The discriminating value for a fingerprint ridge feature is determined by the value for both it's type and position. 

For purposes of simplicity the discriminating value for a ridge feature position is defined in terms of "intervening ridge count to the nearest 1 neighbor".  The values are based on the following frequency of occurrence experiment performed by the author:

The number of intervening ridges observed between 1170 pairs of the two most frequently occurring ridge formation types, the ending ridge and bifurcation, in 39 randomly selected flat fingerprint impressions, was counted.  The percentage distribution and respective discriminating values  for each were subsequently defined.  For purposes of simplicity, frequency of occurrence for “0” and “1” ridge counts were combined and equated to a baseline discriminating value equal to 1.  For purposes of conservatism discriminatory values for intervening ridges to the nearest neighbor were rounded down to the nearest whole number (see below table). 

Discriminating Values for Ridge Feature Positions PDF Click HERE

Experiment to Check Value for Ridge Feature Position

Based on the above frequency of occurrence study of intervening ridge counts to the nearest Level II ridge feature neighbor in flat, distal fingerprints, the conservative, lower-bound discriminatory value (i.e., rounded down to the nearest whole number) for 4 intervening ridges located to a nearest neighbor was estimated to be 41. 

Based on data gathered as a result of previous experiment, the discriminatory value for 1 ending ridge, not located in a diminishing area or funnel, is estimated to be 14.25. Subsequently, for an arrangement of 2 ending ridges not in a funnel, with 1 ending ridge positioned 4 intervening ridges to its nearest neighbor, the T-Model predicts there will be not more than 3.6 close matches (i.e., look-alike arrangements) in 1000 distal, flat fingerprints.

The following close match experiment was performed to check this prediction. 

Close Match Experiment

An arrangement of 2 ending ridges separated by 4 intervening ridges was used for the experiment.  The arrangement contained 1 ending ridge positioned 4 intervening ridges away from its nearest neighbor which was a 2^nd ending ridge feature.  The ending ridges are marked in red and there are 4 intervening ridges from the ending ridge located near the bottom of the image to its nearest neighbor which is ending ridge located near the top of the image (Image 1). 

Image 1

1000 visually clear, minimally distorted flat, distal fingerprint impressions (i.e., 100 right thumb prints, 100 right index fingerprints, 100 right middle fingerprints, and so on), selected at random from a ten-print database and each from a different individual, were used to see how many times the above ridge feature arrangement occurred. 

The number of close matches, i.e., look-alike arrangements, found in the 1000 flat, distal fingerprint sampling was 2. 

Based on results from this experiment the discriminatory value of 41 for 4 intervening ridges to the nearest neighbor was deemed to be corroborated. 

It is significant to note that the number of close matches predicted by the T-Model was very close to, and slightly more than, the actual number observed.  This conservative, fairly accurate result is subsequently extremely appealing for purposes of criminal casework.



Example #1

The discriminatory value for an ending ridge (not in a funnel) is 14.25.  The discriminatory value for a feature position 4 intervening ridges to itws nearest neighbor is 41.

What is the total discriminatory value for the 2 ending ridges marked in red in the below image?

Answer

The discriminating value for the ending ridge at the bottom of the image is 14.25 x 41.  The value for the ending ridge feature type (i.e. 14.25) is mutiplied to the value for its position (i.e., 41) because there are 4 intervening ridges to its nearest neighbor. 

The discriminating value for the ending ridge near the top of the image is 14.25 x 1.  The value for the ending ridge feature type (i.e. 14.25) is mutiplied to the value for its position (i.e., 1) because there is 1 intervening ridge to its nearest neighbor.  

The total discriminating value for the above cluster of 2 ridge feature types in position is calculated as follows:

(14.25 x 41) x (14.25 x 1 ) = 8,352.56

Note:  The match probability for the above cluster of ridge features is defined as the inverse of the discriminating value, i.e., 1 / 8,525.56.



Example #2

The intervening ridge count to the nearest neighbor for each of the ending ridges shown in Image 1 is "2".   As a result, the positional discriminating value for each ending ridge is estimated to be "4".  Subsequently, the total discriminating value for each ending ridge (not in a funnel) is estimated to be 14.25 x 4 = 57.  Therefore, the discriminating value for the two ending ridges in combination is estimated to be 57 x 57 = 3,249.  The match probability (the reciprocal of the discriminating value) is therefore estimated to be 1 / 3,249.  

NOTE

It is significant to note that the discriminating values for a ridge feature position can be  refined by using more than 1 nearest neighbor.  For example, a frequency study can be performed that measures the frequency distribution for the sum of 3, 4, 5, or all nearest neighbors to a ridge feature.  The results will likely provide a broader range of discrimination values that can be relatively easy to compute.  However, at this time, and for purposes of simplicity, and for purposes of performing criminal casework in a timely manner, the discriminating values listed above are based on the frquency distribution for intervening ridges to the 1 nearest neighbor.     

NEXT PAGE >>> 

For purposes of conservativeness, discriminating values for ridge feature position should be used only for ridge counts that are clearly defined.  Ridge counts between ridge formation types at the periphery of the fingerprint impression, for example, should default to the lower ridge count present between the two nearest formation types or between the ridge formation type and the edge of the impression, whichever is less.

Example

The ridge count between an ending ridge located in a non-pattern force area and its nearest neighbor, i.e., another ending ridge also located in a non-pattern force area, is 7. 

Absent additional values for continuous ridge units and pores, and not taking into consideration ridge feature quality, the total discriminating value for this 1 ending ridge feature is defined as follows:

14.25 x 588 = 8,379

As a result, the match probability is defined as follows:  1 / 8,379.