AFIS search results, which include both a fingerprint image and a numeric value for each respondent, are then examined by the AFIS Technician.  The name Automated Fingerprint Identification System is actually a misnomer since the computer - for crime scene print to inked print search cases - NEVER makes an identification but only matches data files that are possibilities. * * The AFIS Technician, following an identification process involving an identification philosophy and scientific methodolgy, determines if the unknown crime scene print came from the same source as the respondent image (known database print).  Any identification made by the AFIS Technician using the computer screen images is verified by another qualified AFIS Technician or Identification Officer.  The hard copy crime scene print and 'person of interest' fingerprint impressions on his/her actual fingerprint form is then compared and evaluated.   In order to complete the scientific process, a second AFIS Technician or Identification Officer completes a verification.
A.F.I.S. is a useful TOOL which can assist crime investigators in many areas, such as:
                              
                              Identifying crime scene suspects
                              Identifying "liars" (Suspects who give officers a false name)
                              Identifying deceased parties
                              Identifying unknown parties (e.g. Alzheimer patients etc.)

Fingerprint images and fingerprint classification details including minutiae (major ridge path deviation locations) are electronically stored in a computerized database.

Full sets of 'ten print' fingerprint records, after being examined by a trained technician, are stored on AFIS.  Each fingerprint image is 'filed' based on its pattern type (similar to the
Henry Classification System), its core-delta distance* (applicable to loops and whorls only) and minutiae.  RCMP and local police services maintain their ten print records on a daily basis.
(* core-delta distance is not applicable to all AFIS systems)

How Does AFIS fit into the Investigation Process?

Evidence prints (latents) located at crime scenes or on documents are first collected by the Identification or Scene of Crime Officer.  The Ident Officer reviews each print for possible manual comparison (if a suspect is known) and AFIS suitability.

The crime scene print is submitted to AFIS.  The AFIS Technician (perhaps better described as a Friction Ridge Identification Specialist) analyzes the impression in order to determine possible pattern type, major ridge path deviations, different types of distortion present, proper orientation and possible digit determination.   Analysis is the first step in the
identification process.  The technician then begins the process of loading the print on the AFIS computer using as much of the criteria previously mentioned as can be determined from the crime scene print.  Quite often crime scene prints are only partial prints and extremely distorted.  AFIS is capable of enhancing the print using contrast and brightness adjustments but the technician must be a fully trained latent print examiner in order to accurately 'load' the print. 

Specific search parametres are set and AFIS searches the specified 'known' ten print file databases using the crime scene print information such as  pattern type, core-delta distance* and specified digit information (if possible)  to filter out file prints that do not match.  Obviously the more information about the print that can be specified will help narrow the search.  The minutiae (major ridge path deviation locations) data is then compared and a match report is generated listing the respondents.
AFIS search results, which include both a fingerprint image and a numeric value for each respondent, are then examined by the AFIS Technician.  The name Automated Fingerprint Identification System is actually a misnomer since the computer - for crime scene print to inked print search cases - NEVER makes an identification but only matches data files that are possibilities. * * The AFIS Technician, following an identification process involving an identification philosophy and scientific methodolgy, determines if the unknown crime scene print came from the same source as the respondent image (known database print).  Any identification made by the AFIS Technician using the computer screen images is verified by another qualified AFIS Technician or Identification Officer.  The hard copy crime scene print and 'person of interest' fingerprint impressions on his/her actual fingerprint form is then compared and evaluated.   In order to complete the scientific process, a second AFIS Technician or Identification Officer completes a verification.
Automated Fingerprint Identification System (A.F.I.S.)
A.F.I.S. Fingerprint Characteristics

          *  Pattern Type (based on Henry Classification System)
          *  Core (central area) and Delta (triradius area) distance
          *  Minutiae locations (at bifurcations, ridge endings, enclosures                and short ridges.                                                    

................Short Ridge
.............Ridge Ending
">
">
">
">
Bifurcation..............................
">
">
">
">
">
">
">
">
">
">
">
">
Enclosure ....................................................
Left Loop
**Update on AFIS from ED German (10Feb2002):
"As of last fall there were four states in the US where completely "lights-out" ten print to ten print identifications were effected by AFIS with no human intervention... and reported back to agencies as positive identifications (again, we are talking only ten print to ten print). The threshold matching score required for such lights-out identifications was based on years and years of experience with their systems. The error rate has been zero, though as with human comparisons they do routine QA/QC reviews and special reviews if a defendant questions the ident, etc.. The no-human-intervention-identification only occurs when the matching score is sufficiently high... but, it is a reality.

One or more additional states may have started limited lights-out automated identifications (TP to TP only, again) since last fall. It makes sense, especially based on a performance record of zero errors when a sufficiently high threshold score is utilized."
">
">
">
">
..................Ridge Dot