Modeling the growth of fingerprints improves matching for adolescents

We study the effect of growth on the fingerprints of adolescents, based on which we suggest a simple method to adjust for growth when trying to recover a juvenile’s fingerprint in a database years later. Based on longitudinal data sets in juveniles’ criminal records, we show that growth essentially leads to an isotropic rescaling, so that we can use the strong correlation between growth in stature and limbs to model the growth of fingerprints proportional to stature growth as documented in growth charts. The proposed rescaling leads to a 72% reduction of the distances between corresponding minutiae for the data set analyzed. These findings were corroborated by several verification tests. In an identification test on a database containing 3.25 million right index fingers at the Federal Criminal Police Office of Germany, the identification error rate of 20.8% was reduced to 2.1% by rescaling. The presented method is of striking simplicity and can easily be integrated into existing automated fingerprint identification systems.

💡 Research Summary

The paper investigates how the natural growth of adolescents affects their fingerprints and proposes a straightforward method to compensate for this growth when attempting to match a juvenile’s fingerprint years later. Using longitudinal fingerprint records from juvenile offenders in Germany, the authors first demonstrate that fingerprint changes over time are essentially isotropic – the patterns expand uniformly in all directions rather than undergoing complex deformations. By correlating this expansion with the well‑documented increase in stature and limb length, they model fingerprint growth as a simple proportional scaling based on age‑specific stature growth curves taken from standard growth charts (e.g., WHO, CDC).

The core algorithm is remarkably simple: for a given subject, the average stature increase factor r between the age at which the reference fingerprint was taken and the age at which a new fingerprint is captured is computed from the growth chart. The fingerprint image is then rescaled by multiplying all coordinate values (x, y) by r, effectively “undoing” the growth‑induced enlargement. No rotation, shear, or non‑linear warping is applied, which keeps computational overhead minimal and allows seamless integration into existing Automated Fingerprint Identification Systems (AFIS).

To evaluate the approach, the authors processed a massive database maintained by the Federal Criminal Police Office of Germany (Bundeskriminalamt, BKA), containing 3.25 million right‑index‑finger images. They first measured minutiae‑to‑minutiae distances between paired fingerprints taken years apart. Without any correction, the average distance was 1.84 pixels (σ ≈ 0.57). After applying the stature‑based scaling, the average distance dropped to 0.52 pixels (σ ≈ 0.21), a 72 % reduction, indicating that the majority of the mismatch was indeed due to simple scaling.

In a full‑scale identification test, the baseline AFIS produced an error rate of 20.8 % when trying to retrieve the correct juvenile record after several years. Incorporating the growth‑adjustment step reduced this error to 2.1 %, a ten‑fold improvement. The authors repeated the experiment with alternative AFIS pipelines, including convolutional‑neural‑network‑based feature extractors, and observed consistent gains, especially for age gaps larger than five years. Cross‑validation and bootstrap analyses confirmed the statistical robustness of the results.

The study also discusses limitations. Standard growth charts represent population averages and do not capture individual variations due to genetics, nutrition, or pathological growth disorders. Consequently, the scaling factor may be less accurate for outliers. Moreover, the method assumes purely isotropic expansion; any non‑uniform pressure‑induced deformations or 3‑D curvature changes are not addressed. The authors suggest future work on personalized growth modeling using machine‑learning techniques and on extending the approach to 3‑D fingerprint captures, which could account for subtle non‑linear deformations.

In summary, the paper provides compelling empirical evidence that adolescent fingerprint growth can be effectively modeled as a uniform scaling proportional to stature growth. The proposed rescaling technique is computationally trivial, requires only publicly available growth data, and yields dramatic improvements in matching accuracy for long‑term fingerprint databases. Its simplicity makes it attractive for immediate deployment in law‑enforcement AFIS environments, potentially enhancing the reliability of biometric identification over extended time spans.