Page 129 - Automated Fingerprint Identification Systems (AFIS)
P. 129

114  AUTOMATED FINGERPRINT IDENTIFICATION SYSTEMS



                              results of a name search of the MNI. Following evaluation, “lights out” may be
                              extended to additional types of tenprint searches.



                              5.4.2 LATENT TO TENPRINT (LT/TP) SEARCHES
                              Because the latent print examiner does not work with the same volume or type
                              of identification material as the tenprint examiner, latent print search proce-
                              dures are somewhat different. Whereas the tenprint examiner has two or four
                              impressions for a single search, the latent print examiner typically handles each
                              latent print as a separate search. Whereas the tenprint examiner works with
                              clear images that may have passed a quality check prior to transmission from a
                              livescan device, the latent print examiner has only the images left behind at a
                              crime scene, which may be smudged, have another latent print overlay, or be
                              on a background that must be neutralized for the image to appear.
                                 The output of the LT/TP search also differ from the TP/TP search. Like the
                              tenprint examiner, the latent print examiner is presented with a candidate list.
                              However, the latent print examiner may request that all candidates above a
                              certain threshold, or a specific number of candidates, be presented for com-
                              parison. Because there are fewer minutiae to work with, the latent print exam-
                              iner will look at matcher score, pattern, and ridge flow in addition to the
                              placement of the minutiae in the side-by-side comparisons. If a candidate does
                              match the latent print, the examiner may request that a card be printed or
                              retrieved and compared with the latent print. The examiner may also request
                              that a second or senior examiner confirm the ident.
                                 If the first candidate does not match, the examiner continues on through
                              the list of candidates above the threshold; if necessary, he or she may also look
                              at candidates that are below the threshold. Given the combination of a low-
                              quality latent print and low-quality minutiae on the enrolled image, it is possi-
                              ble that the score for the minutiae match could be below the threshold but still
                              be within the requested list of candidates.
                                 When two latent prints (or more) are searched on AFIS, the appearance of
                              the same SID number on candidate lists for both images means that a candi-
                              date for the first latent print image also appeared as a candidate on the second
                              image. This information can aid the examiner in making an identification. It
                              may be, for example, that these images are fingers 2 and 3 (right hand index
                              and middle fingers) of the same person left at the crime scene. Again, the AFIS
                              search only provides information, not a determination, for the latent print
                              examiner.
                                 If the search does not produce an ident, the examiner may continue to
                              search the latent print against the database by changing variables. Depending
                              on the system, these variables could include geographic area searched, change
   124   125   126   127   128   129   130   131   132   133   134