Draft Finger Image and Minutiae Data Standard Version 1

7/31/2019 Draft Finger Image and Minutiae Data Standard Version 1

1/25

Draft Document Fingerprint Image and Minutiae Standard----------------------------------------------------------------------------------------------------------------------

Version 1.0 August 2010 Page 1 of 23

Fingerprint Image and Minutiae Data Standardfor

e-Governance Applications in India

Draft Version 1.0

August, 2010

Government of IndiaDepartment of Information Technology

Ministry of Communications and Information TechnologyNew Delhi 110 003


2/25



CONTENTS

1 Introduction ...................................................................................................................................... 41.1 Scope ......................................................................................................................................... 41.2 Objective/Purpose ...................................................................................................................... 51.3 Description ................................................................................................................................. 5

2 Target Audience .............................................................................................................................. 53 Type of Standard Document: ......................................................................................................... 64 Definitions and Acronyms Refer Annexure I ............................................................................... 65 Specification of Fingerprint Image Standard ................................................................................ 6

5.1 Device Specifications ................................................................................................................. 65.2 Image Specifications .................................................................................................................. 7

5.2.1 Impression Type ................................................................................................................. 75.2.1.1 For Enrolment / Identification ...................................................................................... 7

5.2.2 Finger Position .................................................................................................................... 85.2.2.1 For Enrolment / Identification ...................................................................................... 8

5.2.3 Rotation Angle .................................................................................................................... 85.2.4 Number of Fingers .............................................................................................................. 8

5.2.5 Image acquisition format ..................................................................................................... 8

5.2.5.1 For Enrolment / Identification and for Verification in highly sensitive applications ...... 8 5.2.5.2 For Verification in less demanding applications. ......................................................... 9

5.3 Quality Specifications ................................................................................................................. 95.4 Storage Specifications ............................................................................................................... 9

5.4.1 Fingerprint Image Storage Format: .................................................................................... 95.4.1. For Storage / Archival .................................................................................................... 9

5.5 Minutiae data Format Specifications ...................................................................................... 106 Recommended Best Practices ..................................................................................................... 116.1 Best Practices for Implementation of Fingerprint image Standard....................................... 11

6.2 Best Practices for various processes ....................................................................................... 136.2.1 Enrolment Process ............................................................................................................ 136.2.3 Process for capturing rolled fingerprint images ................................................................ 14

6.3 Fingerprint Record Format ....................................................................................................... 147 Annexure ........................................................................................................................................ 18

Annexure 1 ................................................................................................................................... 18Definitions and Acronyms .............................................................................................................. 18


3/25



8 References ..................................................................................................................................... 239 List of Members ............................................................................................................................. 24

9.1 List of Members of Expert Committee ..................................................................................... 249.2 Technical Expert on Biometrics .............................................................................................. 249.4 Other Contributors .................................................................................................................... 259.5 List of Research Associates ..................................................................................................... 25


4/25



1 IntroductionThe Indian government proposes to use fingerprint identification and verification in many e-Governance applications. Fingerprints are proposed as a means of recognition in theseprograms. There are many vendors selling devices that capture fingerprint images. In orderfor all Indian e-Governance applications to read these images, it is necessary to have astandard header that completely specifies the stored images. There are also manyalgorithms developed by vendors that compare fingerprint features to decide on a match.Each algorithm may use/store intermediate fingerprint features that may not be usable byother algorithms for comparing biometric records. To allow the application developer themaximum flexibility to use algorithms from different vendors, the fingerprint image shouldbe stored. From this image, the features for any algorithm may be extracted. To addressinteroperability issues, the fingerprint image must be stored in a specified format. In

addition, the extracted features also need to be in a specified format.

The Government of India would adopt ISO 19794-4:2005(E) ISO 19794-4 as fingerprintImage standard, and ISO 19794-2:2005(E) as Minutiae data format standard. The standardspecifications are tailored to meet specific needs of civilian e-Governance applications inIndian context. Best practices for fingerprint devices selection, enrolment, verification,identification and storage formats of fingerprint images are also recommended for adoptionin Indian context.

1.1 ScopeThis standard specifies fingerprint image specifications, acquisition / storage / transmissionformats, etc. and minutiae storage specifications.. It also includes best practices for

implementation of the standard specifications in different categories of e-Governanceapplications based upon the volume of data, and verification/ accuracy requirements. Thecurrent version of the standard is applicable to all civilian e-governance applications.Certain law enforcement applications may need features like latent fingerprint, which arenot specified in the current version.

The fingerprint image data enrolled (addressed in the current standard) is stored for1:1(verification) and 1: N (Identification) matching. Minutiae extraction can be done fromthese images.

This standard is structured as follows:

Sections 1 to 4 include scope, objectives of this standard and a brief description, targetaudience, type of enforcement category, definitions & acronyms etc. The standardspecifications are mentioned in Section 5. Section 6 includes best practices for variouscategories of e-Governance applications.

Biometric Standards being formulated by the Expert committee under DIT would beapplicable for all e-Governance applications at National level, and the principles forstandards formulation would be governed by Governments Policy on Open Standards.


5/25



This standard is sufficiently broad to cover the requirements of all e-Governanceapplications. The applications may judiciously select the specifications relevant to theirneeds.

1.2 Objective/PurposeThe Indian e-Government applications will deal with Fingerprint data during multiplestages. Some of these stages are given below.:

a. Image acquisition, and its storage in the Enrolment stageb. Image acquisition and storage for off line / on line verification of fingerprint data in

1:1 matching stagec. Image acquisition and storage for the purpose of identification in 1:N matching staged. Transmission of finger image data to/from smart card or for data exchange with

other e-Governance applications.e. Extraction of minutiae from fingerprint images (enrolment or recognition stage), their

storage, and minutiae matching

It is possible that different fingerprint capture devices and software (compressionalgorithms and matching algorithms) are used at different stages as mentioned above.

The purpose of this standard document is to standardize fingerprint devices, fingerprintimage, storage / transmission and minutiae specifications to ensure interoperabilitybetween various fingerprint sensors and algorithms by which the fingerprint images arecaptured / stored.

From the stored fingerprint images in enrolment stage, minutiae would be extracted formatching of minutiae of captured fingerprint images at the verification/identification stage.

1.3 DescriptionA fingerprint is an impression of the friction ridges found on the inner surface of a finger ora thumb. The ridges follow a global pattern identified as whorl, right loop, left loop, arch,tented arch and twin loop etc. Skin pores also present a detailed pattern in fingerprints.There are also local patterns where ridges end or bifurcate, known as minutiae. Localand/or global patterns of fingerprints are matched to provide a means of identification orverification. The science of fingerprint recognition constitutes accurate means of positiveidentification known to humans.

This version of standard addresses the standard specifications for the process of

verification using minutiae data. Pattern features and skin pore features are not addressedin this version of the standard.

2 Target Audience- All E-Governance projects of the central and state Government or any other

organization which need to comply with this standard for the purpose ofinteroperability.

- Vendors for Fingerprint acquisition devices or required software for conversion ofimages as per the standardised format, Quality checking S/W, minutiae extractionand matching algorithms etc.


6/25



- All Integrators/Service providers for Indian e-Governance applications.

3 Type of Standard Document:Standard Specification and Best Practices

Enforcement Category: Mandatory

4 Definitions and AcronymsRefer Annexure I

5 Specification of Fingerprint Image StandardIndian e-Governance Standard will adopt ISO-19794-4 Fingerprint Image Data Standard asIndian Standard. However for this version of the Standard the prescriptive values,

exceptions, deviations and additions, if any are listed in the following five sections asfollows5.1. Device Specifications5.2 Image Specifications5.3 Quality Specification.5.4 Storage Specification5.5 Minutiae Specifications5.6 CBEFF Header Specifications (This is optional and is used for encrypting and

signing of the data)

5.1 Device SpecificationsFor the purpose of acquisition of fingerprint image data at the time of enrolment, verification

or identification, the fingerprint scanner devices need to be used. There is a need tostandardize device specifications to ensure interoperability of fingerprint images of thesame person taken at different stages.

Device specification covers scan resolution, pixel depth and dynamic range. A higherresolution device does not necessarily produce better images1. The biometric samplescaptured during enrolment need to be the best samples possible. The specifications aretailored from the ISO 19794-4:2005(E) standard to meet the civilian e-Governanceapplication requirements.

5.1.1 For Enrolment and Identification

The minimum Aquisition setting level allowed is 31. (Refer Section 7.1, table 1 of ISO19794-4:2005(E).) The minimum pixel depth allowed is 8 bit. Pixel depth of 8 bit and more,up to 16 bit is also allowed. Acquisition setting levels 31 and 41 are allowed

1 It should be noted that two devices with identical scan resolution, pixel depth and dynamicrange do not provide similar quality images. A number of laboratory tests have shown thata 197 dpcm (500 dpi ) device from one vendor performs better than a 394 dpcm (1000 dpi)device of another vendor. Nevertheless, these attributes are the only transparent way tospecify the minimum device requirements for now.


7/25



Settinglevel

Scanresolution

(ppcm)

Scanresolution

(dpi)

Pixel depth(bits)

Dynamic range(gray levels)

31 197 500 8 20041 394 1000 8 200

For best practices refer Section 6.1.1

5.1.2 For VerificationThe minimum Aquisition setting level allowed is 30. Pixel depth of 8 bit and more, up to 16bit, is allowed. Setting level 30 /31 / 41 as shown in Section 7.1, table 1 of ISO 19794-4 areallowed.

Setting

level

Scan

resolution(ppcm)

Scan

resolution(dpi)

Pixel depth

(bits)

Dynamic range

(gray levels)

30 197 500 8 8031 197 500 8 20041 394 1000 8 200

For other levels specifications refer Section 5.1.1


5.2 Image Specifications

5.2.1 Impression Type

5.2.1.1 For Enrolment / IdentificationImpression types having the following codes will only be allowed for Enrolment andIdentification/De-duplication: 0(Live-Scan Plain), 1(Live-Scan Rolled), 2(Non Live-ScanPlain), 3 (Non Live-Scan Rolled) or 9 (Live-Scan Contactless).

5.2.1.2 For VerificationAllowed Impression type codes for Verification are: 0(Live-Scan Plain), 1(Live-ScanRolled), 2(Non Live-Scan Plain), 3 (Non Live-Scan Rolled), 8(Swipe) or 9(Live-ScanContactless).

Refer Section 6.2.1 for best practices

Refer table 7 in Section 8.3.7 of ISO 19794-4:2005(E) for the different impression typecodes.


8/25



5.2.2 Finger Position

5.2.2.1 For Enrolment / Identification

The valid values for finger position are 0 through 10, 13, 14, 15

0 - Unknown Finger1 - 5 Right thumb through right four fingers6 10 Left thumb through left four fingers13- Plain right four fingers14- Plain left four fingers15- Plain both thumbs.

5.2.2.2 For Verification

The valid values for finger position are from 0 to10.0 - Unknown Finger1 - 5 Right thumb through right four fingers6 10 Left thumb through left four fingers


Refer Table 5 of ISO 19794-4 for the complete table

5.2.3 Rotation Angle

No rotation angle is permitted.


5.2.4 Number of Fingers

In general, every additional finger increases accuracy and improves the possibility of bettermatching. However in view of constraints of storage space, the number of fingers to becaptured should be optimized depending upon the purpose, sensitivity and accuracyrequirements of the e-Governance applications.

5.2.4.1 For Enrolment/ IdentificationMaximum no. of fingers to be captured as 10, and minimum number is 1

5.2.4.2 For VerificationMinimum 1 finger to be captured

5.2.5 Image acquisition format

5.2.5.1 For Enrolment / Identification and for Verification in highly sensitiveapplications

Uncompressed image format (Code No 0 and 1) of ISO 19794-4:2005(E)


9/25



5.2.5.2 For Verification in less demanding applications.JPEG 2000 Part I (Code No 4 of ISO 19794-4:2005(E))

Note: The purpose of mandating acquisition of image in lossless format is to avoid asituation in which a vendor may acquire the image in some lossy compressed format, andmay convert the same into PNG format for storage described in section 5.4. In such casesthe original loss will get carried over to the stored image.


5.3 Quality SpecificationsCaptured images must be checked for image quality. While many proprietary algorithmsclaim superior image quality indicators, NIST Fingerprint Image Quality (NFIQ) is publiclyavailable and has been widely used. Images captured with NFIQ value of 1, 2 and 3qualifyfor acceptable quality.

NFIQ level 4 or 5 are poor quality images for minutiae data creation and are discouragedfrom use for enrolment / Verification/Identification purposes. However, if it is not possible toobtain desired quality images even after 5 attempts, best out of these attempts may beaccepted for storage / matching.


5.4 Storage SpecificationsOnce the fingerprint image gets qualified for storage as above, it needs to be retained /

stored for future reference. The data format specifications for storage / archival shouldadhere to the Policy on Open Standards to ensure interoperability, long-term usage andoptimal utilization of storage space, without affecting the quality of image.

5.4.1 Fingerprint Image Storage Format:

5.4.1. For Storage / Archival

PNG (lossless compression with compression ratio 1:2- Code No 5 of ISO 19794-4:2005(E)) format of fingerprints will be used for storage during enrolment, verification andidentification. Archival of enrolled fingerprints will also be the same. Code for PNG format is

5.4.1. 2 For Smart Card storage

For smart card applications having limited storage capacity, compressed fingerprints inWSQ format will be stored. The Compression ratio with respect to the originaluncompressed fingerprint can be up to and should not exceed 1:15 .

5.4.1.3 Fingerprint image Transmission Format for Verification

- For Low bandwidth the transmission may be either in the form of minutiae data instandard minutiae template or Fingerprint image itself in WSQ (Compression up to1:15 - Code No 2 of ISO 19794-4:2005(E)) format.


10/25



- For High bandwidth the transmission to be in PNG (lossless compression ratio of1:2 - Code No 5 of ISO 19794-4:2005(E)) format.


5.5 Minutiae data Format SpecificationsISO /IEC 19794-2 Finger Minutiae data standard would be adopted in this version of thestandard. The mandatory values in this format specified in the ISO 19794-2 standardshould be used for the purpose of matching. The extended data, even if extracted, neednot be utilized. The ridge count data, core and delta data, and zonal quality data that areoptional in this standard should be stored in the open format specified in this standard.Vendors are not allowed to store these optional values in a proprietary format. Everyvendors minutiae matching algorithm must have a version that utilizes just the mandatoryvalues in the minutiae storage format for matching. This ensures that each vendorsalgorithm has at least one version that is completely interoperable.



11/25



6 Recommended Best PracticesThe e-governance standards are intended for a variety of applications with varying degreeof sensitivity and volume requirements. The specifications in Section 5 are broad in natureto cater to all types of e-Governance applications. This Section covers the best practicessuitable for various categories of applications

6.1 Best Practices for Implementation of Fingerprint imageStandard

6.1.1 Device Specifications

For enrolment/identification at image acquisition setting level 31 and 41, EFTS/FCertification of the device is advised.

6.1.2 Image Specifications

Refer Section 5.2.1

Impression TypeSpecifications

Although Live scan and non live scan images meeting thestandard specifications do not have interoperability issues, butstill matching a scanned image with a live print, may not givesame accuracy as matching two live scan prints. Hence theimpression types 0 and 1 should be preferred in e-Governanceapplications for enrolment and verification.In case of Enrolment / Verification in legacy applications for

specific local requirements, impression types 2, 3 can beallowed for operational requirements.However, in the new versions of the legacy application, it isadvised to collect Fingerprint image data from live-scan devices(0,1,9)

Refer Section 5.2.2 forFinger positionspecifications

The choice of finger position would be application dependent,and there should be clear directions indicating which finger (s)data is to be captured for enrolment/ verification/ identification.In case of enrolment using 13, 14 or 15 (4+4+2), thesegmentation of individual fingers with appropriate finger

position should be done and stored.

Refer Section 5.2.3 forRotation Anglespecifications

Operational Instruction:It is expected that there is no rotation in the finger. Supervisedcapturing is recommended to ensure that there is no rotation orthe device must ensure automatic rotation correction.Specificallyfor enrolment and identification, the vendor needs toensure minimal rotation (not necessarily 0angle) with the helpof rotation detection and correction algorithm.


12/25



Refer Section 5.2.4 forNo. of Fingersspecifications

For Enrolment / Identification

The number of fingers to be captured should be based on the

(a) Total population of the enrolment (b) interoperability needsand (c) de-duplication plans. If there are no plans to perform de-duplication or identification, one or two fingers may suffice. Forany state wide or larger enrolment, ten fingers should becaptured for de-duplication. A smaller enrolment for accesscontrol or district level benefit scheme could use less number offingers. If capturing 1 or 2 fingers, the index finger(s) should becaptured. If capturing 4 fingers, two index fingers and twothumbs should be captured.

Any finger option

The normal practice is to use one specific finger, say the indexfinger for verification. However, current technology could allowthe person to use any finger. This is not merely a question ofconvenience. Certain fingers, depending on the condition of thefinger, will perform better in matching. While one cannot easilydetermine this a priori, a user will learn it by experience. Thisimproves subsequent user experience and could potentiallyimprove match accuracy.

No. of fingers to be enrolled will be application dependent.However, for National database of person identification,capturing of 10 fingers data is recommended. For operational

ease, in case of 10 fingers enrolment, 4+4+2 slab could beconsidered. The option of capturing by other devices also can beconsidered.

There should be clear operational instructions to ensure that thecorrect finger positions are linked with the stored images

For VerificationNormally only 1 finger is required. In situations requiring higherlevel of accuracy, two or more fingers may be used.

6.1.3 Image Quality SpecificationsRefer Section 5.3 forFinger Image Qualityspecifications

The capture software should have an automatic mechanism tocheck the NFIQ level of captured image, on the spot. Thesoftware should accept images with NFIQ quality levels 1 / 2 / 3only, or accept best out of five attempts, in case the image of aperson does not fall within acceptable NFIQ level.

You may get poor quality image under circumstances like:a) Due to environmental conditionsb) Due to the ridges being worn outc) Due to injury to the finger(s)In case of b and c, best quality fingerprint image out of 5


13/25



attempts may be considered.In case of a following operational instructions may be followed:

- Operator to guide enrolee hand and apply pressure if

necessary to obtain best possible image quality.- For corrective measures and re-tries, operator to wipe the

finger(s) of enrolee with wet cloth or apply lotion.Note: The operators need to be trained on the above

6.1. 4 Storage SpecificationsAcquisition for Enrolment /Identification

Only those devices to be used, which have option to capturefingerprint images in raw or lossless PNG format. Imagescaptured in raw format will have to be converted into PNGformat for storage at Client / Server system

Acquisition for verification In case of e-Governance applications requiring finger printimage verification from local data base only, and the accuracy /sensitivity level is not so high then the image acquisition canbe even in WSQ ( up to compression ratio 1: 15)

6.1.5 Minutiae DataEvery minutiae matching algorithm must have a version thatutilizes just the mandatory values in the minutiae storageformat for matching. This ensures that each algorithm has atleast one version that is completely interoperable.The matcher and extractor algorithm should have beenparticipated in the MINEX interoperability test or any other

equivalent test.

6.2 Best Practices for various processes

6.2.1 Enrolment Process

I. Process at Client end

a. Capture fingerprint Image data through a Fingerprint Image Scanner

connected with a computer for on line processing, as per the standardspecifications

b. Do quality check of captured image online as per the standardspecifications.

c. Store Biometric Data in secure manner on Client machine along with demographicdata of the enrollee, if available.

II. Process at Server end

d. Create biometric template to be used for identification or de-duplication.


14/25



e. Compare enrolees biometric data against entire data base (1: N identification) storedon the server to ensure that enrolee is not being enrolled 2nd time.

f. Depending upon the type of e-Gov. Application, decide either to reject the biometricdata of enrolee or identify the subject accurately.

g. Store / Archive the Biometric data of the enrollee in lossless compression PNGformat, as per standard specifications, on the server database, in case the data wasnot found on the server earlier.

6.2.2 Verification Process

a. Capture Fingerprint Image data through a Fingerprint Image Scannerconnected with a computer for online verification, as per the standardspecifications.

b. Do quality check online as per the standard specifications.

c. Create minutiae data

e. For on line verification, do 1:1 matching of captured biometric data withthe corresponding data from the server.

f. For off line verification, do 1:1 matching of captured biometric data withthe corresponding image data stored on the smart card / Client systemitself or do 1:1 matching of earlier enrolled minutiae data and to-be-verified minutiae data.

6.2.3 Process for capturing rolled fingerprint imagesThe rolled image, common in forensic applications, contains twice as much information asthe plain image. The plain image is easier to capture and the capturing of rolled image ofenrolee requires guidance by trained operator.

6.3 Fingerprint Record Format

This is a format to store biometric data within a biometric data record to cater tointeroperability requirements of the biometric data taken by various image acquisitiondevices. This format also stores specific information related to the fingerprint images

CBEFF (Common Biometric Exchange Formats Framework) described in ISO 19794-4 willbe adopted. This format is based on ISO 19785-1. As per ISO 19785-1, ISO 19794-4 , theCommon Biometric Exchange Format Framework (CBEFF) for fingerprint data is structuredas follows:

- SBH (Standard Biometric Header )

-BDB ( Biometric Data Block) for Finger PrintGeneral Fingerprint Record HeaderImage Record Header


15/25



-Image Data (Compressed / Uncompressed)Fingerprint dataFor best practices refer Section 6.2 of this document

Certain values in the Finger image format would be static as per the standard. Certainvalues will be dynamic in nature.

In the following tables, values in shaded rows are variable in nature and need to be enteredon case to case basis. Values in other rows are static in nature as per the standardspecifications, and should not to be altered.

Fingerprint Biometric Data Block (BDB) (Extracted from ISO 19794-4)

a. Values in shaded rows are variable in nature and need to be entered on case- to case

basis.

b. Values in other rows are fixed in nature as per ISO 19794-5:2005(E) specifications andshould not to be altered

Table 1 for General Record Header

Field Size Value

Format Identifier 4 bytes 0x464952 (F I R) FingerImage Record

Version Number 4 bytes 0x30313000 (0 1 0 0x0)

Record Length 6 bytes General Record Header length+No. of Views * ( Image RecordHeader Length + Image DataLength)

32+ Number Views *(14 bytes + Data length )32+1*(14+12028)=12074 bytes

Capture Device ID 2 Bytes 0x0 by default

Image Acquisition Level 2 bytes Value as per Section 5.1.1requirements

Number of fingers 1 byte Value as per Section 5.1.2requirements

Scale units 1 byte 2 representing pixels per cm.

Scan Resolution 2 bytes Value as per Section 5.1.1requirements


16/25



Scan Resolution (Vert.) 2 bytes Value as per Section 5.1.1requirements

Image Resolution 2 bytes Value as per Section 5.1.1

requirements

Image Resolution (Vert.) 2 bytes Value as per Section 5.1.1requirements

Pixel Depth 1 byte Value as per Section 5.1.1requirements

Image compressionAlgorithm

1 byte Value as per Section 5.4requirements

Reserved 2 byte 0x0 default value


17/25



Table 2 for Image Record Header

Field Size Values

Length of finger datablock (bytes)

4 bytes To be calculated

Finger position 1 byte Values as per Section 5.2Count of views 1 byte Values as per Section 5.4

requirementView number 1byte

Finger image quality 1 byte Actual value as per Section 5.3requirements

Impression type 1 byte Values as per Section 5.2

Horizontal line length 2 bytes Value as per input fingerprintimage

Vertical line length 2 bytes Value as per input fingerprintimage

Reserved 1 byte 0x0 by defaultFinger image data < 4310

bytesValue as per input fingerprintimage.


18/25



7 Annexure

Annexure 1

Definitions and AcronymsAcquisition

Process of accepting a biometric sample(s) in accordance with the defined policy, that isdeemed suitable for creating a biometric reference or a biometric probe.

Note: In addition to capture, acquisition may include segmentation, biometrics featureextraction, quality control and other pre-processing steps.

Biometrics[Automated] recognition of [living] persons based on observation of behavioural andbiological (anatomical and physiological) characteristics.

Biometric systemAn automated system capable of:1. Capturing a biometric sample from an end user;2. Extracting biometric data from that sample;3. Comparing the biometric data with that contained in one or more reference templates;4. Deciding how well they match; and5. Indicating whether or not an identification or verification of identity has been achieved.

Biometric dataThe data representing a biometric characteristic

EXAMPLE: sensor data, image data, behavioural data, feature data

NOTE For the purpose of this document, biometric data refers to fingerprint Image data.Biometric Data Block (BDB)Block of data with a defined format that contains one or more biometric samples orbiometric templates

Biometric Information Record (BIR)Data structure containing one or more BDBs together with information identifying the BDB

formats, and possibly further information such as whether the BDB is encrypted.

Biometric sampleData obtained from a biometric device, either directly or after processing.

Biometric TemplateBiometric sample or combination of biometric samples that is suitable for storage as areference for future comparison

biometric sampleinformation obtained from a biometric device, either directly or after processing


19/25



CaptureThe process of taking a biometric sample from an end user.

Capture device type IDThe capture device type ID shall be a unique identifier for the type of capture devicedeployed to acquire a biometric sample. The capture device type ID shall be recorded intwo bytes. A value of all zeros indicates that the capture device type ID is unreported. Thevalue unreported may not be allowable in some applications. The value field isdetermined by the vendor possibly depending on requirements for the respectiveapplication

Friction ridgeThe ridges present on the skin of the fingers and toes, the palms and soles of the feet,which makes contact with an incident surface under normal touch. On the fingers, theunique patterns formed by the friction ridges make up fingerprints.

EnrolmentIn enrolment, a transaction by a subject is processed by the system in order to generateand store an enrolment template for that individual.

Enrolment typically involves:

Sample acquisition;

Segmentation and feature extraction;

Quality checks, (which may reject the sample/features as being unsuitable forcreating a template, and require acquisition of further samples);

Template creation (which may require features from multiple samples), possibleconversion into a biometric

Data interchange format and storage; Test verification or identification attempts to ensure that the resulting enrolment is

usable;

Should the initial enrolment be deemed unsatisfactory, repeated enrolmentattempts may be allowed (dependent on the enrolment policy).

NOTE: An accept or reject decision is then based on whether this score exceeds the giventhreshold.

IdentificationIn identification, a transaction by a subject is processed by the system in order to find anidentifier of the subjects enrolment. Identification provides a candidate list of identifiers that

may be empty or contain only one identifier. Identification is considered correct when thesubject is enrolled, and an identifier for their enrolments in the candidate list. Theidentification is considered to be erroneous if either an enrolled subjects identifier is not inthe resulting candidate list (false-negative identification error), or if a transaction by a non-enrolled subject produces a non-empty candidate list (false-positive identification error).

Identification typically involves:

Sample acquisition;

Segmentation and feature extraction;

Quality checks (which may reject the sample/features as being unsuitable forcomparison, and require acquisition of further samples);


20/25



comparison against some or all templates in the enrolment database, producing asimilarity score for each comparison

Judgment on whether each matched template is a potential candidate identifier for

the user, based on whether the similarity score exceeds a threshold and/or isamong the highest scores returned, producing a candidate list;

An identification decision based on the candidate lists from one or more attempts,as dictated by the decision policy.

intermediate biometric samplebiometric sample obtained by processing an acquired biometric sample, intended for furtherprocessing

LatentA fingerprint collected from an intermediate surface, rather than directly via a live capturefrom the finger itself.

Live captureThe process of capturing a biometric sample by an interaction between an end user and abiometric system.

Live-scan printA fingerprint image that is produced by scanning or imaging a live finger to generate animage of the friction ridges.

MatchingThe process of comparing biometric data with a previously stored biometric template andscoring the level of similarity.

NOTE: An accept or reject decision is then based on whether this score exceeds the giventhreshold.

PolicyCourse or principle of action adopted or proposed by an organization or individual

ResolutionNumber of pixels per unit lengthNOTE: Pixels per cenimeter(ppcm) or( pixels per inch (ppi) ) will be used in this part ofBiometrics Indian e-Governance standard as the units of resolution.

Standard Biometric Header

Provides encodings for abstract values of CBEFF data elements and enables anapplication to obtain knowledge about the format and other properties (such as creationdate) of the BDBs that are contained in the BIR without having to process the BDBsthemselves

NOTE: BDBs are not required to be (and generally are not) self-identifying. Identification ofBDB formats is provided in CBEFF data elements.

SwipeA method of fingerprint collection where the finger is manually moved across a one-dimensional sensor to produce the two-dimensional image system: An organized scheme


21/25



or method; a complex whole; a set of things working together as mechanism orinterconnected network.

VerifyMake sure or demonstrate that something is true, accurate or justified.

VerificationIn verification, a transaction by a subject is processed by the system in order to verify apositive specific claim about the subjects enrolment (e.g. I am enrolled as subject X).Verification will either accept or reject the claim. The verification decision outcome isconsidered to be erroneous if either a false claim is accepted (false accept) or a true claimis rejected (false reject). Note that some biometric systems will allow a single end-user toenrol more than one instance of a biometric characteristic (for example, an iris system mayallow end-users to enrol both iris images, while a fingerprint system may have end-usersenrol two or more fingers as backup, in case one finger gets damaged).

Verification typically involves:

sample acquisition,

segmentation and feature extraction,

quality checks (which may reject the sample/features as being unsuitable forcomparison, and require acquisition of further samples),

comparison of the sample features against the template for the claimed identityproducing a similarity score,

Judgment on whether the sample features match the template based on whetherthe similarity score exceeds a threshold,

A verification decision based on the match result of one or more attempts asdictated by the decision policy.

EXAMPLE In a verification system allowing up to three attempts to be matched to anenrolled template, a false rejection will result with any combination of failures-to-acquireand false non-matches over three attempts. A false acceptance will result if a sample isacquired and falsely matched to the enrolled template for the claimed identity on any ofthree attempts.

Abbreviated Terms

API Application Programming InterfaceBDB Biometric Data BlockBIR Biometric Information Record

CBEFF Common Biometric Exchange Formats FrameworkPPCM pixels per centimetrePPI pixels per inchSBH Standard Biometric HeaderSB Security BlockPNG Portable Network GraphicsWSQ Wavelet Scalar Quantization


22/25




23/25



8 References1. ISO 19794-4 Information Technology Biometric data interchange formats Part 4

Finger image data.

2. ISO/IEC 19785-1, Information technology Common biometric exchange formatsframework Part 1: Data element Specification.

3. MTR 04B0000022 (Mitre Technical Report), Margaret Lepley, Profile for 1000Fingerprint compression, Version 1.1, April 2004 Available athttp://www.mitre.org/work/tech_papers/tech_papers_04/lepley_fingerprint/lepley_fingerprint.pdf

4. IAFIS-IC-0110 (V3), WSQ Gray- scale Fingerprint Image Compression Specification1997.

5. ANSI/NIST-ITL 1-2000, Information systems Data Format for the Interchange ofFingerprint, Facial and Scar Mark & Tattoo (SMT) Information.

6. ISO/IEC 19784-1 Biometric Application Programming interface Part1: BioAPIspecification.

7. NISTIR 7151 August 2004 Fingerprint Image Quality


24/25



9 List of Members

9.1 List of Members of Expert Committee

S. No Name & Designation Phone No/Email

1. Professor Manindra Agrawal(Department of computer Science &Engineering Indian Institute ofTechnology (IIT) Kanpur) Chairman

[email protected]

2. Professor Phalguni GuptaIndian Institute of Technology(IIT)Kanpur

[email protected]

3. Shri G.S. Raghu Raman

CMC Limited Hyderabad

[email protected]

4. Dr. M.V.N.K. PrasadAssistant ProfessorInstitute for Development & Researchin Banking Technology (IDRBT),Castle Hills, Road No. 1;Masab Tank, Hyderabad,AP-5000 057

[email protected]

5. Mr. S K Sinha (Nodal Officer)Scientist F , NIC, New Delhi

[email protected],24365085

6. Mr. R RameshScientist E , OTC, NIC, Chennai

[email protected]

7. Dr. Mayank VatsaAsst. Professor IIIT Delhi

[email protected]

8. Dr. Richa SinghAsst. Professor IIIT Delhi

[email protected]

9. Dr. Semant DubeyAsst. Professor IIT Kanpur(checkspellings in email)

[email protected]

9.2 Technical Expert on BiometricsS. No Name & Designation Phone No/Email

1. Dr. Krithika VenkataramaniAsst. Professor , IIT Kanpur

+91 512 259 [email protected]

2 Mr. Rajesh MashruwalaMember UIDAI

[email protected]


25/25



9.4 Other Contributors


1. Mrs. Aruna Chaba, Scientist F&Head, e-Governance StandardsDivision, NIC, New Delhi

[email protected]

2. Dr. P. BalasubramanianScientist G & Head OTC, NIC,Chennai

[email protected]

3. Ms. Renu Budhiraja, Director (EgovDivision), DIT

[email protected]

4. Ms. Anita Mittal, Senior Consultant ,DIT

9.5 List of Research Associates


1. Mr. Rajnish Kumar SharmaJunior Research Fellow, NIC

[email protected]

Documents

Draft Finger Image and Minutiae Data Standard Version 1