echnJLca1 Fiel ,

This invention relates, to a banking system and method for the processing of data-carrying documen-ts.

Background Art

Bank accounting systems track debit and credit entries for the bank and its customers. Entries and transactions are the physical entities (documents) of the System which show what action and what needs are required. Over the years, procedures evolved for the processing of entries and transactions which improved the routing of information and the transfer of funds.

Prior art bank accounting systems are general- ly organized into three general areas:

1. The point of acceptance, where the transaction physically enters the accounting system for processing. Included are those transactions accepted from depositors and those entries processed through other banks. 2. The point of processing, where transactions from the point of acceptance are processed.

3. The point of payment, where the account re¬ sides. At the point of payment, entries are reviewed, verified, amended, and distributed. An accounting entry is the basic document which substantiates a transfer of funds in any book¬ keeping system, either manual, mechanical, or electronic. An entry in banking procedures is a single document con¬ taining information about the account and giving the bank an instruction. The entry may either "take from" or "add to" the account. The physical document convey¬ ing the entry is important only as far as it preserves the information and the instructions of the entry.

The information associated with an entry is composed of five levels:

- -

1. Information of value. This information tells the amount of money involved in the transfer.

. v r>_

2. Information of control. This information gives the account number and the date of the transfer ^" . 5 3. Information of instruction. This information tells what is to be done and indicates any special con¬ ditions which must be met.

4. Information of authorization. This informa¬ tion, composed of signatures or initials, gives the

10 account holder's authorization to transfer funds.

Authorization has legal weight should the transfer be challenged.

5. Information of record. This information, ^" con¬ sisting of teller stamps, endorsings, proof machine ^"*" " ^'

15 embossings, gives the bank the ability to trace the ^' physical documents through the system, i.e., who brought in the document; what teller handled it; what day was the transaction made; and which machines processed it? Information of value, control, instruction,

20 and authorization forms the essence of the accounting entry, while information of record forms the means of controlling that entry. The document comprising the accounting entry (a check, deposit, withdrawal form, etc.) is important as a vehicle for these five levels

25 of information.

An accounting transaction consists of two or more balancing entries that direct the bank to move funds from one account to another. The customer directs the flow of funds from a source to a need by presenting

30 a transaction to the bank. The transaction indicates ^" nr the amount to draw from the source, and the amount to apply to the need. Because this is a transfer of funds, the amount removed from the source must always equal the amount applied to the need.

35 Most of the problems which occur in such banking systems stem from controlling the document or paper. That is, for the information to be available.

O

the document or paper must be available. Consequently, all systems which are dependent on documents or paper to carry information suffer from the problems associated with moving and controlling the documents or paper. Some of the problems encountered in such bank¬ ing systems are alleviated by a banking system proposed in the article "The message...without the medium" pub¬ lished in "Banking, Journal of American Bankers Associ¬ ation", October 1977. In this system, an electronic image is generated of all information carried on a document presented at the point of acceptance and each document and its associated image is provided with identification information. The obtained image infor¬ mation is then forwarded to the processing center where it is translated into machine useable information ^■ and is further processed without using the documents themselves.

An advantage of this system is that the move¬ ment of paper between banks to support financial entry processing is greatly reduced and information about a financial entry may be given to a customer without having to physically locate and retrieve the original documents.

Disclosure of Invention The present invention represents a further improvement in the banking system of the above kind in which transaction balancing is performed on the machine- useable data developed from the image information to produce qualified data for a group of documents when in balance. The sorting of a group of documents according to out-clearing destinations is then prepared by using the associated qualified data therefor. Moreover, means are provided which interface the qualified data for said group with conventional application systems. ^An advantage of the banking system of the invention is that the documents are sorted using the

_ _

associated qualified data instead of having to rely on non-controlled information contained on the document itself. Further advantages are:

1. Without proofing operations at the point of 5 acceptance, the scheduling of work and peak-load staf¬ fing can be stabilized at said point in addition to lowering the level of skill of the operators.

2. The level of skill of operators at the pro¬ cessing center can also be reduced as balancing and con-

10 solidations are completed automatically.

3. Key entry operators are not required to handle documents or make proof or sort decisions. Op¬ timum key efficiency is thus realized from the key entry operators.

15 Thus, according to the invention, there is provided a banking system for processing data-carrying documents, including a point of acceptance having an apparatus for generating an electronic image of the data carried on each document presented at said point

20 of acceptance and for assigning identification informa¬ tion to each document and to its associated image; a processing center operatively connected to said point of acceptance for developing machine useable data from the obtained image ^" information and for performing pro-

25 cessing operations with respect to such data; and a point of payment operatively connected to said proces¬ sing center, characterized in that said processing center includes image storage means for storing said image information for a group of documents; data develop-

30 ment means for receiving said image information for ' said group of documents from said image storage means to develop said machine useable data therefrom, and in¬ cluding computer means for performing transaction balancing on said machine useable data to produce

35 qualified data therefrom for said group of documents when in balance; data storage means for storing said qualified data for said group of documents; sorting

means for sorting said group of documents according to out-clearing destinations utilizing the associated qualified data therefor; and means for interfacing said qualified data for said group of documents with conven- tional application systems for the preparation of bank¬ ing reports and statements and for reproducing on an archival file said qualified ^' data and said image infor¬ mation for those documents for which account responsi¬ bility rests with said point of payment. According to another aspect of the invention, there is provided a method of processing data-carrying documents in a banking system including a point of acceptance, a processing center and a point of payment, in which an electronic image is generated of the data carried on each document presented at said point of acceptance and identification information is assigned to each document and to its associated image, and in which machine useable data are developed from the ob¬ tained, image information and processing operations are performed with respect to such data, characterized by the steps of storing said image information for a group of documents in an image file located in said processing center; withdrawing said image information for said group of documents from said image file to develop said machine useable data therefrom; performing transaction balancing for said group of documents while utilizing said machine useable data and said identification in¬ formation therefor, to produce qualified data for said group of documents when in balance; storing said quali- fied data in a data storage means located at said pro¬ cessing center; sorting said group of documents according to out-clearing destinations utilizing the associated qualified data therefor; interfacing said qualified data for said group of documents with application sys- terns for the preparation of banking reports and state¬ ments; and producing an archival file of said qualified ^d ata an ^d said image information for said group of _d ocu-

-6-

ments for use by said point of payment in the preparation of reports and copies of said documents.

Brief Description of the Drawing

One embodiment of the invention will now be 5 described, by way of example, with reference to the accompanying drawings, of which:

Fig. 1 is a general schematic diagram, in block form, showing an embodiment of this invention which includes a point of acceptance, an image proces- 10 sing center, and a point of payment and their associated elements;

Fig. 2 is a general flowchart representing general functions which are performed at a point of acceptance with regard to financial documents presented 15 thereat;

Figs. 3A and 3B taken together show a general schematic diagram, in block form, of an image lift unit shown in Fig. 1;

Fig. 4 is a general flowchart indicating how 20 data is developed from images forwarded to the image processing center from a point of acceptance;

Figs. 5A through 5F taken together show the various elements included at an image processing center; Fig. 6 is a flowchart showing, mainly, error 25 correction procedures at a point of acceptance for bringing a transaction group of documents into balance; Fig. 7 is a flowchart showing the movement of documents at an image processing center for a qualified document in a transaction group; 30 Fig. 8 is a flowchart showing the creation of reports for the associated point of payment;

Fig. 9 is an organizational chart showing the flexibility of the preferred embodiment;

Fig. 10 is a chart showing the general func- 35 tions performed by the data development means shown in Fig. 1; and

OMP

Fig. 11 is a chart showing the layout of Figs. 5A through 5F which comprise the image processing center.

Best Mode of Carrying Out the Invention Fig. 1 is a general schematic diagram, in block form, showing a preferred embodiment of this invention as used in a banking environment, with the preferred embodiment being referred to as the banking system 10. The banking system 10 includes, typically, a plurality of: (1) points of acceptance, (2) image processing centers, and (3) points of payment; however, for ease of illustration, only one point of acceptance 12 (hereinafter referred to as POA), one image processing center 14 (hereinafter referred to as IPC) , and one point of payment 16 (hereinafter referred to as POP) are shown in Fig. 1.

The function of the POA 12 is to capture (in image form) all the information from documents (checks, deposits, etc.) presented thereat and to prepare and transmit this information to the associated IPC 14. The POA 12 is also responsible for correcting error transactions.

The IPC 14 provides services for several POAs 12 and POPs 16 and is geographically located to provide the services thereto. Basically, the function of the IPC 14 is to extract data from the information received from the associated POAs 12 and put it into a form which can be used for electronic processing for account post¬ ing, preparation of reports, and the like. At a certain stage in the processing of the information for a group of transactions at the IPC 14, the associated documents for this group are physically transported from the POA 12 to the IPC 14 where the documents are physically sorted according to destinations determined while data was developed at the IPC 14. Checks drawn on banks outside of the banking system 10 are qualified for shipment to these outside banks.

OMPI

The POP 16 is that part of the system 10 which has the ultimate responsibility for particular customers' accounts. After processing at the IPC 14, a permanent record medium, (as for example, a video disc) containing the information about accounts for a particu¬ lar POP 16 is physically shipped to that POP 16 for information support of the customers' accounts retained at that POP 16.

The functions performed by a POA 12 and a POP 16 are such that both functions are typically performed at a branch bank of the banking system 10, and may also be performed at an IPC 14 as will be described later herein.

In explaining the general functioning of the banking system 10, it is appropriate to start at the POA 12 (Fig. 1). At a POA 12, a teller receives documents 18 (Fig. 2) such as checks and deposits representing financial transactions. Banking transactions are pro¬ cessed in transaction groups, with each group having approximately 150 to 200 documents in transaction order.

A certain amount of preparation 20 (Fig. 2) is performed on the documents 18 to prepare them for entry into an image lift unit 22 (hereinafter called ILϋ) shown in Fig. 1. The preparation 20 includes scrutinizing the documents 18 to eliminate foreign objects and folds and to align them to face in the same direction for entry into the ILϋ 22.

The ILU 22 is shown as a schematic diagram, in block form, in Figs. 3A and 3B. The ILϋ 22 is basically similar, in physical form, to a conventional card reader; therefore, the actual physical construction of the ILϋ 22 need not be shown.

The ILU 22 includes a conventional input hopper 24 (Fig. 3A) for receiving the documents 18, a conventional transport mechanism 26 for moving the documents along a transport path 28, and an output

OM

hopper 30 for receiving the group of documents 18 after processing by the ILϋ 22.

A transaction group control document 32 (Figs. 2 and 3A) is placed in the input hopper 24 along with the transaction documents 18 in order to identify the particular group of documents which is to be pro¬ cessed. The transport mechanism 26 sequentially moves each one of the documents 32, 18 along the transport path 28 past a conventional imaging device, designated generally as 34. The imaging device 34 is the means for generating an image of each of the documents passing thereby.

The imaging device 34 (Fig. 3A) includes a light source 36 for illuminating the front of a document 18 and an imager 38 for generating an image of or

"imaging" the front while a similar light source 40 and imager 42 are utilized for simultaneously illuminating and imaging the back side of the same document 18. In the embodiment described, imaging of the front of the document 18 is effected by scanning from the upper right hand corner (as viewed in Fig. 3A) to the lower right hand corner, and then proceeding from the right edge of the document to the left edge thereof until the entire front thereof is scanned, although other scanning pat- terns may be employed. Accordingly, the back of the document 18 is simultaneously scanned from the left edge thereof (as viewed from the imager 42 in Fig. 3A) to the right edge thereof until the entire back thereof is scanned. Each of the imagers 38, 42 shown in Fig. 3A may, for example, be a conventional imaging device such as a CCD imager which is manufactured by Fairchild Inc. or a photodiode array imager which is manufactured by Reticon. These imagers 38 and 42 produce a high reso- lution digital representation or "image" of each of the documents 18 which allows a representation or image of the document to be recreated in a computer memory, shown

on video displays or reproduced on a record medium via graphic printers. The image capture step is shown by block 44 in Fig. 2. The outputs 46 and 48 of the imagers 38 and 42, respectively, are fed into an image enhancement means 50 (Fig. 3B).

The image enhancement means 50 (Fig. 3B) may be any conventional enhancement means which eliminates certain unwanted background data while intensifying meaningful information. At the present time, some of the documents 18 which are scanned by the imaging device 34 contain scenic backgrounds, such as mountains, shore¬ lines, etc., which do not contain meaningful information and can be eliminated as these backgrounds, at times, interfere with the electronic processing of data. The image enhancement means 50 contains a processor such as a microprocessor 52 and conventional algorithms 54 which may be implemented by software or logic in hardware to eliminate the unwanted background data. Essentially, the image enhancement means 50 reduces the gray scale resolution of a digitized image of a document 18 by suppressing the image's low frequency spatial components which are those components changing slowly from light to dark and from dark to light (representing faint back¬ ground images) . The high frequency spatial components of the digitized image of a document 18 are those com¬ ponents which change rapidly from light to dark and from dark to light (compared to the low frequency components) and represent the desired information such as account numbers, dates, amounts, signatures, and the like. The low frequency spatial components which are suppressed are replaced by "white" components with the desired data being replaced by "black" components, thereby reducing the number of gray levels which in turn reduces the number of bits necessary to represent the documents 18.

After the digital representation or image of a document is enhanced by the image enhancement means 50,

O ^W

the enhanced image is routed to a conventional image compression means 56 (Fig. 3B) which includes a processor such as a microprocessor 58 and the necessary algorithms 60 to compress the image received. The purpose of image enhancement means 50 and image compression means 56 is to reduce the image bit representation to the lowest possible number while retaining all pertinent information to lower transmission time, and to reduce computer memory, and disk storage requirements. The input hopper 24, transport mechanism 26, imaging device 34, image enhancement means 50 and the image compression means 56 are controlled by a conven¬ tional control means 62 (Fig. 3B) which includes, for example, a processor 64 such as a microprocessor and its associated program 66.

As the document 18, shown at the imaging device 34 in Fig. 3A, is moved further along the trans¬ port path 28, it passes a reader 68 which reads a pre¬ defined area of one side of the document to sense for a Document Identification number (DI#) which may have been previously assigned to this particular document. For the moment, assume that no previously assigned DI# is present on the document 18 which is further moved along the transport path 28 to a printer 70 (Fig. 3A) . The printer 70 is a conventional printer (under the control of the control means 62) which prints a unique DI# on the particular document 18 passing thereby. The unique DI# is printed in conventional coding (such as height modulated or bar-no bar coding) on one side of the document 18 as for example the back side. The printer 70 may be, for example, of the ink-jet type.

After the document passes the printer 70 (Fig. 3A), it is moved further along the transport path 28 to a reader 72 which reads the DI# which has just been printed or was previously printed thereon by the printer #70. The DI# which is in digital form is then combined with the compressed image and a unique tracing,

^ zTtj^.

location, and retrieval number (TLR#) for interval control of the document 18 to form a collection or package of image information which is called an entry record 74 (Fig. 3B) which is transmitted to the IPC 14 by means to be later described herein.

The image enhancement and compression steps for the image enhancement and compression means (50, 56) are shown as step 76 in Fig. 2, and the addition of the DI# is shown as step 78. The contents of the entry record 74 are shown only diagrammatically in Fig. 3B; they include the digitized image of both the front and rear of the document 18, the TLR# shown as 82, and the DI# shown as 80. The TLR# is a unique number in digital form which represents the particular bank and branch, the particular ILϋ 22 which processed the document 18, and the date and a sequence number assigned to each of the documents 18. Basically, the TLR# is used to trace a document back to the particular entry point or POA 12 an _* d is also used as a retrieval number for both the document 18 and the associated digitized image. Essentially, the DI# is used by automated document handling equipment to facilitate distribution of the documents 18.

If a document 18 is entering the banking system 10 for the first time, the DI# and the TLR# will be the same. In some situations, it is necessary to reprocess certain documents which have been through the banking system 10 at a prior time. For example, the same check which was initially refused at the bank for "insufficient funds" may be again used by the receiver of the check by re-depositing the check after the drawer of the check has rectified the insufficient-funds status of his account. In this situation, the prior-assigned DI# for the document 18 is sensed or read by the reader '68 (Fig. 3A) and is added to the entry record 74. In summary, only one DI# is ever encoded on a document. The transaction group number (Trans. Gr.#) is entered

• O

A. wi

into the banking system 10 via the transaction group control document 32 (Fig. 3A) and this number represents the particular transaction group or "batch" of 150-200 documents. The Trans. Gr.# may be read by the reader 72 (Fig. 3A) and transferred to the associated entry record 74 (shown as 86 in Fig. 3B) via the control means 62. The step of adding the Trans.. Gr.# is represented by block 87 in Fig. 2. After the transaction group control document 32 and the associated documents 18 have been processed at the ILϋ 22, they are deposited in the output hopper 30. All these documents (32, 18) are then removed from the hopper 30 and are kept together as a batch at the POA 12 until they are later physically transported to the IPC 14 at a certain point in the processing of the associated data at the IPC 14 as will be described later herein. The entry records 74 (Fig. 3B) formed by the image information for the documents 18 including the associated identification information are transmitted to the IPC 14 via a communication means 88 (Fig. 1) located at the IPC 14 and a high speed digital communications network or in-house line 90 (Fig. 1) which connects the POA 12 with the IPC 14.

To enable operators at the POA 12 (Fig. 1) to communicate with the associated IPC 14, a conventional keyboard 92, an optionally-provided display 94 (such as a CRT display) and terminal printer 96 are used. The keyboard 92, display 94 and printer 96 are coupled through a conventional controller 98 to the ILϋ 22 and to the communication means 88 at the IPC 14 via the line 90 to conventionally provide communication therebetween. Fig. 1 contains the elements of the IPC 14 in block form, with the content of these elements being described in more detail later herein. It will be recalled that the IPC 14 performs all data processing and document handling functions within the banking system 10. As the image information of the documents 18 included in a transaction group or batch are received in

OMPI

/., WIPO

he form of entry records 74 (Fig. 3B) by the communi¬ cation means 88, they are routed to the image file means 100 via a system bus 102 which may be any conventional high-speed bit serial bus. All image information and data coming into or going out of the IPC 14 are con¬ trolled by the communication means 88, which performs all handshake protocol, logical addressing and communi¬ cations packaging, and which directs all incoming images and data to the appropriate file means, as for example, image file means 100. The image file means 100 is processor controlled and broadly includes a primary storage 104 which represents, for example, a plurality of high-capacity magnetic discs and a back-up storage or archival file system, shown, for example, as a video disc 106.

^• A system manager 108 at the IPC 14 (Fig. 1) ^• provides common support functions such as operator consoles 110 (only one being shown), line printers (not shown) , program libraries, and non-volatile storage and retrieval of system information needed by other subsys¬ tems. The system manager 108 also provides the operator interface to all subsystems of the banking system 10, and conventionally provides the control of initiation, termination and re-start processes. This aspect will be discussed in more detail hereinafter.

The IPC 14 (Fig. 1) also includes a data development means 112 which essentially converts the images of a transaction group of documents 18 into machine useable data, e.g. data for machine processing and transaction proofing. Some of the functions per¬ formed by the data development means 112 include: (a) data scheduling, which insures that the correct images from the image file means 100 are extracted for proces¬ sing and which utilizes computer pattern recognition of information in the image; (b) data completion, which provides a human interface to complete the image-to-data conversion where necessary; (c) data qualification,

O

which employs verification routines to prove the cor¬ rectness of the developed data and to apply document routing information thereto; (d) batch consolidation, which arranges the data as developed at the data develop- ment means 112 along with the information comprising the entry record 74 without the image information (herein¬ after referred to as data records 74-D) into input transaction groups for batch proving; and (e) batch analysis and proving which verifies that debits equal credits for those documents within a transaction group or batch.

Once the batch analysis and proving is effected by the data development means 112 (Fig. 1), the data records 74-D for a proved transaction group are con- sidered qualified data and are then stored in a par¬ ticular location within the data file means 114. The data file means 114 is processor controlled and broadly includes a primary storage 116 which represents, for example, a plurality of high-capacity magnetic discs and magnetic tape units, and an optionally-provided back-up storage or archival file system, shown for example, as a video disc 118. With both the image file means 100 and the data file means 114, information therefrom is acces¬ sed using the TLR# of a specific document 18 or the transaction group number 86 associated with a particular batch of documents.

When a particular batch or transaction group of documents 18 has been proved or qualified (indicating that the debits equal the credits for that group) and the associated data records 74-D are placed in the data file means 114, a release from the IPC 14 is made. The associated POA 12, upon receiving this release, then physically ships the documents 18 associated with that transaction group to the IPC 14 for physical sorting of the documents 18 by the sorting means 120 (Fig. 1).

During the process of developing the data for a trans¬ action group of documents 18 at the data development

means 112, certain sorting destination routing informa¬ tion indicative, for example, of checks drawn on banks other than those associated with the banking system 10 and of checks drawn on banks within the system 10 is added to the data which is placed in the data file means 114 for that transaction group of documents 18 so as to facilitate the physical sorting of the documents 18 at the sorting means 120.

The first sort at the sorting means 120 (Fig. 1) is made to separate the "transit" documents (those drawn on other banks) from the "on-us" documents (those drawn on banks included in the banking system 10). The data records 74-D for a transaction group are extracted from the data file means 114 and utilized by the sorting means 120 to perform the sorting. In this regard, the DI# on a document 18 is read by the sorting means 120 and matched with the DI# contained within the associated qualified data record 74-D to retrieve the associated routing information for sorting the document 18 into the appropriate pocket within the sorting means 120. In this way the sorting means 120 is controlled by data derived by the data development means 112 instead of having to rely on non-controlled information contained on the document itself during sorting by sorting means 120. This enables each document 18 to be accounted for during the sorting means function which reduces errors and reduces manual handling.

In order to satisfy current banking require¬ ments, those "transit" documents which have been sorted out by the sorting means 120 (Fig. 1) must be "MICR qualified" which means that the monetary amount of the document 18 must be MICR encoded thereon; a document encoder qualifier means 122 is provided for this pur¬ pose. The encoder qualifier means 122 is basically a conventional MICR printer; however, it includes a reader (not shown but similar to reader 72 in Fig. 3A) for reading the associated DI# on the document 18. The

monetary amount for the document is obtained by associ¬ ating the DI# on the document with the associated data record 74-D in the data file means 114 and the monetary amount is then MICR printed on the document. A MICR reader (not shown) associated with the encoder qualifier means 122 then reads the newly printed data, and a check is performed at the encoder qualifier means 122, insuring that the monetary amount just printed on the document 18 corresponds to the amount stored in the data, f le means .. 114. Those "transit" documents which exit from the encoder qualifier means 122 are again sorted at the sorting means 120 according to data identifying the particular banks outside the banking system 10 for the usual processing. Non-qualified "transit" documents 18 are separated, appropriately marked, and sent to the outside banks. ' ^•

Those "on-us" documents 18 which represent documents drawn on banks or branches within the banking system 10 which were sorted via the sorting means 120 (Fig. 1) are then arranged in an appropriate order for delivery to their final destination.

The data associated with a transaction group of documents 18 is extracted from the data file means 114, and is put in the appropriate format by a conven- tional interface 124. From the interface 124, the data associated with the "on-us" documents 18 is presented in the desired format to the conventional application systems 126 where reports and application posting are performed. Because there is generally a large investment in the application systems with regard to posting, report, account preparation functions and their associ¬ ated programs, the interface 124 enables the processing by the banking system 10 to interface with the applica¬ tion systems 126 without any change such as reprogramming in these application systems; this is a feature of this system.

After processing by the application systems 126 (Fig.-l), the data records 74-D from the data file

OMPI ^IP

means 114 associated with the production of reports by the application systems 126 and the images of the docu¬ ments from the image file means 100 are placed on a permanent record or archival file, as for example, a video disc 128. This video disc 128 is physically transported to the associated POP 16 where the disc 128 may be placed on a conventional display terminal 130 having an associated keyboard 132 for selecting portions of the data thereon to be displayed for visual inspec- tion; hard copies of this data to be displayed, in¬ cluding surrogate copies of documents 18, may be ob¬ tained by a conventional graphic printer 134 located .at the POP 16 and associated with the display terminal 130. The preparation of reports and account statements, etc., will be discussed in detail hereinafter.

Before proceeding with a detailed discussion of the apparatus represented by the various components shown in block form for the IPC 14 shown in Fig. 1, it would appear appropriate to provide a functional des- cription of the general flow of data and information thereat.

One of the premises for the banking system 10 is that the original documents 18 contain all the necessary inform.ation for updating the various accounts and for processing within the system 10 whether the information appears in pre-applied form or in human- written form. Therefore, the image of the document 18 which is forwarded from the ILϋ 22 must represent all this information in a readable form or the document 18 cannot be processed.

As previously stated herein, the image file means 100 contains the raw images of the documents 18 along with their TRL#s which indicate the sequence in which the documents were received at the associated POA 12, and the data development means 112 (Fig. 1) is utilized to extract data from the raw images of the documents 18 and to put it into a form which can be

O

used for electronic or machine processing. The various functions performed by the data development means 112 are shown in Fig. 10 and will be discussed prior to discussing the elements (Fig. 5C) included therein as some of the processors shown in Fig. 5C perform more than one of the functions included in Fig. 10.

A data scheduling function 136 shown in Fig. 10 is included in the broad retrieval control step 138 shown in Fig. 4. The scheduling function 136 determines when all the entry records 74 for a transaction group of documents 18 are ready for processing; when ready, these entry records 74 are entered into the data development process by the data development means 11 ^' 2\

The data scheduling function 136 (Fig. 10) utilizes a conventional decompression unit which decom¬ presses the images of the _, entry records 74 and also utilizes an algorithm which interprets the bit patterns for an entry record 74 and formats the decompressed image to prepare it for machine character recognition functions. A conventional machine pattern recognition

(MPR) unit 140 (shown in Figs. 10 and 5C) is utilized to interpret pre-applied and pre-structured fonts such as E13B and handwritten information (such as monetary amounts which appear on the front of a document) . While the interpretation of handwritten monetary amounts is not necessary for viability of the banking system 10, it is a means for increasing the efficiency of the system 10. The data which is machine or computer de¬ veloped by the MPR unit 140 is conventionally analyzed in step 142 (Fig. 4) to determine what d-ata information is incomplete, requiring human or operator input. If all the data required for processing is complete as represented by dashed line 144 (Fig. 4), the human interpretation step 146 may be eliminated. The human interpretation step 146 (Fig. 4) is effected in the data completion function 148 (Fig. 10). The particular area of an entry record 74 which requires

OMPI /.. WIPO .<

human interpretation is displayed on a cathode ray tube (CRT) 150 (Fig. 5C) to enable an operator to interpret the particular area of the image thereon and key in the appropriate data (such as the monetary amount) while

5 using a keyboard 152. The data which was just keyed in by the operator is added to the computer-derived data from step 142 to complete the data completion function 148. A plurality of CRTs 150 and keyboards 152 is provided as shown in Fig. 5C to enable a plurality of ^• 10 operators to process a plurality of entry records 74.

The step 146 (Fig. 4) completes the development of data to enable a transaction group of documents 18 to be machine processable.

After the data completion function 148 (Fig.

15 10) is effected, the developed data for a transaction group of documents 18 is then processed for the data qualification function 154 (Fig. 10) which basically checks the developed data against user-verification routines to prove data correctness and to apply routing .

20 information (outclearing destinations etc.) to the developed data.

After the data qualification function 154 (Fig. 10) is completed, the developed data for a trans¬ action group of documents 18 is processed for a batch

25 consolidation function 156. Because the developed data may be out of order due to the operators working at different rates of completion in step 146 (Fig. 4), it is necessary to arrange the developed data for a trans¬ action group of documents 18 according to the TLR num-

30 bers for that transaction group to complete the function 156.

After the batch consolidation function 156 for a transaction group of documents 18 has been com¬ pleted, a batch analysis and proving function 158

35 (Fig. 10) is initiated in which a qualification and proof step 160 (Fig. 4) is performed. During this step 160, a determination is made of those entries which have

O V WI

incorrect data which would prevent account processing and proving (zero balancing). As earlier stated herein, the data is proved or qualified when the debits equal the credits for a transaction group of documents 18. The qualified data (block 162 in Fig. 4) for a trans¬ action group of documents is stored in a particular area of _. the primary storage 116 of the data file means 114. The data for a transaction group of documents 18 which does not result in a zero-balance is considered non- proved and is shown as non-qualified data (block 164 in

Fig. 4) and is stored in a particular area of the primary storage 116 of the data file means 114 until certain correction procedures (to be later described herein) are performed. The status of each group of qualified and non-qualified data for a transaction group of documents 18 is available at all times to the system manager 108 via the console 110. (Fig. 1) and to the associated POAs 12 via the communication means 88 and the associated keyboard 92 and display 94 or printer 96 located at the POA 12. Signature cards or images 166 which are input into the system 10 via the ILϋ 22 in Fig. 2 are data completed as non-dollar batches by the data development means 112 and are used to derive account and control information therefrom; they are placed in the data file means 114 (Fig. 1) for use in updating master files which are used in reporting as will be later described herein.

The responsibility for bringing the non-proved or non-qualified data 164 (Fig. 4) for a transaction group of documents 18' into balance resides with the particular POA 12 which forwarded the associated entry records 74 to the IPC 14 because the particular docu¬ ments 18 which comprise the transactions and the asso¬ ciated depositor account relationship along with the tools for effecting the necessary changes all reside at this POA 12. As data for a transaction group of docu¬ ments 18 is developed by the data development means 112,

it is stored by the data file means 114. An operator at a POA 12 can then request (step 168 in Fig. 6) from the IPC 14 the associated transaction group status (step 169 in Fig-. 6). The transaction groups having qualified data are then transmitted via the communication means 88 to the associated POA 12, and the transaction group numbers may be printed on a list by the terminal printer 96. An operator at the POA 12 then physically ships to the IPC 14 (via step 170 in Fig. 6) the documents 18 for those transaction groups whose data is qualified, and the groups of qualified data 162 (Fig. 4) are released by the IPC 14 for subsequent processing to be later described herein.

For a group of non-qualified data 164 (Fig. 4) the operator at the POA 12 requests the IPC 14 to send the data which has been developed for each docu¬ ment 18 included within the group. The non-qualified data 164 for a transaction group is transmitted to the POA 12 and printed on a printout report by the terminal printer 96. using the printout report and the associ¬ ated documents for the group of non-qualified data, an operator is able to locate the associated errors as shown in step 172 in Fig. 6. These errors are those described earlier herein, and generally include errors made by a depositor, errors due to missing or extra documents, and errors due to faulty data interpretation. Correction documents are prepared by the operator (step 174 in Fig. 6) which correction documents identify the particular document 18 or account involved, and missing documents 18 are located and are included with the correction documents to form a transaction correction group (step 176 in Fig. 6). A transaction group control document similar to document 32 (Fig. 3.A) is also pre¬ pared. Because the ILU 22 is the only input device to the banking system 10, a transaction group control document (similar to document 32) along with the neces¬ sary correction documents are then processed through the

-^t_

ILU 22 to forward the associated entry records 74 there¬ for to the IPC 14. The entry records 74 for this trans¬ action correction group, when received at the IPC 14, are then processed by the data development means 112 as previously described to update the associated non¬ qualified data 164 stored in the data file means 114. The correction and release-of-documents process des- . cribed in relation to Fig. 6 takes place throughout a processing day at the POA 12 to provide a steady stream of work to the IPC 14.

A major requirement of any financial entry processing system entails the movement of documents to their final outclearing destinations. -At the IPC 14, all document movement is performed under processor control so that when documents cannot be accounted for, corrective action can be initiated.

Fig. 7 is a flowchart showing the movement of released documents for a qualified transaction group from a POA 12 to their outclearing destinations. The documents for a qualified transaction group are placed in the sorting means 120 (Fig. 1) where the documents are sorted in an initial breakout (step 178 in Fig. 7) based on the sort decisions that were placed in the qualified data (step 162 in Fig. 4) for that transaction group which was placed in the data file means 114 (Fig. 1). Because the sorting techniques have already been described in detail in relation to the sorting means 120 of Fig. 1, this aspect of Fig. 7 will be described only briefly. The transaction group control document 32 (Fig. 3A) for the qualified documents in a transaction group is also placed in the sorting means 120 along with the associated documents 18 in order to extract the qualified data 162 from the data file means 114 for that transaction group. All the qualified data (shown as pocket control data 186) for .that transaction group is then placed in the processor memory associated with the sorting means 120. The transaction group control docu¬ ment 32 is then placed in an internal document pocket

OMPI

within the sorting means 120. As each document 18 for the transaction group passes through the sorting means 120, the document control number (from document 32) that was placed in the bar code on that document 18 is read and associated to the associated qualified data record 74-D. The sort decision which was placed in the data record 74-D is retrieved and the associated docu¬ ment 18 is directed by the sorting means 120 into the appropriate destination pocket therein. The data record for this document 18 is then written into a control file (in the memory of the sorting means 120) for that pocket.

During the initial breakout or sort 178 (Fig. 7), the "transit" or other bank documents 182 are separ¬ ated from the "on us" documents 184. The other bank documents 182 are then amount or dollar MICR encoded at step 188 in Fig. 7. Using the qualified control data 190, the documents 18 are further sorted by other bank end points as shown in step 192. Any discrepancies or exceptions found are noted by updating the qualified control data 190. The exception documents are removed from the qualified documents and broken out separately. Detail listings 194 are produced using the qualified control data 190 in conjunction with a printer 318 (Fig. 5D) associated with the system manager 108 (Fig. 1). The transit documents 18 along with their associated detail lists 194 are prepared for shipment in step 196. The on-us documents 184 in Fig. 7 are sorted according to an intermediate sequence which best accom¬ modates the preparation of statements for customers' accounts using the associated pocket control data- 198 as previously described. In some situations, on-us docu¬ ments are rejected by the associated POP 16 as described later herein as a result of a review 226 shown in Fig. 8. The associated data in the data file means 114 for these returned documents is updated to reflect this fact and these returned documents 200 are charged to the particular POA 12 in step 202 and returned thereto for

corrective action. Certain on-us documents 18 which are received and initially processed at one IPC 14 are also separated in step 184 and are forwarded to another IPC within the banking system 10 for processing at that IPC.

The on-us documents 204 (Fig. 7) which are to be processed by this particular IPC 12 are then se- quenced by customer account number in step 206 using the associated control data 208 therefor. The functions performed within the dashed line 210 in Fig. 7 occur on statement preparation day. The control data 208 at this time has been rearranged into the appropriate machine code and data format by the interface 124 (Fig. 1) to make the data compatible with existing application systems for posting and for the preparation of customer account statements 210 and for the preparation 2i2 of the statements along with the documents for mailing to the account holders. If a particular document is mis¬ sing, -a surrogate copy of the document may be produced at the associated POP 16 by using the video disc 128 and graphic printer 134 as previously described herein. The qualified data 162 (Fig. 7) which was rearranged by the interface 124 (Fig. 1) to produce the control data 208 (Fig. 7) is also retained in the qualified data format within the data file means 114 to become the basis for account reporting. The applica¬ tion systems 126 (Fig. 1) prepare journals which reflect the daily postings just completed. These journals are analyzed, and the results of the posting for each entry, i.e., post or reject, are determined. The qualified data record is located in the sorted data file within the data file means 114, and the posting results are now ready for the reporting subsystem shown in Fig. 8. A feature of the banking system 10 is that the application systems 126 (Fig. 1) may be re-programmed to utilize the qualified data in data file means 114 directly and update them as a result of posting. In this event, the interface _, 124 on Fig. 1 could be eliminated.

Fig. 8 is a flowchart which illustrates the creation of electronic reports for a POP 16 for review, audit and customer service functions performed thereat. This is a very important concept of the banking system 10, that is, to put all information to support funds movement at the POP 16 location which has responsibility for the associated customers' accounts.

The qualified data for each of the transaction groups for each POA 12 is used to create an entry order journal. The image (front and back) of each document 18 from the image file means 100 plus the qualified data from the data file means 114 forms a data base which provides information about all entries within a trans- ^• action group. This entry order journal is used to answer deposit questions and to create surrogates for those documents 182 (Fig. 7) which were shipped to other banks not included in the BANCR system 10.

With regard to Fig. 8, the various reports (non-image application reports) shown as 214, various reporting data 216, the associated images 218 from the image file means 100, qualified transaction data 220 from the data file means 114 and the associated signa¬ tures 222 from a signature file means located at IPC 14 are used to create image reports 224 at the associated IPC 14. These image reports ^" 224 include the unique TLR, account control signatures, the qualified data, any special information put out by the posting or applica¬ tion systems 126 (Fig. 1), and the front and back image of the document. These reports 2.24 are prepared elec- tronically, using any conventional permanent record or archival techniques, and are placed on the permanent record which is shown, for example, as a video disc 128 in Fig. 8. The video discs 128 (Figs. 8 and 1) con¬ taining the information about a particular POP 16 are physically transported to that POP 16 for review thereat.

At the POP 16, the video disc 128 containing information is utilized to display the data on the

display terminal. The review of posted entries (step 226 in Fig. 8) permits an examination of all the entries without access to the physical documents 18 themselves (by using the images thereof) to insure that each entry contained all the required information such as valid data, signature, amount, endorsement, etc. Decisions on the payment or refusal of posting rejects is made by bank officers using the image and data in¬ formation from the video disc 128. Those entries which are to be refused as a result of the review 226 are placed in the banking system 10 by preparing a rejected item document (step 228) and this document is then entered into the ILϋ 22 which is located at the POA 12. Each branch bank within the banking system 10 generally performs the functions represented by the POA 12 and POP 16. The rejected item document is then identified in a transaction group of refusal documents when entered at the ILϋ 22. The entry record 74 for a rejected item document is then forwarded to the associated IPC 14 for processing thereby according to the techniques already explained. These rejected item documents are shown as returned documents 200 in Fig. 7.

In providing customer service in relation to Fig. 8, for either a deposit or account inquiry, for example, the associated video disc 128 for the entry in question is accessed. The images relating to the ques¬ tioned entry may then be viewed on the display terminal 130, or a hardcopy surrogate may be produced on the graphic printer 134 and given to the customer. The record medium, such as the video disc 128, selected for the banking system 10 should have an indefinite life to meet the financial institution's archival requirements.

Fig. 9 is an organizational chart which shows the flexibility of the banking system 10 and how infor- mation flows between the various IPCs 14 within the system 10.

The banking system 10 as described in relation to Fig. 1 was kept simple so as to facilitate a descrip-

tion of the basic functions of the system 10. The relationship between POAs 12, IPCs 14, and POPs 16 in an expanded banking system 10 is shown in Fig. 9 and the various POAs, IPCs and POPs are identical to those already described. The IPC-A shown in Fig. 9 has a plurality of POAs (POA A-l through POA A-N) and a plurality of POPs (POP A-l through POP A-N) associated therewith as shown. Because IPC-A may itself be a bank, it may cater to an "in-house" POA and POP. IPC-B similarly has a plurality of POAs (POA B-l through POA B-N) and a plurality of POPs (POP B-l through POP B-N) associated therewith. The particular IPCs may communi¬ cate with each other through a conventional high speed communication network shown as lines 229, which is connected to the communication means 88 already des¬ cribed in relation to Fig. 1.

In general, the IPCs (Fig. 9) are located geographically as close as possible to the associated POAs so as to reduce transportation time for trans- porting documents 18 etc., to improve work flow, and to reduce high-speed communication costs.

A particular IPC such as IPC-A in Fig. 9 receives entry records 74 from its associated POAs and it processes these records 74 as previously described. The qualified data developed as previously described for "on-us" documents received at IPC-A are used to prepare the necessary reports, application posting, etc., for those POPs A-l through A-N associated with it. The image data for those on-us documents which are processed by the IPC-A which have POPs which are associated with IPC-B, for example, will be sent via the communications network 229 to the IPC-B and placed in the image file means 100 thereat to enable the IPC-B to prepare the necessary reports, application posting, etc., for its associated POPs B-l through B-N. The communications network 229 facilitates inquiry on an on-line basis.

IPC 230 in Fig. 9 may be configured to handle special entries such as those associated with the use of

a ^' credit card (as for example, VISA). In this situation the images or entry records 74 (Fig. 3) could be pro¬ duced at any POA within the banking system 10 and trans¬ mitted to the IPC 230 for processing thereat as already explained.

The banking system 10 is flexible enough to permit the processing performed by the system 10 to be performed for banks outside the system 10.

Fig. 5A shows an expanded block diagram of the communication means 88 shown in Fig. 1. The com¬ munication means 88 includes a conventional buffered intelligent adapter 230 (BIA) which is used for buf¬ fering the information (entry records 74) from the ILϋ 22 at a POA 12 to a communication controller (CC) such as 232, 234 or 236 via a conventional, digital-data- system switch 238. In the banking system 10, only two of the CC's 232, 234, or 236 are used at any one time, and the remaining CC is used simply as a back-up con¬ troller in the event that one of the CCs malfunctions or becomes inoperative. A conventional BIA 240, similar to BIA 230 is used to operatively couple the associated controller, like CC 232 with the DDS switch 238. The DDS switch 238 is an electronic switch which performs the necessary switching function to switch out an in- operative communications controller (any one of CC 232, 234, or 236) and substitute therefor, the back-up con¬ troller. Each of the CCs like CC 232, for example, has a conventional, bit serial link adapter BSLA 242 and a conventional electronic switch 244 operatively coupled to the communication nodes 246 and 248 via the conven¬ tional bit serial link adapters 250. The communication node 246 includes a processor such as an NCR Criterion 8470 with 750K bytes of memory, although other conven¬ tional processors and their associated conventional interfaces such as BSLA 242 and BSLA 250 may be used.

The communications controllers 232, 234, and 236 (Fig. 5A) act as buffers in controlling the flow of

the entry records 74 to the communications nodes 246, 248 which also include memory to store portions of an entry record 74. Conventional direct link adapters 252 are used to couple the communication nodes 246, 248 to the system bus 102. When all the portions of an entry record 74 are received at one of the communication nodes 246, 248 all of these portions of an entry record are then routed to the image file means 100 (Fig. 1) under the control of an image file processor 254 (Fig. 5B) which is included in the image file means 100. When all the entry records 74 for a transaction group are re¬ ceived at the image file means 100, an end of documents 18 signal from the input hopper 24 shown in Fig. 3A indicates this fact to the system manager 108. The image file means 100 (Fig. 1) is shown in more detail in Fig. 5B. Basically, the function of the image file means 100 is to store the raw images or entry records 74 received from the POAs 12, and consequently, any conventional storing means may be used. For example, the processor 254 may be a conventional processor such as an NCR Criterion 8570 with two megabytes of memory, with the processor 254 being used to write the entry records 74 on conventional memory units such as magnetic disc units 256, 258, and 260 (such as NCR 6550 disc units), which comprise the primary storage 104 (Fig. 1). The disc units 256, 258, and 260 are conventionally interfaced with the processor 254 via conventional peripheral adapters 262 and 264 (PA) , electronic switches 266 and 268 (SW) , and bit serial link adapters 270 and 272 (BSLA) as shown in Fig. 5B. The processor is inter¬ faced with the system bus 102 via a conventional direct link adapter (DLA) 255. The back-up storage or archival storage system shown as a video disc 106 in Fig. 1 may include any conventional system such as the video re- corders 274, 276, and 278 shown in Fig. 5B. These video recorders 274, 276, and 278 are interfaced with the processor 254 via conventional peripheral adapters (PA)

280 and 282, electronic switches 284 and 286 and the BSLAs 270 and 272. Conventional video disc readers 288 and 290 are similarly, conventionally interfaced with the processor 254 via PA 292, electronic switch SW 294, and the BSLA 274 as shown in Fig. 5B. The processor 254 has a system console 296, card reader 298, and printer 300 associated therewith due to the particular processor 254 selected for use in the system 10; however, these elements are not absolutely necessary for the system 10. The printer 300 is interfaced with the BSLA 270 via a conventional peripheral adapter (PA) 302 and electronic switch 304 as shown in Fig. 5B. The processor 254 has a conventional integrated disc controller 306 associated therewith, with the associated program or operating instructions being recorded on a conventional medium such as a magnetic disc unit 308.

The system manager 108 (Fig. 1) includes a system service processor 310 (Fig. 5D) which may be a conventional processor such as an NCR Criterion 8570 with one megabyte of memory, with the processor 310 being interfaced to the system bus 102 via a conven¬ tional direct link adapter (DLA) 312. The processor 310 has two bit serial link adapters (BSLA) 314 and 316 associated therewith. The electronic switches 284, 286 and 266 shown in Fig. 5B are interfaced with the BSLA 314, and the electronic switches 294 and 268 are inter¬ faced with the BSLA 316 to enable the processor 310 to provide a back-up function for the processor 254 (Fig. 5B) . The processor 310 (Fig. 5D) has associated therewith, a printer 318, PA 320, electronic switch 322, console 324, and card reader 326 which are identical to the same counterparts already described in relation to the processor 254 shown in Fig. 5B. The electronic switch 322 and the electronic switch 304 (Fig. 5B) are interfaced with the BSLA 316 as shown.

The processor 310 (Fig. 5D) has associated therewith a conventional integrated communications

controller 328 which is used for controlling the remote consoles 330, 332, 334 and 336. The processor 310 also has associated therewith a common trunk unit 338 (C/T) which is used for interfacing the processor 310 with the magnetic tape drives 340, 342, and 344 via the common trunk units 346 and 348, and the electronic switches 350 and 352 as shown in Fig. 5D. The processor 310 has a conventional integrated disc controller (IDC) 354 which is interfaced with the magnetic disc unit 308 (Fig. 5B) and the IDC 306 associated with the processor 254. A second IDC unit 356 (Fig. 5D) associated with the pro¬ cessor 310 is coupled to the magnetic disc units 358 and 360 (Fig. 5D) and IDC 362 associated with a processor 364 (Fig. 5E) which is included in the data file means 114 shown in Fig. 1. The IDC 356 (Fig. 5D) is also interfaced with the magnetic disc units 366, 368, and 370 (Fig. 5E) and a second IDC 372 associated with the processor 364. The IDC 354 (Fig. 5D) is interfaced with the magnetic disc units 374, 376, and 378 (Fig. 5E) and the IDC 362 as shown. The disc units 374, 376, and 378 comprise the primary storage 116 shown in Fig. 1, and the disc units 366, 368, and 370 comprise the optionally-provided backup storage 118 (Fig. 1) which may also be video discs. The processor 364 (Fig. 5E) included in the data file means 100 is identical to the processor 310 (Fig. 5D) and has the usual console 380 and card reader 382 associated therewith. The processor 364 is inter¬ faced with the system bus 102 via a conventional DLA 384 and is also interfaced to the magnetic tape drives 340, 342, and 344 (Fig. 5D) via a common trunk (C/T) 386. The processor 364 has a BSLA 388 associated therewith for interfacing the processor 364 with the printer 318 shown in Fig. 5D, and it also has an ICC 390 (similar to ICC 328) for interfacing the processor 364 with the remote consoles 330-336 and the ICC 328 (Fig. 5D) via a communications switch 392, which may be a high speed

communications switch manufactured by T-Bar Corpora¬ tion.

The data development means 112 (Fig. 1) is shown in more detail in Fig. 5C and includes a plurality of conventional processors 394, 396, 398, 400 and 402

(such as NCR Criterion 8470s) which are interfaced with the system bus 102 via conventional direct link adapters (DLA) 404, and which perform the various functions shown in Fig. 10. The processor 396 has a BSLA 406 associated therewith, and similarly, the processor 398 has a BSLA 408 associated therewith to interface both of these processors with the image display controllers 410, 412 via the BSLAs 414 and 416 and the electronic switches 418 and 420 as shown in Fig. 5C to provide a back-up function as previously described. Each of the display controllers 410, 412 has the plurality of CRTs 150 and keyboards 152 associated therewith to provide for the human interpretation step 146 (Fig. 4) and to provide for the data completion function 148 (Fig. 10) already discussed herein.

The processor 394 (Fig. 5C) has the machine pattern recognition unit (MPR) 140 associated therewith to perform automatic reading of the monetary amounts and the conversion of pre-encoded E13B information, for example, on the images of the documents 18 to machine- readable data as already discussed in relation to the data scheduling function 136 (Fig. 10). The MPR unit 140 has a conventional decompression unit included therein to decompress the compressed images of the documents 18 received from the image file means 100. Processor 400 has a similar MPR unit 140 associated therewith to provide a back-up function for processor 394.

The processors 396 and 398 (Fig. 5C) are dedicated to handling the images of the documents 18 coming from the image file means 100, while the pro¬ cessors 394, 400, and 402 are utilized to perform most

of the functions shown in Fig. 10. The processors 396 and 398 have sufficient memory capacity associated therewith to buffer the images of the documents 18 and to maintain a smooth flow of images to the CRTs 150. Conventional inquiry and flagging techniques are utilized to indicate when one of the processors 396, 398 requires, for processing, the images and entry data for specific types of documents 18, and a request is made over the system bus 102 to the processor 402 which performs the data scheduling function 136. The system manager 108 is aware of the transaction groups of documents 18 which are complete in the image file means 100. The processor 402, when performing the data scheduling function 136, selects a transaction group to be routed from the image file means 100 to the waiting processor 396 or 398 via the system bus 102. Generally, each processor 396 or 398 handles the entry records 74 as individual items; therefore, it is conceivable that the processing of a single transaction group can be shared by these pro- cessors. The data scheduling function 136 (Fig. 10) includes scheduling the images (using, for example, processor 402 in Fig. 5C) for a transaction group ac¬ cording to the TLR# so as to facilitate the integration of elements of data which are developed at the data development means 112 and eventually filed in the data file means 114.

The MPR unit 140 includes a conventional character recognition reader which reads the decom¬ pressed image of a document 18 and ascertains the mone- tary amount thereon. After completion at the MPR unit

140, all the developed data for a document 18 is analyzed for completeness. When data is missing, the associated image is routed to one of the processors 396, 398 for display on one of the CRTS 150 where an operator keys in the appropriate data on an associated keyboard 152. The image display controllers 410 and 412 have conventional decompression units associated therewith for the purpose

of permitting operator viewing of the images from the file means 100. The operators complete the data com¬ pletion function 148 (Fig. 10) by keying in the appro¬ priate data such as monetary amounts (if necessary) . while using the keyboards 152. The processors 396 and 398 have appropriate programs associated therewith for controlling the flow of images to the appropriate CRTs 150 (Fig. 5C).

In addition to performing. the back-up func- tion for processor 394, the processor 400 (Fig. 5C) typically performs the data qualification function 154 and the transaction group consolidation function 156 shown in Fig. 10. Essentially, the qualification function 154 performed by processor 400 relates to verifying the data contents to insure completeness and correctness of the developed data and also relates to adding document routing instructions which are used by the sorting means 120 to "break out" the documents 18. The qualification function 154 also insures that all data records 74-D for a transaction group have been stored in the data file means 114. The entry records 74 for a transaction group of documents 18 most likely arrive in a random order in preparation for the quali¬ fication function 154; because they arrive in a random order, they are out of order according to the TLR#s within a transaction group. As a result of the entry records 74 being worked on by a plurality of operators at the CRTs 150 and keyboards 152, a check-off operation (to insure all entry records 74 are present and arranged by TLR# within a transaction group) completes the

"batch" or group qualification function 156. When the data records 74-D for a particular transaction group are qualified as discussed in the data qualification func¬ tion 154, they are considered qualified data 162 (Fig. 4) and accordingly, the data records 74-D for this transaction group are placed in the data file means 114 via the processor 402, system bus 102, and the processor

364 (Fig. 5E). At this time, it should be noted that the image of a document from the entry record 74 is not filed with the data which has been developed in machine- readable form by the data development means 112 and the image is not used during the data qualification, batch consolidation, batch analysis and proving functions 154, 156, and 158, respectively, being performed at the data development means 112. Consequently, the data which is put into the data file means 114 for a particular docu- ment 18 has been referred to as data record 74-D. The image for a particular document 18 whose data record 74- D is stored in the data file means 114 can be obtained from the image file means 100 by using the associated TLR#. After the batch consolidation function 156 is completed, the previously described batch analysis and proving function 158 (Fig. 10) is performed on processor 402 in Fig. 5C. The data records 74-D for a particular transaction group which d-o not balance out during the proving function 158 are considered non-proved or non¬ qualified data 164 (Fig. 4) and are placed in the data file means 114 where they are retained until the neces¬ sary correction entries are resolved by the associated POA 12 as previously explained herein. Whenever the data records 74-D for a transaction group are placed in the data file means 114 via the associated processor 364 (Fig. 5E), the status of these data records (qualified or non-qualified) is relayed to the processor 310 in the system manager 108. Periodically during the day, a request is made by an operator at the POA 12 via the keyboard 92 (Fig. 1) for those transaction groups whose data records 74-D have been qualified, and the processor 310 (Fig. 5D) responds thereto via the system bus 102 and the communication means 88 by printing on the printer 96 at the POA 12 a list of the qualified transaction groups. The documents 18 for qualified transaction groups are

then physically shipped from the POA 12 to the IPC 14 as previously discussed in relation to Fig. 6.

When the documents 18 for a qualified trans¬ action, group are received at the IPC 14, they are physically sorted at the sorting means 120 and encoded at the document encoder qualifier means 122 as previously discussed in relation to Fig. 7.

The sorting means 120 (Fig. 1) and the encoder qualifier means 122 are shown in more detail in Fig. 5F. The sorting means 120 includes conventional sorter controllers 422, 424, and 426 (such as NCR 6780 controllers) which are interfaced with the system bus 102 (Fig. 5F) by appropriate direct link adapters (DLA) 428. Each of the sorter controllers 422, 424, and 426 is coupled to a conventional sorter 430 such as an NCR 6780 sorter; however, these sorters 430 have been con¬ ventionally modified to include a bar code reader to read the coded information which was placed on each document 18 at the ILϋ 22 as previously explained. The sorting techniques effected on the sorting means 120 have been previously described herein in relation to Fig. 7; however, it is important to stress that the sorting is controlled by the data records 74-D for the documents 18 rather than the documents 18 themselves. The document encoder qualifier means 122

(Fig. 1) is shown in more detail in Fig. 5F and in¬ cludes an encoder qualifier processor 432 and a back-up encoder qualifier processor 434 which are interfaced with the system bus 102 by conventional direct link adapters (DLA) 436. The encoder qualifier processors 432 and 434 are conventionally interfaced with a plurality of encoders 438, 440, 442, and 444 by bit serial link adapters (BSLA) 446 and electronic switches (SW) 448, and by the BSLAs 450 and 452 as shown in Fig. 5F to provide a back-up function. The encoders like 438 are conventional and perform the function of encoding or printing the monetary amount of a check for

"transit" documents in magnetic ink. After encoding, the entire MICR band of data is MICR read to verify that the associated FRD-ABA (Federal Reserve District-Ameri¬ can Bankers Association) numbers are MICR readable by correlating the document 18 with its associated data record 74-D. The document encoder qualifier means 122 is conventional and utilizes an application program and general software for file and control.

Some of the basic functions of the system 10 which are performed through using the images of the documents 18 are:

(1) Deriving all valid data required for updating financial accounting systems;

(2) Auditing each financial entry or data record (derived from the image of the document) to insure that all required data such as date, bank, account number, amount, signature, and endorsements are present.

(3) Performing transaction balancing;

(4) Displaying the data records and associated images for making acceptance or rejection decisions with regard to the associated documents and for answering account holders* inquiries with regard to the documents;

(5) Preparing surrogates or copies of the docu¬ ments when required by an account holder or a bank in answering protests; and

(6) Maintaining an archival record of the data records and associated images for the documents.

It should be clear that the just described banking system has a number of advantages over conven- tional banking systems. These advantages are:

(1) The physical handling of accounting entries is held to a minimum.

(2) Immediate information, including reproduction of original documents, of deposit transactions, of ac- count movements, and of specific entries is available for customers, government agencies, and the courts.

(3) Images can be duplicated so that they may be used simultaneously in geographically dispersed locations

as well as multiple locations. This removes the re¬ striction of sequential processing required in paper based systems of the prior art.

(-4) Computer controlled image display of all in- formation pertaining to financial entries is available for performing all review and audit functions.

(5) All information about a physical document is retrieved using computer directed random access devices to locate and present the specific information. (6) The retrieving of physical documents for processing functions except when required by the customer is eliminated.

\ 7 ) Surrogate documents can be computer derived from electronic images if the original entry is lost or destroyed.

(8) Identifiable accounting images are reproduced in required sequences that enhance item location.

(9) Since the responsibility for the input trans¬ action rests with the point of acceptance, all data developed from the transaction image information that is not in balance is returned to the respective point of acceptance for correction. This eliminates adjust¬ ments between the processing center and the respective point of acceptance. (10) Those operations which are data processing in nature are removed from the points of acceptance. The only processing requirements for the points of accept¬ ance are to feed transaction documents into an imaging unit and to make transaction corrections.

"gfREA ^"

OMPI /,, WIPO ,