Data Format for the Interchange of Fingerprint Information

Also Known As: 1-1993 Fingerprint Format


Type Bitmap
Colors 1- and 8-bit
Compression None, ANSI/EIA 538 (CCITT Group 4), and others
Maximum Image Size 64Kx64K
Multiple Images Per File Yes
Numerical Format Big-endian
Originator NIST-CSL
Platform All
Supporting Applications Many
Specification on CD No
Code on CD No
Images on CD No
See Also FBI Fingerprint Format

Usage
Use to interchange fingerprint data between agencies and to process transactions involving fingerprint data.

Comments
1-1993 actually defines a data stream to be used for defining a transmission protocol for processing and exchanging fingerprint data.


Data Format for the Interchange of Fingerprint Information is a standardized data format for the exchange of human fingerprint information in the form of electronic images. Storing fingerprints as digital information greatly enhances the ability of law enforcement agencies to compare and exchange fingerprint data.

Contents:
File Organization
File Details
For Further Information

Fingerprints are typically stored as inked impressions on a ten-print card. The FBI has over 30 million of these cards currently on file. Each day the FBI receives 40,000 requests for searches and 14,000 updates for these cards. Searches and updates are mostly conducted manually by examiners. In 1995, only 37,000 search a day were processed, leaving a backlog of over one million searches uncompleted. The number of search and update requests are expected to double by the end of the century.

Digital fingerprint information is typically collected using an Automated Fingerprint Identification System (AFIS). Fingerprints may be scanned from ten-print cards, or be taken directly from a subject's fingers using a live-scan reader. The fingerprints are stored as digital images using the 1-1993 data format. The 1-1993 files are then archived in an Image Storage and Retrieval (ISR) system, where they are searched for possible matches. They are also routinely exchanged between law enforcement and security agencies, and any organizations participating in the National Fingerprint File (NFF) program.

When a search is performed, the AFIS system compares a set of one or more fingerprints against its master database of fingerprint data. A selection of "best fit" matches are then presented to a human examiner for the final identification. Using AIFS, tens of thousands of searches may be performed in a day, with most search results being returned to the requester in 2 to 24 hours. It is expected that by the year 2000 75% of all fingerprint transactions will be conducted electronically rather than by mail.

The 1-1993 specification does not specifically define a name for this file format. A proper name such as "Fingerprint Interchange Format" is never given. This article therefore refers to files created using the 1-1993 specification as "1-1993 files".

File Organization

Data in 1-1993 files is stored as a series of logical records. A 1-1993 file may contain up to nine different types of logical records (Table Fingerprint-1). Type-1, Type-2, and Type-9 records contain only 7-bit ASCII data. Type-3 through Type-8 records contain only binary data. All binary data is stored most-significant bit first and most-significant byte first. There is never any padding to align data on physical boundaries. And the fingerprints of only one individual subject may be stored in a 1-1993 file.

Table Fingerprint-1: Logical Record Types
Record Type ID Logical Record Data Max Number Per File
1 Transaction Information 1
2 User-defined Text Unlimited
3 Low-res grayscale Image Data 14
4 High-res grayscale Image Data 14
5 Low-res bilevel Image Data 14
6 High-res bilevel Image Data 14
7 User-defined Image Data Unlimited
8 Signature Image Data 2
9 Minutiae Data 10

1-1993 files are also used to initiate transactions and respond to queries for data processing and information retrieval. Eight types of transactions are defined by the 1-1993 specification. Table Fingerprint-2 lists these transaction and the type and quantity of logical records that may appear in the 1-1993 files.

Table Fingerprint-2: 1-1993 File Transactions
Record Ten-print Latent File Image Search Image Request
Type Inquiry Inquiry Maintenance Request Response Response
1 1 1 1 1 1 1
2 1 0-1 0-1 1 1 1
3 0-14 0 0-14 0 0-14 0-14
4 0-14 0-10 0-14 0 0-14 0-14
5 0-14 0 0-14 0 0-14 0-14
6 0-14 0-10 0-14 0 0-14 0-14
7 0 0-N 0-N 0 0-N 0-N
8 0-2 0 0-2 0 0-2 0-2
9 0-10 0-10 0-10 0 0 0

A ten-print inquiry attempts to match a set of fingerprints of a known subject with any prints stored in a data base. The inquiry will contain some or all of the subject's fingerprints, a text comment identifying the inquiry, and possibly handwritten signature and minutiae data as well.

A latent inquiry is an attempt to match fingerprints of an unknown subject to one or more records stored in the master database. The transaction contains only high-resolution images records and possibly user-defined images and minutiae data. No low-resolution images or signature data is present.

A search response is the reply to a ten-print or latent inquiry. The response may contain a partial or complete record of the fingerprints, or an indication that a match was not found.

An image request is used by an agency to specify one or more ten-print files that it wishes to receive. No image data is present in the request. The requested fingerprint data is returned in the form of an image request response.

File maintenance is used to send an update to a fingerprint database. Any type of record information may be found in a maintenance update.

You will notice the wide latitude that 1-1993 gives to the content of these transactions. In part this is because not every fingerprint file will contain a complete set of prints and information. Also, different agencies will have differing requirements for the information that must, or must not, be present in each of these types of transactions.

File Details

This section describes various types of the 1-1993 format.

ASCII Record Format

Type-1, Type-2, and Type-9 records contain only ASCII data. The basic syntax of these records is as follows:

    <record type>.<field number>:<field
data><GS>

Each record field begins with the logical record type, followed by a period, the field number, a colon, the field data, and a group separator character. If the field is the last in the record, then the group separator is replaced by a field separator character.

Information in these records is delineated by the four separator characters defined in the ASCII character set: File Separator (FS, ASCII 1Ch), Group Separator (GS, ASCII 1Dh), Record Separator(RS, ASCII 1Eh), and Unit Separator (US, ASCII 1Fh). Only one separator characters may appear between two data items and may never adjacent to another separator. A segment of a 1-1993 data stream would look something like this:

Type-1 Transaction Record

The first record in every 1-1993 file is a Type 1 record. This record stores basic interchange information that you would expect to see in a file header. Type-1 record field are composed entirely of ASCII data, several of which are considered optional and may not appear in a 1-1993 file if the creating agency decides that they are not needed.

Each Type-1 record contains a File Content field. This field is an index of all the logical records in the file and stored in the order that they appear in the file. Each 1-1993 logical record contains an Image Designation Character (IDC) field which holds a numeric value that identifies the logical record. The IDC may be any positive integer and typically start at zero and increment upward. A 1-1993 file reader can check the contents of a 1-1993 file by looking only at the File Content field.

Each Type-1 field is separated by a GS character. The last field is followed by an FS character. All fields are mandatory to appear in the Type-1 record unless otherwise noted. Type-1 record fields are laid out as follows:
Logical Record Length
Version Number
File Content
Type Of Transaction
Date
Priority
Destination Agency Identifier
Originating Agency Identifier
Transaction Control Number
Transaction Control Reference
Native Scanning Resolution
Nominal Transmitting Resolution

Logical Record Length contains the total size of the record described as an ASCII string. For example, if the total number of bytes in all fields were 241 bytes, this field would contain the string "241". This field is always begins with the field number "1.01:".

Version Number is a four-byte field specifying the version of the standard used by the software creating the 1-1993 file. The first two characters specify the major version and the last two specify the minor version. The initial version of this specification is "0000"; the ANSI/NBS-ICST 1-1986 specification is "0100", and the ANSI/NIST 1-1993 specification is "0200".

File Content is an index of all the logical records in the 1-1993 file and the order in which they appear. This field contain one subfield for each logical record in the 1-1993 file.

The first subfield contains information on the Type-1 record. The first information item is the record type ("1"). The second information item is the number of records that follow the Type-1 record. This number is also the remaining number of subfields in this field. All remaining subfields store the record type and the assigned IDC value for that record. Here's an example of this field for a 1-1993 files with 5 records:

  1.03:1<US>4<RS>2<US>00<RS>4<US>01<RS>4<US>02<RS>8<US>03<GS>

Type Of Transaction contains information on the type of processing required by the file data. The format of the data in this field must conform to the requirements of the agency that is receiving the file.

Date is the date that the transaction was initiated (file was created or transmitted). The date appears in the format of YYYYMMDD representing the year, month, and day of the date. For example, the date 12 May 1996 would be stored as "19960512".

Priority contains a single character whose value specifies the urgency in which a response is required. The value many be "1", "2", "3", or "4", with "1" denoting the highest priority. This field is optional, and if not present in a Type-1 record a priority of "4" is assumed.

Destination Agency Identifier contains information that identifies the person, group, or organization that is to receive this file. The format of the data in this field must conform to the requirements of the agency that is receiving the file.

Originating Agency Identifier contains information that identifies the person, group, or organization that created and is sending this file. The format of the data in this field must conform to the requirements of the agency that is receiving the file.

Transaction Control Number is a unique number assigned by the originating agency to this file for purposes of identification. Any responses by the destination agency to the originating agency will refer to a specific 1-1993 file using this number.

Transaction Control Reference is an optional field used when responding to a transaction (that is, the reception and processing of a 1-1993 file). This field will contain the Transaction Control Number of the 1-1993 information that the response is in reference to.

Native Scanning Resolution is a five-byte field used to specify the scanning resolution of the device used to capture the image data. It may also be used to specify the desired resolution of data that is being requested. The resolution is specified in pixels per millimeter and stored as two numeric characters, a period, and two more numeric characters. For example, a scanning resolution of 20.47 p/mm would be stored as "20.47".

Nominal Transmitting Resolution specifies the actual resolution of the image stored in the 1-1993 file. The resolution is specified in pixels per millimeter and stored as two numeric characters, a period, and two more numeric characters. For example, a scanning resolution of 19.69 p/mm would be stored as "19.69".

Type-2 User-Defined Text Records

Type-2 records are used to store agency-specific ASCII data. The data may be in any format, including 1-1993 field syntax or free-form ASCII text. Type-2 records follow the same field syntax as Type-1 records, except that the field numbers are three characters in length rather than two. Type-2 record fields are laid out as follows:

Logical Record Length
Image Designation Character
User-defined Data

Logical Record Length contains the total size of the record described as an ASCII string. For example, if the total number of bytes in all three fields of a Type-2 record were 238 bytes, this field would contain the string "238". This field is always begins with the field number "2.001:".

Image Designation Character is the numeric value used to associate this record with the corresponding value in the File Content field of the Type-1 record. This field always begins with the field number "2.002:".

User-defined Data contains one or more fields of ASCII data, the format of which is specified by the agency defining the Type-2 record. All user-defined fields will typically begin with field numbers in the range of "2.003:" to "2.999:", although this is not mandatory. Each agency defining the fields in a Type-2 record is responsible for publishing the format of these fields for use by other agencies.

An example of a Type-2 record that contains a total of 56 bytes of data is as follows(<GS> and <FS> represent single ASCII characters):

    2.001:56<GS>2.002:00<GS>2.003:Joe D. Eyeball Detective
Agency<FS>

Type-3, Type-4, Type-5, and Type-6 Image Records

Type-3, Type-4, Type-5, and Type-6 records (referred to as bitmap data records) store image data of finger, thumb, and palm pints. Type-3 and Type-4 records store low- and high-resolution grayscale data respectively; Type-5 and Type-6 records store low- and high-resolution bilevel data respectively.

All of the bitmap data records have the same structure. All fields defined for bitmap data records contain only binary data and all fields must be present in each record.

typedef struct _Type3456Records
{
    DWORD LogicalRecordLen;   /* Total size of record in bytes */
    BYTE  ImageDesignChar;    /* Image Designation Character */
    BYTE  ImpressionType;     /* Image subject collection method */
    BYTE  FingerPosition[6];  /* Finger code */
    BYTE  ImageScanRes;       /* Image Scanning Resolution */
    WORD  HorzLineLen;        /* Image Width */
    WORD  VertLineLen;        /* Image Height */
    BYTE  Compression;        /* Type of compression algorithm used */
    BYTE  ImageData[];        /* Image data */
} TYPE3456RECORDS;

LogicalRecordLen is the total length of this record in bytes, including all fields and the image data.

ImageDesignChar is a synchronization value that identifies this record as belonging in a group to all records with the same designation value. This value is always a binary representation of the Image Designation Character for this record found in the file content field of the Type-1 record.

ImpressionType specifies how the fingerprint image data was collected. This field will contain a value corresponding to the following set of codes:

0 Live-scan plain
1 Live-scan rolled
2 Nonlive-scan plain
3 Nonlive-scan rolled
4 Latent impression
5 Latent tracing
6 Latent photo
7 Latent lift

FingerPosition indicates the finger, or set of fingers, that is the subject of the image. This field is actually an array of six byte values. The first byte contains the known or "most probable" finger position. The remaining five bytes may also contain alternate finger position values (presumably in order from "most probable" to "least probable"), or contain the value FFh indicating that the byte is unused. Valid values for each of the six bytes in this field are as follows:

0 Unknown finger
1 Right thumb
2 Right index finger
3 Right middle finger
4 Right ring finger
5 Right little finger
6 Left thumb
7 Left index finger
8 Left middle finger
9 Left ring finger
10 Left little finger
11 Plain right thumb
12 Plain left thumb
13 Plain right four finger
14 Plain left four finger
255 Unused

ImageScanRes will contain a value of 0 if half the minimum scanning resolution is used and a value of 1 if half the native scanning resolution is used.

HorzLineLen indicates the number of pixels in a single scanline of the image data.

VertLineLen indicates the number of scan lines in the image data.

Compression indicates the type of encoding algorithm used to compress the image data in this record. For all record types a value of 0 indicates that the data is uncompressed. For bilevel images (Type-5 and Type-6 records) a value of 1 indicates the image data is compressed using the ANSI/EIA 538 facsimile algorithm. For grayscale images (Type-3 and Type-4 records) a non-zero value indicates a compression algorithm is used that is registered with the FBI.

Type-7 User-defined Image Record

Type-7 records are used to store user-defined image data. When an agency needs to transmit or store a subject not defined by the 1-1993 specification (palms, soles, toes, and so forth), then image data is stored in a Type-7 record. It is also possible to store a mug shot of the subject in a Type-7 record as well.

Type-7 record data must follow the same criteria as Type-3 through Type-6 records (that is, bilevel or grayscale image data, 0 is black, 1 or 255 is white, same compression algorithms, and so forth). The format of the data in the Type-7 record is similar to the other image data records, but a section of user-defined fields is left to be determined by the agency defining the Type-7 record:

typedef struct _Type7Record
{
    DWORD LogicalRecordLen;   /* Total size of record in bytes */
    BYTE  ImageDesignChar;    /* Image Designation Character */
    BYTE  UserData[];         /* User-defined fields */
    BYTE  ImageData[];        /* User-defined image data */
} TYPE7RECORD;

LogicalRecordLen is the total length of this record in bytes, including all fields and the image data.

ImageDesignChar is a synchronization value that identifies this record as belonging in a group to all records with the same designation value. This value is always a binary representation of the Image Designation Character for this records found in the file content field of the Type-1 record.

UserData is zero or more user-defined fields used to interpret the data in the ImageData field. Each agency defining a Type-7 record is responsible for publishing the format of these fields for use by other agencies.

ImageData is the image data stored in this record.

Type-8 Signature Record

Type-8 records store a bilevel bitmap image or a vector drawing of a handwritten signature. This signature may be from either the person whose fingerprints are stored in the FIF file, or from the official who created the original fingerprint card or file. There may be up to two Type-8 records per FIF file, each with the following format:

typedef struct _Type8Record
{
    DWORD LogicalRecordLen;   /* Total size of record in bytes */
    BYTE  ImageDesignChar;    /* Image Designation Character */
    BYTE  SignatureType;      /* Originator of the signature */
    BYTE  SignatureRepType;   /* Signature data type */
    BYTE  ImageScanRes;       /* Image Scanning Resolution */
    WORD  HorzLineLen;        /* Image Width */
    WORD  VertLineLen;        /* Image Height */
    BYTE  ImageData[];        /* Subject Signature Image Data */
} TYPE8RECORD;

LogicalRecordLen is the total length of this record in bytes, including all fields and the image data.

ImageDesignChar is a synchronization value that identifies this record as belonging in a group to all records with the same designation value. This value is always a binary representation of the Image Designation Character for this record found in the file content field of the Type-1 record.

SignatureType indicates the origin of the handwritten signature. A value of 00h indicates it is the signature of the subject whose fingerprints are stored in the FIF file. A value of 01h indicates the signature is of the official who recorded the finger prints. If two Type-8 record exist in the same FIF file they will each store a different signature type.

SignatureRepType specifies the form in which the data of the signature is stored. A value of 00h indicated a scanned, uncompressed bitmap; a value of 01h indicates a scanned, compressed bitmap, and a value of 02h indicates a vector drawing.

ImageScanRes will contain a value of 0 if half the minimum scanning resolution is used or a value of 1 if half the native scanning resolution is used.

HorzLineLen indicates the number of pixels in a single scanline of the image data. For vector data this field is always 0000h.

VertLineLen indicates the number of scan lines in the image data. For vector data this field is always 0000h.

ImageData is the image data stored in this record.

The signature image data may be a bilevel bitmap packed eight bytes per pixel or vector data. Bitmap signatures may be uncompressed, or compressed using the ANSI/EIA 538 algorithm. Vector data is a series of five-byte values describing the signature as collection of points connected by line segments. Each point has the following data format:

typedef struct _Type8RecordVector
{
    WORD XCoord;    /* X location of point */
    WORD YCoord;    /* Y location of point */
    BYTE Pressure;  /* Pen pressure */
} TYPE8RECORDVECTOR;

XCoord and YCoord are the X and Y position of the point respectively. Coordinates are expressed in units of 0.0254mm (0.001 inches) and are referenced from the origin at the bottom-left corner of the image. Positive values of XCoord move from left-to-right; positive values of YCoord move from bottom-to-top.

Pressure is the relative thickness or intensity of the current line segment. The current segment is the line drawn from the point defined by the XCoord and YCoord values of this point to the next point). A value of 00h indicates no pressure; a value of 01h is the lightest possible pressure; a value of FEh is the heaviest possible pressure, and a value of FFh indicates the final point in the vector data.

Type-9 Minutiae Record

A Minutiae record describes the minutiae data and other details related to a single finger. Such descriptions can greatly decrease the amount of time to find a positive match when searching thousands or millions of records. Up to ten Type-9 records may be stored in a single 1-1993 file.

Logical Record Length
Image Designation Character
Impression Type
Minutiae Format
Originating Fingerprint
Reading System
Finger Position
Finger Pattern Classification
Core Position
Deltas Position
Number of Minutiae
Minutiae Ridge Count Indicator
Minutiae and Ridge Count Data

Logical Record Length contains the total size of the record described as an ASCII string. For example, if the total number of bytes in all fields were 91 bytes, this field would contain the string "91". This field always begins with the field number "9.01:".

Image Designation Character is the numeric value used to associate this record with the corresponding value in the File Content field of the Type-1 record. This field always begins with the field number "9.02:".

ImpressionType specifies how the fingerprint image data was collected. This field stores a numeric value in the same range as the binary value stored in the Impression Type field of Type-3 though Type-6 records. This field always begins with the field number "9.03:".

Minutiae Format indicates if the remainder of the Type-9 record complies to the 1-1993 format standard for Type-9 records or is user-defined. This field will contain the value "S" if it is standard, or "U" if it is user-defined. This field always begins with the field number "9.04:".

Originating Fingerprint Reading System contains two or three subfields. The first is the name or designation of the system that created this record. The second is the method by which the minutiae data was read, encoded, and recorded. The following are the valid values for the second subfield:
"A" Automatic data collection; no human editing possible
"U" Automatic data collection; human editing possible, but not performed
"E" Automatic data collection; edited before encoding and recording
"M" Manual data collection

The third subfield is optional and contains a two-character, user-defined value that identifies the equipment used to create the data.

Finger Position is the identification of the finger from which this minutiae information is derived. If the exact finger position is not know, multiple guesses may be entered in this field separated by RS characters. The following values may appear in this field:

"0" Unknown finger
"1" Right thumb
"2" Right index finger
"3" Right middle finger
"4" Right ring finger
"5" Right little finger
"6" Left thumb
"7" Left index finger
"8" Left middle finger
"9" Left ring finger
"10" Left little finger

Finger Pattern Classification specifies the type of fingerprint pattern that contains the minutiae. This field contains two subfields separated by RS characters. The first subfield contains a single character that indicates if the pattern classification code is defined by the 1-1993 standard ("S"), or is user-defined ("U"). The second subfield contains the pattern classification code. These codes as defined by 1-1993 are:

"PA" Plain Arch
"TA" Tented Arch
"RL" Radial Loop
"UL" Ulnar Loop
"PW" Plain Whorl
"CP" Central Pocket Loop
"DL" Double Loop
"AW" Accidental Whorl
"WN" Whorl (type not designated)
"RS" Right Slant Loop
"LS" Left Slant Loop
"SR" Scar
"XX" Amputation
"UN" Unknown or Unclassifiable

Core Position is an 8-character field that contains the X and Y coordinates of the fingerprint core. Data in this field is stored using the XXXXYYYY format. For example, a core location of 348,12 would be stored as "03840012". This optional field always begins with the field number "9.08:".

Deltas Position stores the X and Y location of each delta in the fingerprint. Each location is stored using the XXXXYYYY format. Multiple locations are separated using the RS character. This optional field always begins with the field number "9.09:".

Number of Minutiae is a single-character field that stores the number of minutiae recorded in the fingerprint. This field always begins with the field number "9.10:".

Minutiae Ridge Count Indicator specifies if minutiae ridge count information is present ("1") or not present ("0") in the file. This field always begins with the field number "9.11:".

Minutiae and Ridge Count Data contains all of the minutiae and ridge count data for this record. This field will contain as many subfields as indicated by the Number of Minutiae field value. Each subfield contains information for one minutiae, is separated from other subfields by the RS character, and has the following format:

Index Number
X, Y, and Theta Values
Quality Measure
Minutiae Type Designation
Ridge Count Data

Index Number is a unique number assigned to each minutiae subfield. This index begins with "0" and increments upwards.

X, Y, and Theta Values are the coordinate vales of the minutiae data. The X and Y values are 4-digit, positive numerical values. Theta is a 3-digit value in the range of "000" to "359". The three values are stored as a single 11-digit value using the format XXXXYYYYTTT. If the minutiae is of Type C or D, the Theta value will be "000".

Quality Measure is an optional field that indicates the quality of the minutiae data. A value of "0" indicates manually encoded minutiae. A value of "1" indicates no method of quality measure is available. Values of "2" to "63" indicate the confidence level of the accuracy of the minutia data. "2" is the highest confidence level.

Minutiae Type Designation is a optional single character that indicated the type of minutiae. The following minutiae types are valid:
"A" Ridge Ending
"B" Bifurcation
"C" Compound (trifurcation or crossover)
"D" Type Undetermined

Ridge Count Data the optional ridge count data. This field contains a series of subfields consisting of a minutiae number and a ridge count.

For more information on the encoding and storage of minutiae data in Type-9 records, refer to the ANSI 1-1993 standard document.

Image Data

The fingerprint image data stored in a FIF file is a bitmap of 1-bit bilevel or 8-bit grayscale pixels. Grayscale images are stored one pixel per byte with a value of 00h indicating black (least intense) and a value of FFh indicating white (most intense). Bilevel images are packed eight pixels per byte with a bit value of 0 indicating white and a bit value of 1 indicating black. Bilevel pixels are always left-justified and packed into bytes most-significant bit first.

The data in a bitmap records may be compressed or uncompressed. Bilevel data may only be compressed using the ANSI/EIA 538 (CCITT Group 4) compression algorithm. No compression method for grayscale data is given in the 1-1993 specification. This decision is left to the agencies using the 1-1993 data format. A popular grayscale fingerprint compression method used with 1-1993 is the FBI WSQ algorithm.

The 1-1993 specification makes a distinction between "scanner resolution" and "transmitted resolution" for image data. Scanner resolution is the resolution at which the original fingerprint samples were captured and saved to a 1-1993 file.

Transmitted resolution this the resolution at which the images are transmitted during an electronic transaction. Typically, the scanner and transmitted resolutions are the same. But the resolutions may be different if the images are subsampled, scaled, or interpolated before transmission.

The minimum scanning resolution of all images is 19.69 +/-0.20 pixels per mm (495 to 505 pixels per inch). No maximum is specified by the 1-1993 specification. Scanning resolutions greater than this minimum value and with a tolerance of +/-1% may be used.

The transmitted resolution for high-resolution bilevel and grayscale images must be in the range of 19.69 +/-0.20 p/mm (495 to 505 p/in) minimum to 20.47 +/- 0.20 p/mm (515 to 525 p/in) maximum. For low-resolution bilevel and grayscale images the transmitted resolution is half the resolution, 9.84 +/-0.10 p/mm (247.5 to 252.5 p/in) minimum to 10.24 +/- 0.10 p/mm (257.5 to 262.5 p/in) maximum.

For Further Information

The 1-1993 fingerprint format is officially described in the following ANSI document:

ANSI/NIST-CSL 1-1993, Data Format for the Interchange of Fingerprint Information, November 22, 1993.

You can obtain this specification directly from ANSI:

American National Standards Institute
105-111 South State Street
Hackensack, NJ 07601 USA
Voice: (212) 642-4900
Voice: (212) 302-1286
Email:
WWW: http://www.ansi.org/

The algorithm used by 1-1993 for encoding bilevel image data is also described and available from ANSI:

ANSI/EIA 538-1988, Facsimile coding schemes and coding control functions for group 4 facsimile equipment, 1988.

For more information on Automated Fingerprint Identification Systems (AFIS), have a look on the FBI's home page at http://www.fbi.gov/iafis/iafis.html.

ANSI Copyright

Portions of this material are reproduced from American National Standard ANSI/NIST-CSL 1, Data Format for the Interchange of Fingerprint Information, copyright 1994 by the American National Standards Institute. Copies of this standard may be purchased from the American National Standards Institute, 11 West 42nd Street, New York, NY 10036.



Copyright © 1996, 1994 O'Reilly & Associates, Inc. All Rights Reserved.