11.07.2015 Views

and overview of the ANSI/NIST-ITL - NIST Visual Image Processing ...

and overview of the ANSI/NIST-ITL - NIST Visual Image Processing ...

and overview of the ANSI/NIST-ITL - NIST Visual Image Processing ...

SHOW MORE
SHOW LESS
  • No tags were found...

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

The <strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> St<strong>and</strong>ard:Forensic Data Interchange


Existing Capabilities<strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> 1-2011


Latent Friction Ridge PrintsExtended Feature SetMarkupsCoresDeltasDistinctive CharacteristicsMinutiaeDotsIncipient RidgesCreases & Linear DistortionsRidge Edge FeaturesPores & Ridge edge fields-- Automated Matching-- Forensic Analysis &Comparison


Body <strong>Image</strong>sFacePrior -- Ante-mortem:* Passport, driver license,family photos, etc.* Try to get frontal <strong>and</strong>fairly recentPost-mortem or for livingamnesiacs:O<strong>the</strong>r Parts, Scars, Tattoos, InjuriesPrior -- Ante-mortem:* As available* Distinctive features (e.g. club foot, extra finger, brokennose, etc.)Post-mortem or for living amnesiacs:* With ruler* All angles•Mark eye locations using 2Dmarkups-- Automated Matching-- Forensic Analysis &Comparison-- Forensic Analysis & Comparison


<strong>Image</strong> Markups• Forensics: 2D <strong>and</strong> 3D anthropomorphic facial image markup fields


DNA•Rapid DNA <strong>and</strong>Laboratory<strong>Processing</strong>•Pedigree tree• Claimed <strong>and</strong> Validatedrelationships• Data on victim <strong>and</strong>relatives in onetransaction• X-STR, Y-STR,Mitochondrial


Conformance Test ToolsBiometric ConformanceTest S<strong>of</strong>tware for<strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> 1-2011(AN-2011) Transactions *• Test assertions based on<strong>NIST</strong> SP 500-295• CTM for AN-2011• Tests for selected RecordTypes in traditionalencoding:• Installer• Comm<strong>and</strong> Line Interface• Under development: toolto test XML encoding* Developed by <strong>NIST</strong>/<strong>ITL</strong> Computer Security Division. Sponsored, in part, by DHS/US-VISIT


Dental Working Group• Dental Forensics Working with <strong>the</strong> AmericanDental Association to base Type-12 record upon <strong>the</strong>ir Spec 1058 Meetings held in Argentina,Washington, D.C., New York(with SWG-DVI), Lyon (withINTERPOL), Atlanta (withAAFS), <strong>and</strong> San Francisco (withADA) FBI has done <strong>the</strong> update to <strong>the</strong><strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> XML schemafor <strong>the</strong> current draft Will be presented as a supplementto <strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> 1-2011 forvoting


Dental ForensicsDraft available athttp://www.nist.gov/itl/iad/ig/ansi_st<strong>and</strong>ard_dental_forensics.cfmBased on ADA Specification 1058Familial Data SetDental History Data SetTooth Data SetMouth Data Set<strong>Visual</strong> <strong>Image</strong> Data SetRadiograph <strong>Image</strong> Data SetIncludes DICOM data & images (Electronic Health Record)


Key Challenges : Dental ForensicsTeeth numbering not <strong>the</strong> same in US <strong>and</strong> o<strong>the</strong>r nationsDecision: Use ISO st<strong>and</strong>ard for transmissionDental Forensic Systems all use different datadefinitions <strong>and</strong> levels <strong>of</strong> detailDecision:* Use <strong>the</strong> ADA Spec 1058 as a ‘translator’* State <strong>the</strong> level <strong>of</strong> aggregation for each system(i.e. NamUS, WinID, NCIC, UDIM, Plass & FastID)* Include original encoding with transmission tohelp resolves ambiguities


Pattern InjuryNew Fields added to Type-10Field 10.046: Victim / VICcurrent subject statuslivingbody statusdeceasedunknownClass 1 Natural Tissue (Whole or Fragment)Class 2 Decomposed (Whole or Fragment)Class 3 Skeletal (Whole or Fragment)Field 10.047: Data collection organization / DCO


Pattern InjuryField 10.048: Pattern injury description<strong>Image</strong> location (using NCIC codes to define body parts)Type <strong>of</strong> injury (Avulsion, laceration, etc.)Color <strong>of</strong> injurySurface contourShape <strong>and</strong> size (measured using 2D ABFO ruler)Tissue characteristicsUnderlying structurePossible cause (human, animal, object, etc.)


Perioral / CheiloscopicRecommend use <strong>of</strong> Field 10.029: 2Dfacial feature points / FFP to indicatefeatures <strong>of</strong> <strong>the</strong> lip print when sendingan image <strong>of</strong> <strong>the</strong> lips <strong>of</strong> <strong>the</strong> possiblesuspect.Add new code to Table 58 (Type-10image types) as Perioral.Modify Field 10.040 NCIC SMT codeto use <strong>the</strong> NCIC codes to specifylocation <strong>of</strong> <strong>the</strong> lip print on <strong>the</strong>victim’s body.Include information about commoncheiloscopic classifications


New <strong>Image</strong> TypesModify Field 10.012 Color space / CSP to include morepossibilities than currently (grayscale, color <strong>and</strong>undefined)SonogramRadiograph (X-ray), including CT scans<strong>and</strong> cone-beamsMagnetic resonance image (MRI)


Palatine DescriptorsProblem: ADA Mouth Data Set does notinclude descriptors for dental arch,rugae ramification or shape.Solution:Initiate a revision to ADAspecification 1058 instead <strong>of</strong> addingdescriptors to <strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> in <strong>the</strong>Supplement.Impact:Descriptors will not be includedinitially when Dental Supplement isadopted. Expected to take more than ayear.


Forensic VoiceFor investigatory <strong>and</strong> forensic usesnot intended for applications like access controlIncludes audio recordings, as well as related informationdiarization (when speech occurs during <strong>the</strong> recording)transcription (what is said)redaction annotation (‘bleeping’ <strong>of</strong> portions –sometimes required by legal authorities)snipping annotation (cutting out portions <strong>of</strong> a longer recording)audio characteristics (recording device, CODECs, etc.)geographic location <strong>of</strong> speakers <strong>and</strong> recording device(s)speaker characteristics (impairment, language, etc.)Example: Compare a recording <strong>of</strong> a ransom call in a kidnapping caseagainst known voice samples <strong>of</strong> suspects


Mobile BiometricsMobile ID Best Practice Recommendation (BPR)* issued in 2009* defined subject acquisition pr<strong>of</strong>iles* defined military <strong>and</strong> law enforcement pr<strong>of</strong>iles* requires use <strong>of</strong> <strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> st<strong>and</strong>ard<strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> 1-2011* incorporates key points <strong>of</strong> <strong>the</strong> BPRTechnological advancements <strong>and</strong> modified operationalneeds suggest that a new document is neededKey point: What type?


<strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> LITE


Object Oriented DataBullets <strong>and</strong> CartridgesIssue: Advances in imaging analysis are making new approaches toidentification <strong>of</strong> weapons or classes <strong>of</strong> weapons more accurate (i.e. 3Dtopographic data)Need: Databases for matcher system development <strong>and</strong> initial testing as wellas for sequestered testing (much like is usually done at <strong>NIST</strong>)Problem: How to receive <strong>the</strong> data (transmission protocols) <strong>and</strong> how to makeit accessible


Issue:Disaster Victim Identification Procedures using<strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> for data transmissionData to be received from around <strong>the</strong> worldFingerprints, Facial images, Dental records, DNA samplesData to be collected toge<strong>the</strong>r relating to a subject <strong>and</strong> updated as needed (i.e.pedigree trees <strong>and</strong> established relationships)Medical examiners & disaster recovery teams have not worked with policedepartments to transmit data in <strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> formatSolution:Develop a Best Practice Recommendations document


NEW PROPOSAL2013/2014: Prepare interfaces based upon st<strong>and</strong>ard• 2014: Demonstration at ADA meeting <strong>of</strong> a simulateddisaster with to show receiving information from NewYork, overseas, INTERPOL to <strong>the</strong> conference <strong>and</strong>entering it into NCIC, NamUs, WinID <strong>and</strong> checkingresults• focus on dental data exchange• 2015: Data-exchange simulation <strong>of</strong> a disaster, tocommemorate <strong>the</strong> 10 th anniversary <strong>of</strong> Katrina• include capabilities for DNA (including RapidDNA), fingerprints, as well as dental


WORKSHOP ON PROPOSED CHANGES TO<strong>ANSI</strong>/<strong>NIST</strong>-<strong>ITL</strong> STANDARD at <strong>NIST</strong>Monday January 28<strong>ITL</strong>09:00 – 10:00 Welcome <strong>and</strong> <strong>overview</strong> <strong>of</strong> <strong>the</strong> <strong>ANSI</strong>/<strong>NIST</strong>-current status <strong>and</strong> <strong>the</strong> st<strong>and</strong>ards development process10:00 – 10:30 NIEM <strong>and</strong> <strong>the</strong> XML Biometrics Domain10:30 – 11:00 List-driven Programming11:00 – 12:00 Ballistics Discussion13:00 – 17:00 Forensic Voice Supplement DiscussionTuesday January 2909:00 – 12:00 Mobile ID Discussion13:00 – 17:00 Mobile ID DiscussionWednesday January 3009:00 – 12:00 Dental Forensics Supplement Discussion13:00 – 15:00 Disaster Victim Identification Best Practices15:00 - 17:00 Open Forum


• FOR FURTHER INFORMATION:HTTP://biometrics.<strong>NIST</strong>.Gov& click for <strong>the</strong> st<strong>and</strong>ard's pageOrBrad.Wing@<strong>NIST</strong>.GOV

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!