Fingerprint comparison is based on the agreement of ridges in sequence [19].  Agreement of ridges in sequence encompasses all aspects of agreement between ridge formations, i.e. ridge positions, intervening ridge counts, orientations, paths, spatial relationships, structure (ridge lengths, widths, edge contours), pores, creases, scars, and so on.  Ridges should be “run”, analyzed and compared ridge unit by ridge unit, one after the other.  Sequence should be maintained at all times in order to encompass the entire amount of corresponding ridge detail present. 

Based on the product rule, quantitative-qualitative values for pairs of corresponding ridge units accumulate as ridges are run.  The same values that are defined for individual ridge units are applied to pairs of ridge units in two impressions found in agreement in which pairs of “corresponding” ridge formations displaying ridge formation type disagreement default to lower bound quantitative weights with exception to the ending/bifurcating ridge which results when an ending ridge is displayed in one fingerprint impression and a bifurcation is displayed in the other (the value for this type of pair of ridge formations is defined by the average of the two quantitative weights multiplied by the reduction factor for connectivity disagreement).  Otherwise, for purposes of conservativeness, the ridge formation with the lower quantitative weight is applied to the value for the pair of ridge formations.

 

Quantity and Quality Defined During the Comparison Phase

The T Model defines the value for a pair of corresponding ridge features during the comparison phase a latent v. exemplar as the product of the lower-bound quantitative weight (e.g., based on shape and position) and the quality of agreement. 

Qualitative metrics were designed to reflect reduced levels of quality of agreement in terms of ridge feature type, ridge path, and spatial relationship to nearest neighbor.   The reduction factors were designed to reduce the quantitative weight for the pair of corresponding ridge features in the latent v. exemplar in proportion to reduced levels of ridge feature quality of agreement, keeping in mind the need for optimum reproducibility, simplicity, and sensitivity. 

A 5-tier qualitative grading scale (e.g., A, B, C, D, F) with corresponding reduction factors (e.g., 1.00, 0.75, 0.50, 0.25, 1/P), was designed based on objective criteria (see below Ridge Feature Quality of Agreement Grade Scale).

 

The quantity and quality of ridge features compared and the level of agreement and/or non-agreement found in each should be documented such that another qualified examiner can determine what was done, if needed (see Bench Notes). 

 

NEXT PAGE >>>