User's Manual
A PPENDIX A – Q UALITY AND C ONTENT
86
© Copyright 2002, Bioscrypt Inc. All rights reserved.
Quality
The quality score is based on how well the ridge pattern is defined within the
fingerprint image that was enrolled. In other words, quality measures how clearly
the unit imaged the fingerprint. Poor quality enrollments can result in an elevated
rate of false rejection making it difficult for the user to verify reliably.
The score is given in stars («) and ranges from zero to five stars, with five being
the best quality (rarely obtained) and zero being the worst. Quality scores of three
stars and higher perform well with the Bioscrypt verification algorithm. In this
range, the algorithm readily compensates for differences in fingerprint quality. It
statistically is still true that the larger the quality score the better the performance
of an enrollment.
As a general rule of thumb, quality scores less than three stars require intervention
on the part of the Enroller or administrative software. Sources of low scores
include dry fingers and dirty sensors.
If the quality score falls below three stars, Bioscrypt recommends the following
options:
• Ensure that the sensor and finger are clean.
• If the finger and sensor are clean and a dry finger is suspected, try re-enrolling
one more time, leaving the finger on the sensor for several seconds prior to
enrollment. Frequently finger moisture accumulates over time to provide a
good image.
• Fingerprint quality can vary among individual fingers for the same person. Try
enrolling an alternate finger to see if the score improves.
• Alter the security level for that particular template by decreasing the threshold a
minimum of 1 level (e.g., change the value from medium [3] to low [4]). This will
offset the false rejection for that template by making it easier to match. If use of
that template indicates that raising the threshold one level still produces false
rejections, try setting the value to its lowest security (level 5).
Warning: Decreasing a template’s security may increase the risk of a
false acceptance for that template.
Very Low
Quality
Very High
quality