CA1213034A - Optical mark recognition for controlling input devices, hosts, and output devices - Google Patents

Abstract

ABSTRACT OF THE DISCLOSURE

The operation of a document distribution network having one or more input work stations, a linking network with one or more nodes and one or more output work stations, is controlled by a job control sheet. The job control sheet is partitioned into a plurality of control zones. Each zone contains dedicated marked sense information for controlling the input work stations, the network nodes and the output work stations. The input work station includes a marked sense recognition device which coacts with the job control sheet to identify the presence or absence of the control zones. Marked sense information which is associated with the input station control zones is extracted and utilized to control the input work station. The marked sense information which is associated with network nodes control zone is encoded and transmitted with identifying marks to the network nodes for further processing. The marked sense information which is associated with the output station control zone is extracted, encoded and is transmitted to the network node. The network node routes the information to the output work station.

Description

~3~3~

OPTICAL MARK RECOGNITION FOR CONTROLLING
INPUT DEVICES/ EIOSTS, AND OUTPUT DEVICES

CROSS-REFERENCE TO ELATED PATENT APPLICATION

Reference is hereby made to Canadian Patent Application, Serial No. 387,649 filed on October 9, 1981 by K.A. Bushaw et al and entitled "Modular Processor Controlled Document Distribution TerminalO" The subject application describes a noncoded input (NCI) data terminal including a facsimile scanner with multiple interfaces and multiple processors to control the data entry or scan, print and communication functions.

BACKGROUND OF THE INVENTION

1) Field of the Invention .
The present invention relates to document distribution networks and more particularly to a control mechanism for controlling the distribution of documents in a facsimile distribution system.

2) Prior Art The use of facsimile communication systems for distributing documents is well known in the prior art. Such systems include one or more transmitting terminals, one or more receiving terminals and a communication 1ink interconnecting the terminals. Usually the transmitting terminal includes a raster input scanner which scans a document and generates a video stream of pixels representing the informational contents of the document. The video stream of pixels is digitized, imposed on a carrier signal and is transmitted via the communication lin~ to the receiving terminal. A
typical receiving station includes a demodulator which removes the digitized data. The digitized data is utilized by a printer '~

~L3g~3~

to reproduce a copy (facsimile) of the transmitted document. Facsimile communication systems operate either in the so-called point-to-point mode or the distributive mode. When operating in the point-to-point mode, the transmitting terminal communicatesdirectly with a receiving terminal. When operating in the distributive mode, data from the transmitting terminal is redistribu-ted by other agents on the communication link.

10 To facilitate the control of data, the communication ' links are fitted with nodes or switching junctions.
The nodes are further fitted with host controllers ~such as computers). The primary function of the host is to control the flow of data throu~h the communication link and t~ ensure that the link operates at maximum efficiency. To this end, data is often stored at a host and transmitted at a more opportune time. Of course, there are other well-known functions which are performed by the host. These functions will not be described because they are not germane to the present invention.

Substantial effort and cost have been expended to automate facsimile systems. Automation dictates that these systems function with a minimal amount of operator intervention. To achieve this, the prior art has adapted se~eral devices and/or techni~ues to control the systems. In one prior art technique, a coded information (CI) input terminal, such as a keyboard, is used to key controlled information into the system. Such control information includes recipient addresses, distribution commands, printing commands, scanning commands, etc. Although this approach works well for its intended purpose, it undermines the automatic philosophy because an operator is needed to enter the controlled information.

..

~Z~3~3~
B09~100~ 3 U.S. Patents 3,534,495, 4,~86,443 and 4,207,598 disclose another technique used in the prior art to automate document distribution. U.S. Patent 3,594,~95 describes a radio facsimile postal system including a device for transmitting signals representative of a written message and a radio facsimile recorder to receive the transmitted signals. The addressee is identified by unique code signals embedded in the transmitted signals. ~ach recorder is conditioned to respond to a different code sig~al. Therefore, the transmitted message will only be reproduced by the recorder which is conditioned to respond to the transmitted code.

U.S. Patent 4,207,598 describes an automatic mail system wherein a television camera scans a document and the information is transmitted to a storage in a telephone exchange. The identity of the addressee is sent to the exchange where it is deciphered and a path to the destination is set up automatically.

U.S. Patent 4,086,443 describes the use of a job separation sheet which may be mark-sensed by a facsimile scanner to detect the address of the receiver.

U.S. Patent 4,058,672 is yet another example of the prior art control techni~ues. The patent discloses a data communication network where data blocks are pl~ced in packets along with addresses and priority information. The communication network then transmits the packets in the most efficient way possible.
Upon receiving the packets, the receiver puts the data message back together. Although the different automatic techni~ues described in the ~atents seem to be a step in the right direction, none of these techniques disclose or suggest a method wherein the sender controls the entire system.

~Z~L3q;~34 ~U~MARY OF T~IE INVENTION

It is therefore the general object of the present invention to provide a more efficient job control apparatus than has heretofore been possible.

It is another object of the present invention to provide a job control apparatus which enables a sender to control the entire document distribution system.

In accordance with the present invention, the above objects are accomplished by providing a document distribution control sheet having a plurality of control zones thereon.
Each of the control zones is marked with information for controllin~ either the sending unit, the communication nodes and/or hosts and the receiver. A mark-sensing device (including hardware and/or software) senses the marks on the document distribution control sheet and generates a bit stream representative of the marks thereon. Means are provided to extract and decipher the bits which contain instructions for controlling the sending unit and the receiving unit. The information which controls the sending unit is used locally (that is at the sending unit) while the information for controlling the receiver is transmitted over the communicating network to the appropriate recei~er. The bits containing instructions which control the node are encoded and forwarded over the communciation link. Means 2S are provided at the nodes to decipher the transmitted bits and to distribute the transmitted information (documents) in accordance with the transmitted instructions.

BO9-81-00~
.

B0981009 5 12~3~34 The foregoing and other objects, features and advan-tages of the invention will be apparent from the following more particular description of the preferred embodiment of the invention, as illustrated in the accompanying drawings.

BRIEF DESCRIPTION OF TH~ DRAWINGS

FIG. l shows a block diagram representation of a document distribution system with a job control sheet according to the teachings of the present invention.

FIG. 2 shows a block diagram of the components used to process the marked sense indicia on the job control sheet.

FIG. 3 shows the general layout of the job control sheet.

FIG. 4 shows a specific layout for one type of job control sheet.

FIG. 4A shows a schematic of a procedure to identify the skew associated with a sheet.

FIG. 5 shows a block diagram of the hardware detection lo~ic.

FIG. 6 shows a logical flowchart of how the logic works.

FIG. 7 shows a typical network configuration with a controller, at the network node, to process the contents of the -ontrol cover sheet.

' ' , , 6 ~3~

FIGS. 8A-8F describe a program to control the microcomputer which assimilates the data generated by the hardware detection logic.

FIG. 9 shows a block diagram of the controller of FI~. 7.

DESCRIPTION OF THE PRE~ERRED EMBODIMENT

Although the present invention can be used in any type of document distribution system, it works well with a facsimile type distribution system wherein the documents to be distributed are entered by a noncoded input terminal such as an optical scanner. As such, the invention will be described in this environment. However, this should not be construed as a limitation on the scope of the present invention since it is intended that the invention can be adapted for use in any system where data to be transmitted is inputted at an input point and distributed over a communication link to one or remote users.

Referring now to FIG. l, a block diagram representation, according to the teaching of the present invention is shown~
Control sheet 10 contains a plurality of prerecorded marks and user-generated marks. The marks on the control sheet are used to control the input device 12, the host or network of hosts 14 and the facsimile output devices 16 and 18, respectively. The input device 12 is a conventional facsimile optical scanner such as that described in U.S.
Patent 4,086,448 or the above-referenced Patent Application.
The scanner scans the information recorded on the ~ob control sheet or card and generates a plurality of bit , .

B09~1009 7 ~2~3~34 streams representative of the informational contents of the sheet. The control information A and B which is recorded in zone 20 of the control sheet is used to control functions in the scanner 12 and the facsimile output devices 16 and 18, respectively.
Likewise, the control information recorded at area 22 of the control sheet is used to control functions in the hosts 14.

By way of example, the information (identified by A) is used for controlling the scanner 12. Likewise, the information B and C control the network hosts and the output terminal, respectively. To this end, A is designated in proximity to the scanner while B
and C are designated next to hosts and output termin-als. Data rom the scanner 12 is transmitted overcommunication link 24 to the hosts 14 at network node 15. Similarly, information from the hosts is transmitted over communication links 26 and 28, respectively, to facsimile output devices 16 and 18.
It should be noted that information is transmitted from the input device to the hosts and from the hosts to the output device. In a point-to-point communication link, data from scanner 12 is directed through th~ link to one facsimile output terminal.
By preceding each document or group of documents (not shown) to be distributed with an appropriately marked controI sheet l~, the user via the user marks effect batch document input to the distribution network, control device and host functions in the 30 ~etwork, and the distributed batch output across the network. The det~ils of the job control sheet) together with the necessary hardware. devices and software algorithms which are used to process the control information recorded on the control sheet will now be described.
.

B09~1009 8 ~3~34 1. The Control Sheet Referring no~ to FIG. 3, the layout of the control sheet 10 is shown. The function of the control sheet is to convey control information to the work station (input and ou~put devices) and to the document distribution nodes. By way of example, functions that may be controlled for the input device are as follows: coarse or fine scan, si~e of paper, encrypt, etc. Similarly, functions that may be controlled for the outpu~ device are image width, white margin, decrypt, etc. Likewise, functions that may be controlled for the distribution nodes are distribute, file, obtain, etc. To this end, the sheet is lo~ically divided into controlled areas. Each controlled area is fitted with markable locations. The locations are marked with control information which is sensed and is used to control a physical section of the distribution system. Each control area is divided into one or more horizontal control bands which extend from left to right of the sheet. The control sheet comprises a fixed grid of markable positions.
A column of reference marks 2~ is disposed on the left-hand edge of the control sheet. Each row of mark-sensed position is identified by one of the reerence marks on the left-hand side of the sheet.
As will be described hereinafter, presence of a reference mark indicates to the reco~nition hardware and recognition microcode that there is a row of marks extending from the left to right which has to be processed for control information. Stated another way, the presence or absence of a mark in column 24 indicates the presence or absence of a row of markings.

Although the areas on the card can be changed~around without departing from the scope of the present lnvention, in the preferred embodiment of this ~Z~3~

invention, the area of the card identified by numeral 26 is referred to as the top recognition band. The recognizable information in this band is mark-sensed in row 2~. As stated before, the presence of row 2 is indicated by the reference mark disposed in column 24 and identified by numeral 29. The presence of this mark indicates to the recognition hardware and microcode (to be described hereinafter) that there is additional information in the row to be extracted. The information is embodied ln the markings. The information in row 28 is recognized and analyzed by the rocognition logic (to be described hereinafter) and disposed in the input device. The mark-sensed information in the top recognition band 26 and in the top reco-gnition row 28 in particularly, indicates that the sheet represents a control sheet rather than a data sheet. Also, the control-sheet t~pe is included in the recognition row 28. Althoùgh it is within the skill of the art to devise a plurality of types of control sheets, in the pre~erred embodiment of this invention, there are four types of control sheets.

The first type is the so-called "document control sheet~" This sheet is always the first page of a document to be transmitted. Its function is to indicate to the work station, the machine controls and to the document distribution node, the control parameters and addressing information to be used with document transmission.

~nother type of control sheet is the so-called "continuation control sheet." This type~ when used, follows a document control sheet. Its purpose is to provide the sender with additional address specifica-tion areas. The control-sheet type field must be filled in, either manually or preprinted if this

3~

type of control sheet is used. One or more continuous control sheets may follow immediately after a document control sheet.

Another type of control sheet is the so-called l'work station control sheet." Tllis type of control sheet is used to request services of a work sta-tion. The work station type sheet will be processed by the work station without involvement of any other work station or document distribution node. All current documents send/receive operations will be terminated prior to the processing work station type control sheet requests. An example of this type is the encryption type. The encryption type can be used to enter/change and cipher keys of a related keyboard.

Still another type of control sheet is a document distribution node control sheet. This type of control sheet is used to request services of the document distribution node.

Still referring to FIG. 3, the next band in order from the top of control sheet 10 is the machine control band 30. This band is used by the sender to indicate the work station control parameters required for scanning and printing, respectively. Although a plurality of rows performing different types of functions can be positioned within this band, in the preferred embodiment of this invention, the machine control band is divided into two rows. Each row is identified by a recognition mark in recognition column 24. The first row 32 is used to control the setup control parameters for the sending unit or scanner. The type of information which is marked in this row are resolution scan size, secure (encrypted) and darkness, etc. Control information for the output device such as a printer is marked in row 34 of the machine control band.

.

.
.

~2~L3~34 The document distribution node control area is identified by numeràl 36. This area contains marked information which is to be utilized by the document distribution node. The input scanner and associated control have the responsibility of recognizing the presence of marks or no marks in this area and present a corresponding bit stream representative of the mark and/or no mark condition to the document distribution node. As will be described hereinafter, the transmitted string of bits is processed by the host software algorithms in the network nodes and are used for performing routing functions in the distribution system. The document distribution band area is divided into the distribution control bands 38 and distribution address bands 40. The distribution control band 38 includes one or more rows of infor-mation. As before, each row is identified by a recognition mark in recognition column 24. The distribution control band 38 is used by the sender to provide instructions to the document distribution node. By way of example, the distrlbution bands may contain format identification used to identify the mask/table required to translate each marked position.
Delivery and transmission requirements, such as confirmation of delivery, personal, confidential and priority information can be placed in this band.
Print output requirements and sender and document identification data can be included in this band.

The distribution address band 40 contains the auxiliary hand-written, typed names or organizational desig~
nation6 of the document recipients. The operator at the receiving work station must read these names in order to deliver the received document to each of the recipients associated with that work station.
3~ The distribution network addresses of the recipients are indicated by user marks which are recognized by B09alO09 12 ~3~3~
the work station ~nd sent to the document distribution node for interpretation. As before, each of these information is recognized by recognition mark in the recognition æone 24. The addresses may be indicated individually or as distribution list names or a combination of both. As is used in this application, the distribution list name identifies a previously defined collection of destinations, addresses and document recipients stored in the document distribution node.

The bottom recognition area identified by numeral 42 is contiguous to the document distribution area 36.
It contains a band of recognized marks identified by numeral 44. These marks are analyzed and recognized by the input device. The last mark identified by numeral 43 in row ~4 is identical to the first mark identified ~y numeral 29 in row 28~ This correspon-dence allows a system to recognize a sheet as a control sheet even if the sheet is-inserted upside down. (However, if the control sheet 10 is inserted upside down, the system will not process the sheet and the operator will be informed accordingly.) FIG. ~ shows a specific layout for a document c~ntrol type sheet. In normal operation of the facsimile ~5 distribution system, this sheet is appended to one or more documents to be distributed. The sheet is the first to be scanned by the scanner, processed and transmitted in the network system. The data extracted from the sheet controls the distribution of the appended documents. The sheet is scanned from top to bottom and left to right. As stated before, the sheet comprises a reference mark area 46. The mark in this area informs the recognition hardware and associated algorithm that there is additional information to be extracted from the row 3~
B098100g 13 which is associated with a specific reference mark.
The first hori~ontal row of data on the sheet i5 identified by numeral ~8. This row is referred to as the recognition row or recognition band. The information in this band informs the system that the sheet is a control sheet. Likewise, the information in bottom recognition band 50 informs the system that the sheet is also a recognition sheet. Reference mark 5~ indicates to the system that there is an associated row-of marked information which has to be sensed and processed. Mark 54 indicates to the system that the additional mark in the row contains control information for controllin~ the scanner.
Similarly, marks 56 and 58, respectively, would be translated into control parameters for driving the scanner. Likewise, marks 60 and 62 inform the scanner that there is associated control information to be processed on the sheet. Mark 64 informs the system that the inormation associated with that row is to control the output device such as a printer.
Mark 66 contains the type of control information which is transmitted to the printer. It should be noted that the marks in column ~6, rows 48, 50 and the marks in area ~3 all have different sizes (heights 2S and widths). The difference in size is used by the mark sense controiler ~to be described below) tc identify the marks. To this end, it is preferable that the marks be prerecorded.
.
` The area of the sheet between reference marks 62 and 68, respectively, is ïgnored by the scanner. This is so because there is no recognition mark in that area. The area of the sheet between marks 68 and 70 is processed by the scanner. In one embodiment of the present invention, those areas are left blank and can be filled in by an operator. By way of example, the blocks 72, 74 and 76 are reserved for ~z~3~3~

addressees who should receive the documents to be transmitted. The location code extension/list are filled in by the operator. By way of example, if the operator desires the documents to be transmitted to a set of receivers whose names are identified in block 72 and reside in Boston and Dallas, the operator would fill out the card as shown. The extension and list would also be filled out. It should be noted that with this specific type of form, the name o the receiver and the location are free-formatted.
This means that the operator can fill in the informa-tion into grids designated on the form. Likewise, if the set of receivers recorded in block 72 are in Pittsburg and some distribution lists disposed at the distribution node, then the form is filled out as is shown. Likewise, if the receivers recorded in space 76 were in Los Angeles and Miami, the form would'be filled out as marked. It is clear from the' above that this specific form as prerecorded machine information such as the vertical recognition zone 46, the top and bottom recognition zones 48 and 50, respectively. The prerecorded information is used by the recognition hardware and microcode which will be described subse~uently. However, the name of the parties who receive the document and the location to which the document is sent can be filled in by the sender. Of course, the entire form can be free-ormatted (t}lat is prepared by a user) and~or pre-recorded (prepared by a machine) without departing from the scope of the present invention.

In summary, the above-described job control sheet is fitted with a plurality of dedicated'control areas.
Each control area cont~ins a plurality of bands or rows of information which contain control information for controlling the scanner (sending unit), the printer (receiving unit) and the network (nodes) over which the document is routed for distribution.

12~L3~3~

2. The Mark Sense Controller -(a~ Mark Detection Hardware Logic Means FIG. 2 shows a ~lock diagram of the components used to process the indicia on the job control sheet. The component comprises a mark sense recognition means 62. The mark sense recognition means 62 is placed within the input de~Jice 12.
A document distribution control means 64 is disposed at the node 66 of the document distribution network. In the preferred embodiment of this in~ention, the input work station 12 is a noncoded input terminal. The terminal includes a conventional optical scanner and controls 68.
The scanner can be of the t~pe described in the above-referenced patent application assigned to the assignee of the present invention or the scanner described in U.S.
Patent ~,086,443. Since the scanner and controls are conventional, the details will not be described. SufEice it to say that the scanner includes printer controls, data link and transmission controls coacting under the control of microcode to scan a document, buffer a stream of zeroes and ones bit in a buffer (not shown). The bits represent the informational contents of the scanned ~age. To this end, the control sheet 10 is placed with the optical marking facing the scannin~ elements of the scanner. As bits are placed in the scanner buffer, the mark sense recognition means 62 transfer a scanned line of data over conductor 68.
By examining the bits recorded in the recognition bands of the control sheet, the mark sense recognition means determines whether or not sheet 10 is a control sh~et or a data sheet. If it is a control sheet, the information in zone A and B of the sheet is BO9-81~009 L3~

deciphered in the mark sense recognition means 62 and is used for set~ing parameters of the scanner and remote printers 16 and 18.

Likewise, information in zone C of the control sheet lO is formulated into a bit stream by the mark sense recognition means and is transferred over the communication link to document distribution control means 64. The document distribution control means 64 (to be described hereinafter) analyzes the transfer data and formulates meaningful parameters for distributing the data to the appropriate receiving station 16 or 18, respectively. In the preferred embodiment of this invention, the distribution control means 64 includes a plurality of software algorithms. By way of example, the software algorithms of block 61 transform some of the transmitted bit strings into document distribution parameters. Likewise, the software algorithms of block 63 control the network node document routing functions. Of course it is within ~he skill of the art to use hardware and/or other types of software without departing from the scope of the present invention. ~ata is trans~erred over conductor 70 and 72 to the respective receiving terminal. At the terminal, the printers are controlled by parameters which are extracted from the control sheet in mark sense recognition means 6~ and trans-mitted over the communication channel and through tlle node66 to the appropriate printer. Data transmitted from other input units to input unit 12 is printed out on printer 74.
It is worthwhile noting that other types of conventional network devioes and controls are positioned at node 66.
Since those devices are conventional, details will not be given. By way of example, an article entitled "An Anatomy of SNA" by Gary D. Schultz and published at pgs. 35-38 Com~uter World/Extra describes a communication ,~:, ..

~Z1~3~

system with a network node with conventional data processing and routing devices. The article is incorporated herein by reference and can be consulted for details of the conventional elements.

Still referring to FIG. 2, in the preferred embodiment of this invention, the mark sense recognition means -62 comprises mark detection hardware logic means 76 coupled over conductor 78 to cover sheet recognition microcode algorithm 80. Controlled information from cover sheet recognition microcode algorithm 80 is transmitted over conductor 8~ to the mark detection hardware logic means. The mark detection hardware-logic means acts as an assistant to the cover sheet recognition microcode algorithm to determine whether a given scanned document is or is not a control sheet. Once a given document is determined to be a control sheet, the mark detection hardware logic means continues ta detect and read mark informatibn from the page. 'rhe logic keeps track of what it 20 sees on the page in a horizontal dimension. It does not make a decision on whether or not it saw valid marks on the page. It keeps a tally of how many black PELs ~picture elements) it sees as it scans across valid mark areas. This tally is stored into memory for the software to examine. Software uses this tally result along with the counts from other scans, if necessary, in order to make mark decision.

Before the components in mark detection hardware logic 76 are described, a logical description of the process steps utilized by the logic to determine the presence or absence of a mark, will be given. ~IG.
6 shows a graphical representation of how the logic operates. The control sheet 10 (FIG. 2) is scanned by scanner 1~ in the direction shown by arrows 82 and 84, respectively. In other words, the sheet is ~2~3Q3~

s~anned from top to bottom and left to right. The bit stream representative of the informational contents of the page is loaded in sequential order ir the buffer (not shown). Aft~r four scan lines have been accumulated, the logic 76, in processing the data, shifts four scan lines from the scanner buffer (not shown) into its own buffer. A moving window (preferably a 4X5 shift matrix) is moved over the buffered lines. It is shifted right 12 PELs and then checks for the pattern identified by numeral 86. This pattern is a nearly solid white block. If the white block is not found, the window shifts one PEL to the right and tests again until the contents o~ the shift counter is equal to 256. If the contents of the counter is 256 (test block 88) and the test pattern is not found a mark sense counter is set with a bit which indicates no reference mark found.
If, however, the shift counter is less than 256, the test is performed (block 90) to see if the pattern is found. If the pattern is not found, the proyram loops along path 92 into blocks 88 and 90, respec-tively. If the pattern is found, the program proceeds along path 94 and shifts the window right one PEL
and examines for another pattern indicated by block 96. The pattern in block 96 is any 16 black PELs out of a possible 20 black PELs. Stated another way, the program is now looking for an are~ at least 75% black. The program then checks to see if the contents of the shift counter is equal to 256 ~bloc~
98). If it is, the program ends and sets the no reference mark found in the reference mar~ counter.
If the counter is less than 256, the program tests for the pattern tthat is black) block 102. If it is not found, the program loops along path lO0 and betwèen blocks 98 and 102 until the pattern is found. Once the pattern is found, the size of the moving window which is used to determine the presence BOg81009 19 ~213~4 or absence of the reference mark, is changed. The new window is two vertically adjoining shi*t matrixes identified by numerals 104 and 106, respectively.
In the preferred embodiment of this invention, the size of that matrix is 2Y,2.

Still referring to FIG. 6, a software program then starts a down count and shifts the new windows one PEL position right for each count. The total count is equivalent to the position where the other reference mark is located on the identified row. Stated ` another way, once a reference mark is identified in column a6 ~FIG. 4j, a count is generated in a counter.
The count is equivalent to the PELs (picture elements) re~uired to move the window from the reference mark to the first mark area on the row. The count is referred to as a "down count." ~1hen the count is completed, the program starts an "up count." The accumulated co~nt which is zero the first time, is outputted. The test in block 108 is then performed.
The test requires that there is four black PELs in the upper quadrant of the new matrix 104. If there is four black PELs, the PEL count is incremented by one (block 110). The`program then checks to see if there are four black PELs in the lower quadrant. If there are four black PELs, the program traverses along path 116 into block 118. In block 118, the PEL counk is incremented by one and the window is shifted right (block 120). The window is shifted right also if there is less than four black PELs in the lower quadrant. The hardware assist program then proceeds into block 122.

In block 122, the program tests to see if the number of shifts equals the 77Up count." If it does, the program saves the count, block 124, and traverses along path 126 to start down count snd shifts the `
` ' .

3~

windows duri~g the down count intervalve. If the number of shifts is less than the contents of the up count (block 122), the shift windows 104 and 106, respectively, are shifted one PEL to the right (block 128). The program then traverses path 130 to block 108. The above-enunciated routine continues until the end of the scan line is reached and then the program stops the hardware. The end of a scan line is reached when a predetermined number of PELs has been processed.

The hardware remains in the stop mode until it is restarted by the hardware microcode. To start the hardware, the microcode sets a control bit in a control register. Whenever the hardware completes the above-enunciated routine, the bit is reset.
.
In essence, the mark sense hardware and its associated microcode (mar~ detection hardware logic means 76, FIG. 2) senses for a reference mark which is disposed on the left side of the control sheet. The logic servo onto`the mark by observing a solid white area followed by a substantially black area. Once this determination is made, a relatively small matrix window moves across the horizontal line or horizontal row to determine marks. The digital representation of those marks are tallied and are placed into memory where it is accessed by the cover sheet recognition microcode algorithm 80. The algorithm will be fully described hereinafter. Suffice it to say that the microcode algorithm generates meaningful control parameters from the tallied counts.
:
Referring now to FIG. 5, a block diagram ~or the hardware components of the detection hardware logic means ~6 (F~G. 2) are shown. The hardware comprises a buffer l28. The b~ffer stores the scan lines of ~2~3~;33~

data which are operated upon by moving windows 130, 104 and 106, respectively. Data (that is scan lines) is conveyed over conductor 132 and 134, respectively. Data for buf~er 128 is obtained from S the regular scanner buffer. ~hen the matrix window 130 is scanning buffer 128, the output from buffer 128 is coupled over conductor or buss 138 into solid white area decoder 140 and black area decoder 142.
Twenty of the transmitted bits are transferred into the substantially black area decoder 142 and 16 of the transmitted bits are transferred into solid white area decoder 140. Buffer 128 is controlled by buffer control unit 136. The function of bold white area decoder 140 is to decode the 16 bits coming in to determine whether or not there is a bold white area. Similarly, the function of bold black area decoder 1~2 is to decode the 20 bits coming in to see if a bold black area follows the bold white area. It is worthwhile noting that the 16 and 20 bits respectively are those generated from the relatively wide matrix window 130.

The output from the white bold area decoder laO is coupled over conductor 1~4 into reference mark detection 146. Likewise, the output from bold black area decoder 142 is coupled over conductor 14~ into the reference mark detection 1~6. The function of the reference mark detection 145 is to detect tile reference mark. One of its outputs is coupled over conductor 150 into the no reference mark counter.
The output from the no reference mark counter is fed over conductor 152 to inform the logic assist microcode that no reference mark is found. The other output from the reference mar~ detection 146 is coupled over conductor 156 to valid sample count generator - 35 158. The valid sample count is used to determine up - and down counts with software settable counts. The B0981009 22 ~2 3~34 valid sample count generator 158 is controlled by information outputted on buss 160 from up-count register counter 162 and information on buss 164 outputted by down-count register counter 166. A~
stated before, the up-count reyister counter and the down-count register counter are loaded by the hardware assist microcode. The ou-tput from the valid sample count generator 158 is fed over conductor 168 into PEL sampler 170. Up to eight PELs can be inputted into PEL sampler 170 over conductor 172. The eight PELs are generated from the 2X2 matrix windows 104 and 106, respectively. The output from PEL sampler 170 is coupled over conductor 174 into tally counter 176. The tally coun'er is an eight-bit counter and its output is fed o~er a write multiplexor buss into a shared memory (not shown). The tally count which is written into the shared memory is accessed by the cover sheet recognition microcode algorithm 80 (FIC.
2) to determine the presence or absence of a mark on the sheet. This cover sheet recognition microcode algorithm 80 will be described shortly.

In operation, the scanner adapter is the first hardware to be started. It wil~ place the scanned data into the buffer associated with the scanner.
When four or more lines have been placed in the scanner memory, the mark detection hardware logic means (FIGS. 2 and 5) is started. Mark sense will fetch those four scanned lines and store them in buffer 128 (FIG. 5). The mark sense will not be started again until at least four more lines ha-~e been scanned. By mo~ing the matrix 130 over the buffered lines, a snapshot of the scanned page which covers an image area of four scan lines high by five PEL location wide is examined. This allows the detection of bold white or black image areas and allows PEL densities to ~e counted over four scan 3~3~

lines simultaneously. The bold white area decode 140 is a 16 input and function which examines the complements of the 20 bits from the matrix 130. The bold black area decode 142 examines all 20 bits. A
S given number of missing bits is allowed. This enables detection of bold black areas that are less than 100% black, such as picture mode data.

The mark sense process starts when told to do so by the microcode setting the run bit in a control register ~not shown). The hardware disregards the first 12 PELs and then starts looking for the reference mark. If the reference mark is found when one or more bold white areas are followed by at least one bold black area. The bold area samples occur once each window shift cycle or once each time data shifts one PEL position. Once the bold white area followed by the substantial black area is found, the valid sample time generator 158 is turned on. The up-count of the generator output acts as a gate which allows the tally counter 176 to count. The down-count stops the counter and the count information is transferred to a temporary register (not shown) and then to a designated area in the shared mem~ry.
This area is accessible to the cover sheet recognition microcode algorithm 80 (FIG. 2) which will be described hereinafter. The following up-count start clears the counter so that another count can be made.
During the ne~t up-count, the data in the register is written in time share memory. The up-count from the generator only occurs in valid mark areas.
Immediately following the detection of the reference mark, the generator starts its downtime count. The generator is implemented with two shift registers (nonbinary) counters and two registers which are loaded by software. The down counter runs first, then the up counter, and so on.

.

B0981009 24 ~3~3~
During the up-count (valid time) the PEL sampler operates. The tally counter 176 will count only if P~Ls are seen. Actually, the tally counter only increments when both in valid time and when PELs are detected in clusters of four (2X2). These minor bold area decodes assure that specks and smudges will add little or no significance to -the tally counts. ~ach cluster area (one 2X2 area above another 2X2 area) is sampled during each machine cycle. The clusters shift one PEL position ~efore they are each sampled in the next cycle. The overall effect is that the tally count is equal to half the actual bit count minus one. The t~o clusters are decoded from the two leftmost columns of the larger five-wide, four-high matrix identified by numeral 130. At the beginning of each scan, the no reference mark counter is turned on. If the counter times out before a reference mark is detected, the logic stops and resets. The counter times out when it reaches 256 PEL counts.

b) Cover Sheet Microcode Algorithm Still referring to FIG. 2, when the mark detection hardware logic means 76 completes processing a band or row of marks on the control sheet 10 t it interrupts the cover sheet recognition microcode algorithm 80.
The microcode al~orithm 80 then processes the marks.
As such, the cover sheet recognition microcode algorithm 80 wil~ hereinafter be referred to as the cover sheet recognition interrupt handler. In essence, the microcode interrupt handler is a software pro~ram which resides in an interrupt driven type microprocessor. The microprocessor may be a dedicated processor or the program could be a subprogram in a microprocessor which performs other unctions other than processing the marked information on the control .

3~;~4 B098100g 25 sheet. Also, other types of conventional processors other than irlterrupt driven type processors may be used for processing the data.

The cover sheet recognition interrupt handler (soft-ware) is entered upon the occurrence of an interrupt from the control sheet recognition hardware. Its function is -to accumulate the results of the control sheet recognition hard~are, to recognize valid and invalid control sheets and to produce a bit striny of the marks of the valid control sheets. It dis-tinguishes between marks intended to control the scanner and the recipient printer and marks intended -to control the distribution of the document. The marks which are intended for controlling -the scanner are extracted and are formulated into parameters.
The parameters are used ~y the conventional scanner control mechanisms to control the scanning of the documents. The marks intended to control remote printers are extracted, formulated into usable parameters and forwarded (that is transmitted) through the network to the appropriate printers.
The transmitted parameters are used by the conventional printer controls to reproduce the documents. Marks intended for controlling the distribution of the 2S document are simulated into bit streams with appro-priate headers and forward to the nodes. At the nodes, additional software handlers (to be described hereinafter) converts the control bit stream into meaningful parameters which are used to route the documents to their appropriate dèstinations. It ` should be noted that the control sheet recognition is a continuous control process drîven by hardware interrupts. The control sheet recognition process is controlled by the presence or absence o-f a reference mark positioned along the left margin of control sheet 10 (FIGS. 3 and 4). If the reference mark is 26 ~ 34 found by the mark detection hardware logic means 76 (FIG.
2), the horizontal band or row which with the reference mark is monitored and a set of counts representative of the PELs detected in the marked areas is generated. If the reference mark is not seen within approximately one inch of the left edge of the control sheet, the line is skipped and no counts are accumulated. The counts produced by the hardware are thresholded and, if between certain minimum and maximum values, are translated into a logical "1" bit for a mark position or a logical "0" bit for an unmarked position. The thresholded values used are different for different types of marked positions on the control sheet. For instance, the values used for thresholding a top recognition band mark are different from those values used in working with a user mark.

FIG. 8A shows the main program which is used to accumulate the results of the mark detection hardware logic means 76 ~FIG. ~). The program includes a plurality of subroutines which are s~own in FIGS. 8B-8F. The program is written in a high level structured format. This format is well known -to those having ordinary skill in the programming art and therefore each step of the program will not be described.
The program is initiated by an interrupt signal outputted from the mark detection hardware logic means 76 (FIG. 2).
This step is indicated in FIG~ 8~ by the instruction identified by numeral CSR00041. The next step in the program is identified by numeral CSR00042. This step is a select instruction. The select instruction is a case statement which introduces several different alternatives.
It should be noted that the first step of the program CSR00042 ends with the end select statement identified by instruction number CSR00074. There are two . ~, ~Z~3~3~
~0981009 27 alternatives which follow the select statement. The first alternative is the "when" alternative beginning at the instruction line identified by CSR0004~. If the statement is true, that is no reference mark was found on the current line, the program performs a step-beginning at CSR00045, ending with the end-when statement identified by instruction number CSR00067.
If reference mark was found (instruction number CSR00044), the program goes to instruction number CSR09068. There it tests to see if skew of the scanned line is acceptable (instruction number CSR00069). It then performs the function identified by instruction CSR00070 through the end-when statement CSR00073.

As stated before, there are several subroutines in the program. Each of these subroutines will be identified and the corresponding figures which show the subroutine programs will be described. The first subroutine in the program is the determine skew subroutine CSR00055. The program steps for this routine are given in FIG. 8D. The function of this subroutine is to determine if the skew of the control sheet is within acceptable limits. In other words, this function determines whether or not the counter values supplied by the control sheet recogni-tion hardware can be used for mark determination.
The underlying test is whether the page is skewed in relation to the travel of the scanner. The rightmost counter returned by the device represents a ske~
mark. If the scanner is currently within a reference mark, a minimum value in the skew mark counter indicates that the counter value is representing the marks located between the reference mark on the Ieft-hand side of the control sheet and the skew mark on the right should be valid. If the counter value is less than a specified minimum, there is a ~2~3~34 less than optimum chance that the counter values for the other marks on that line are ~alid. In essence, the relationship between the reference mark on the left and the skew marks (in column 51) on the right of the control sheet are used to determine the skew of the sheet.

The pattern of marks in FIG. 4A represent a row of information marks to be scanned by a linear scanner.
Marks A and C are used to define a scan window. If marks A and C cannot be detected during the same linear scan by the scanning system, the document is skewed relative to the scanner to an exten'c that the information marks B cannot be reliably read. In operation, the scanner accumulates a quantity of black for all of the marks A, ~ and C in the row being scanned. When the A mark is no lonyer detected, the scanning system determi~es how much black has been sensed from the C mark. If the ~uantity detected exceeds a threshold of 80% ~other thresholds might be used), then the scanner knows the skew is small e~ough that the scanner will have also sensed the B
marks if any. The scanner then checks for the presence of accumulated black in other portions of the row indicative of the presence of a B mark. In effect, the dimension lines 11 and 13 between the adges of the A and C mark define an acceptable skew windo~. The size of the skew window is dependent upon the size of the A and C marks and their distance apart. The A and C marks might be placed on each row to be scanned or they might be placed at the first or topmost row of a document. It should be noted that the A mark is recorded in reference mark area 46 (FIG. 4) and the C mark is recorded in column 51 (FIG. 4). Also, the B marks are in the same row with the A and C marks and intermediate said A and B marks. It should be noted that the physical size of the marks are exag~erated for purposes of description.

3~3D~

In this case an indicator flag is set to prevent the mark detection hardware logic means 76 (FIG. 2) processed from using anymore of those values. For each line received with a counter value above a certain minimum, the values of the other counters are used for that line. However, if the counter value of the skew mark does not meet that minimum requirement, the counter values are assumed to be unreliable. As a further check of tolerable skew on the cover sheet, a skew counter is kept. The counter is incremented each time the counter value of the skew mark exceeds the threshold. If the counter has not exceeded a predetermined value, the paye is not skewed. The top and bottom recognition bands have another control mark to specifify that the skew determination is to be skipped. These bands have marks which differ from those of the user and machine control bands thereby allowing detection without the need for skew determination.

The program (FIG. 8D) begins with the IF statement (CSR00180). If-the hardware counter indicates the skew check is to be bypassed, the program indicates a skew is acceptable (CSR00182) and the program goes back to the main program in FIG. 8A~ In the event that the skew check is not to be bypassed, the program performs a skew check defined by program steps CSR00183-CSR00190.

The other subroutine in the main program shown in FIG. 8~ is the convert counter subroutine identified by program step CSR00059. The convert counter subroutine is given in FIG. 8C. The convert counter - subroutine converts the total counters into a bit stream that corresponds to the marked and unmarked posi-tions on the lines or rows of marks on the control sheet 10 (FIGS. 3 and 4). The counters are ~L3034 BO9~1009 30 examined one at a time in the conversion process.
The conversion process consists of comparing a counter value against two thresholded values; a minimum thresholded value and a maximum thresholded value. The counter that falls within the prescribed check values results in a l-blt into stream. If the counter does not fall ~ithin the prescribed check value, the counter is converted to a O-bit.

The first instruction in this program is identified by CSR00118. In this instruction, the program sets the top threshold value that, if exceeded, indicates something other than the mark was seen. The instruc-tion CSR00120 sets the low threshold value, that if not met indicates something other than the mark was lS seen. Once the conditions are set up by the program, the first instruction is a DO-FOR instruction identi-fied by numeral CSR00122. This DO-FOR instruction is applied to all marks beginning from the ~irst mark right of the recognition mark for a particular line until all the marks i~ the line are accounted for. The subroutine completes with the instruction CSR00129. In other words, for each mark, the tally count outputted from the mark detection hardware logic means 76 is tested against the previously set 2~ counter to see whether a 1 or a O is embedded in the bit stream.

FIG. 8E shows the check whether control band subrou-tine. This subroutinè is identified in the main program by numeral CSR00061. The code in this subroutine compares the results of the hardware-~microcode PEL counts to determine whether the row of counts represent a control band. I~ the row is indeed a control band, the bit string representation is moved to the control band array ~or later proces-sing. If the row is n,t a control band, it is noved .

. .
.:

~13~34 to the bit string to be shipped out at a later time.The program begins with an IF s-tatement identified by numeral CSR00209 and ends with an ~LSE statement identified by numeral CSR00212.

The "check control sheet subroutine" is identified in the main program by CSR00063. The check control sheet subroutine is identified in FIG. 8F. The function of this subroutine is to determine whether or not the control band just encountered is a top or - 10 bottom recognition band. If the band is found to be a top recognition band, the sheet is accepted. If the band is a bottom band, the sheet was scanned bottom to top (that is upside down) and is rejected.
The first step in the program is an IF statement which is identified by CSR00229. The instruction identified by CSR00230 is a THEM-DO statement. The TH~N-DO instruction is processed if a control sheet decision has not yet been made. The instruction is terminated by the END-THEN statement CSR00250.
Likewise, the DO-FOR statement CSR00232 terminates at the END-FOR statement CSR00242. The THEN DO
statement CSR00234 is terminated at the END-THEN
statement CSR00241. The THE~l-DO statement CSR00236 is terminated at END-THEN staternent CSR00239.
Finally the ELSE statement CSR00240 is terminated with the END-ELSE statement CSR00248.

FIG. 8B shows the subroutine for the "sum tally counts." This subroutine is identified in the main program by numeral CSR00071. This subroutine adds the tally counts of the control sheet recognition - hardware device to the total counters. These counters are the count of P~Ls within a grid placed over a marked area. One counter is designated for each markable location across the paper. These counters are sùmmed during the up-time. These counters are 3~)3~

summed for the scan lines when a reference mark is seen. Once the bottom of a mark is seen, which is when a no reference found is received, the counters are converted to bit strin~s of ones and zeros representing those positions that are marked and unmarked. The program begins with DO-FOR statement CSR00096 and ends with an END-FOR statement CSR00098.
Althou~h a pariicular set of software programs have been described for processin~ the marks on the control sheet, it should be noted that it is within the skill of the art to generate other programs without departing from the scope and spirit of the present invention. -As stated before, network information relative t~
document distribution parameters are encoded in thescanner control and transmitted to a network node for further processing. At the network node, a plurality of software programs translate the encoded bit stream into network dis~ribution parameters.
The software programs will now be described.

3. The Host Control Sheet SupPort Algorithm Referring now to FIG. ~, a detail block diagram of a network distribution system is shown. The system includes a conventional optical scanner 178. The function of the scanner is to scan control sheets and documents which are to be transmitted to one or more recipient nodes. The output from the scanner is transmitted over a communication link 180 to a first host node 182. The communication link 180 may include a communication controller (CC) such as a 3705. The output from the 3705 is fed into a commun-ication control modem 186. The output from communica-tion modem 186 is fed over communication channel 188. The communication channel 188 may be telephone lines, satellite, microwaves, etc.

~ ` ~

B0981009 33 ~3~34 Communication modem 190 and communication controller 192 are connected in series. Their function is to accept the signal from the communication channel 18 and deliver it to node 182. The host node 182 usually includes a central processing system such as a conventional computer. The function of the computer at host node 182 is to accept the transmitted data and control information from the control sheet, process the information, retransmit it over communica-tion chanllel 196 -to intermediate node 1'3~. Both intermediate node 198 and host node 194 are fitted with auxiliary storage devices 200 and 202, respective-ly. By way of example, these auxiliary storage devices can be direct access drives, etc. From intermediate node 198, the data is transmitted to a destination node 204. Auxiliary storage facility 206 is coupled to no~e 204. The output from destlna-tion node 204 is coupled over communication link 208 to recipient node 210. The communication link 208 incorporates elements substantially equivalent to the element described in communication link 180. As such, the details of these components will not be enumerated. In the preferred embodiment of this invention, recipient nod~ 210 is a facsimile printer.
Except for the control sheet software programs residing in the nodes of FIG. 9, the elements in FIG. 9 are substantially equivalent to the elements in a conventional IBM S/370 or equivalent system.
The IBM S~370 is described in the above-referenced article by Gary D. Schultz. ~he article may be consulted for detailed description of the system.
The teaching in those systems and the article are incorporated herein by reference and details ~rill not be given.

Turning now to FIG. 7, the programming modules or components which reside in node 194, 198 or 204 are shown. The node includes a safe stora~e services ~Z~ 3~

212. The function of the safe s'corage services is to store data. To this end, the safe storage services is a conventional buffer having con~Jentional control-lers for directing the flow of data into and out of the buffer. The terminal document distrihution services 21~ is coupled over conductor 216 to the safe storage services. The output from terminal document distribution services 214 is coupled over communication channel 217 to an originating or recipient node terminal 218. In the preferred embodiment of this invention, the originating node is a conventional optical scanner, while the recipient node is a facsimile printer. The terminal document distribution services includes a plurality of software programs adapted to process requests coming from an originating recipient node such as a scanner involving a covèr sheet. The terminal document distribution services perform the following functions:

Uses the safe storage facility to store trans-mitted documents and bit map on permanent storage.

Interprets the bit map sufficiently to determine that it is indeed a document distribu~ion request.

Transforms the distribution control marks into equivalent distr~bution control parameters using a translation map for distribution requests.

Uses a form number indicated by a portion of the bit map that represents a specific address mark translation map.

Retrieves the translation map and uses it to convert address band marks to network address 1~3L3~3~

of the document distribution node involving intended recipient.

Replaces the bit-encoded document distribution requests header with the actual document distri-bution request header.

Queues the document distribution request for processing by the store-and-forward process.
From here on the request header can be handled as if it had come from a coded information terminal such as a keyboard.

As indicated by controls, or as explicity requested via separate cover sheets or coded information (CI), terminal requests, the terminal document distribution services also establishes a communicating session with each addressed recipient terminal, transmits the terminal print controls originally indicated via marks by the sender and transmits the document contents to the recipient node for printing.

20 ` The store and forward process component 220 is coupled to the safe storage services 212 through conductor 222. The unction of the store and forward process component 220 is to process docume~t routing and transmission algorithms which are governed by distribution controls originally indicated via marks on the scanned cover control sheet. The store and forward process component performs the following functions:

Retrieves the document distribution re~uest header and list of network recipient user identification and document distribution nodes - ~Z~3~3~
BO9~lO09 36 Uses the distribution controls to indicate mode of transmission (example: immediate, delayed, etc.) distribution acknowledgement requests, etc.

Uses the recipient network location address to sort by unique destination node identification and attache-s the specific list of recipient identification for each destination identifica-tion.

- 10 Queues up a request for each adjacent intermediate or destination node.

The internode document distribution services 224 is coupled through conductor 226 to the safe storage services. The intermediate node document distribution services is coupled through communication link 228 to an adjacent intermediate or destination host node. It is worthwhile noting that if an intermediate node is not present in the system, the data moves from the originating node to the document distribution node and then to the recipient node. In the event that the system includes intermediate nodes, the document and controlled information is forwarded through the intermediate node to the destination nodè. To this end, the internode document distribution services 224 performs different functions. The function depends upon whether or not it is positioned at an intermediate node such as node 198 (FIG. 9) or in a destination node such as 204 (EIG. 9). At the intermediate node, the internode document distribution serv~ces performs the following functions:

Retrieves a document from safe storage via the safe strrage facility.

`
.

.

~L~13~34 ~0981009 37 Transmits a single copy of the document data contents and the destination node's distribution request list to each adjacen~ intermediate or destination node using transmission scheduling factors governed by distribution controls.

When its own node is acting as the destination node, queues requests for processing by the terminal document distribution services component.

At each destination's nodes, the internode document distribution services performs the following functlons:

The internode document distribution services component receives the document request and uses the safe storage services facility to store its contents.

The store-and-forward process interprets the request to determine that there is one or more recipients associated with one or more recipient nodes ~terminal) at-tached to this node.

The generation of digital computer programs to perform the above-enunciated functins or steps are well within the skill of an artisan skilled in the art,of computer programming. This being the case, the details of such a program will not be given.
Suffice it tQ say that the program which implements the above factions encodes the bit streams carrying document distribution information into a form which càn be recognized by a conventional store and forward document distribution routine or a conventional -document interchange architecture routine~ These routines distribute the document in accordance with the transmitted instructions. By way of example, an - article in the IBM System Journal entitled "Electronic .

.

Bo981009 38 ~Z~3Q3~
Information Interchange in an Office Environment" by M. ~. DeSousa ~pgs. 4-22) describes an architecture which can be used to distribute the instructions.
The article is incorporated herein by reference.

OPERATION

In operation, documents to be transmitted over the facsimile network are first preceded by a document control sheet. The document control sheet contains information which controls the input scanner, the network node/nodes and the facsimile printer which reproduces the transmitted documents. In the event - that other types of control sheets are needed, those sheets follow the document distribution sheet. In essence, the order in which documents are processed, is as follows: the document control sheet ~hich is the first sheet for each job is scanned by the scanner. If a continuation or other type of sheet is needed, that sheet is next scanned. The actual documents are next scanned by the scanner. The - 20 continuation sheet is an example of other types of control sheets. The control information on the control sheet is prepared by an operator in accordance with desired scanning, distribution and printing requirements.

The operator then places the job control sheet on the document glass of the scanner. The scanner and the scanner control mechanism formulates bits of digital information which represent the controlled instxuctions on the job control sheet. The bits representing instructions are buffered in a conven-tional buffer which is associated with the scanner.
The digital data in the buffer is next accessed by the mark detection hardware logic means and the - cover sheet microcode algorithm means. These means _ 3~3~

identify the instructions which are designated to control the scanner, the network nodes and the facsimile output printer.

The bit streams representing information for control-ling the scanner and the printer are formulated into parameters for controlling these devices. The - parameters for controlling the printers are extracted and are routed to the conventional scanner control means. These conventional scanner control means are part of the scanner mechanism and, as such, details are not given. ~he scanner control parameters are used by the scanner control to control the scanner in accordance with the instructions. Similarly, the parameters for controlling the printers are transmitted lS over the network where it~is routed to the respective printers. At the printers, the conventional printer controls utilize the transmitted parameters for reproducing the document. Because the printer control information is processed (that is transformed into printer sontrol parameters) at the scanning unit, there is no need to incorporate processing circuitry in the printer. It should be noted that such circ~litry may be incorporated without departing from the scope of the present invention.

~he information for controlling document distribution is encoded into bit stream tags with identification numbers and transmitted over the transmission network.
At the network node, the host control sheet support al~orithm accesses the bit streams and converts the 30- bit streams into network parameters. These network parameters are used by the host's computer in the network for distributing and/or routing the data.

While the invention has been particularly shown and described with reference to a preferred embodiment .

B0981009 40 ~3~34 thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.

Claims (4)

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. The combination comprising:
a plurality of facsimile machines for transmitting and receiving signals representative of documents;
a plurality of node means for providing communications of signals from one of said plurality of facsimile machines to a different one of said plurality of facsimile machines, each of said node means including means for coupling each said plurality of node means to at least one of said plurality of facsimile machines, means for coupling each said plurality of node means to at least one other one of said plurality of node means, switching means for routing signals from a coupled one of said plurality of facsimile machines to a coupled other one of said plurality of node means, from a coupled other one of said plurality of node means to a different coupled other one of said plurality of node means, and from a coupled other one of said plurality of node means to a coupled one of said plurality of facsimile machines according to routing command signals; and control document means capable of being read by said facsimile machines for providing control information including said routing command signals.

2. The combination claimed in claim 1 wherein each of said node means further includes control means for controlling a coupled one of said plurality of facsimile machines according to machine control signals and the control information provided by said control document means also includes machine control signals.

3. A method for controlling the interconnection of a plurality of facsimile machines in a communications network wherein each facsimile machine is selectively connectable to every other facsimile machine of said plurality through a series of n switching nodes in response to routing command signals, where n is an integer greater than one, comprising the steps of:

reading a control document by an originating facsimile machine whereby control information from said control document is converted to routing command signals;
communicating routing command signals to a first switching node;
connecting said originating facsimile machine to at least a second switching node through said first switching node according to the communicated routing command signals;
transferring from said first switching node to said second switching node those routing command signals for controlling said second switching nodes and any subsequent switching node in said series;
repeating as necessary said connecting and transferring steps to connect an i-th switching node to an (i+l)-th switching node according to the transferred routing command signals and to transfer those routing command signals for controlling the (i+l)-th switching node and any subsequent switching node in said series, where i is an integer having successive values from 2 to n-l, inclusive; and connecting the nth switching node to a destination facsimile machine according to the transferred routing command signals.

4. The method claimed in claim 3 including the further step of transmitting from said originating facsimile machine to said destination facsimile machine a number of documents controlled by said first switching node according to said control information from said control document.