the extensive computations is performed
utilizing the one or more processors in server 260. Server 260
provides an execution platform for performing image analysis to
extract mail piece characteristics not normally provided by Real
Time Mail Image Profiling (i.e., not provided by the remote image
reading and processing system (RCR)). An example, again, could be a
high-precision Handwriting Matching algorithm.
[0095] In general, the techniques described above may be
implemented, for example, in hardware, software, firmware, or any
combination thereof. The techniques described above may be
implemented in one or more computer programs executing on a
programmable computer including a processor, a storage medium
readable by the processor (including, for example, volatile and
non-volatile memory and/or storage elements), at least one input
device, and at least one output device. Program code may be applied
to data entered using the input device to perform the functions
described and to generate output information. The output
information may be applied to one or more output devices.
[0096] It should be noted that, the term "mail piece" includes any
item, envelope or package being delivered a package delivery
service.
[0097] Elements and components described herein may be further
divided into additional components or joined together to form fewer
components for performing the same functions.
[0098] Each computer program may be implemented in any programming
language, such as assembly language, machine language, a high-level
procedural programming language, or an object-oriented programming
language. The programming language may be a compiled or interpreted
programming language.
[0099] Each computer program may be implemented in a computer
program product tangibly embodied in a computer-readable storage
device for execution by a computer processor. Method steps of the
invention may be performed by a computer processor executing a
program tangibly embodied on a computer-readable medium to perform
functions of the invention by operating on input and generating
output.
[0100] Common forms of computer-readable (computer usable) media
include, for example, a floppy disk, a flexible disk, hard disk,
magnetic tape, or any other magnetic medium, a CDROM, any other
optical medium, punched cards, paper tape, any other physical
medium with patterns of holes, a RAM, a PROM, and EPROM, a
FLASH-EPROM, any other memory chip or cartridge, a carrier wave,
such as electromagnetic radiation or electrical signals, or any
other medium from which a computer can read.
[0101] Although the invention has been described with respect to
various embodiments, it should be realized that this invention is
also capable of a wide variety of further and other embodiments all
within the spirit and scope of the appended claims.
* * * * *
References