AFIS: Entry, Searching, and Candidate Evaluation

What AFIS Is — and What It Is Not

An Automated Fingerprint Identification System (AFIS) is a computerized database and search engine that encodes fingerprint ridge detail as mathematical data, compares that data against stored records, and returns a ranked list of candidates. AFIS dramatically narrows the search field when an unknown print needs to be compared against thousands or millions of records. However, AFIS does not make identifications. Every candidate list it produces is preliminary. A trained human examiner must apply ACE-V methodology before any conclusion is reported. This distinction is not a technicality — it is a foundational principle of forensic science and is required by SWGFAST documentation standards and IAI certification guidelines.

How AFIS Encodes a Print

When you enter a latent print into an AFIS terminal, the system (or you, with the system's tools) identifies and records specific ridge features called minutiae. Each minutia is stored as a data point with three values: its x-coordinate, its y-coordinate, and its ridge direction (angle). The most commonly encoded minutiae types are ridge endings and bifurcations, because these are the most reliably detected features.

For rolled or flat ink prints in the database, the encoding is done at the time of intake and is typically more complete. For a latent print — which may be partial, smeared, or developed under difficult conditions — the examiner must make judgment calls about which minutiae are clear enough to encode reliably. Encoding too many uncertain points degrades search accuracy; encoding too few may miss the correct candidate.

Where the print permits, you also mark the core (the innermost recurving ridge or the center point of a whorl) and delta (the triangular area where ridge flow diverges). These landmarks help the algorithm orient the print before comparing.

How the Search Algorithm Works

Once you submit an encoded latent, the AFIS algorithm compares your minutiae map against every record in the selected database. It calculates a similarity score for each comparison and returns the top candidates ranked by score — commonly the top ten or twenty, depending on system settings. A high score means the algorithm found a strong mathematical similarity. It does not mean the prints match. Score thresholds vary by system and jurisdiction; none of them constitute a legal identification.

Think of the score list as a sorted stack of "look here first" suggestions. Your job as the technician is to take those suggestions and apply rigorous manual analysis.

ACE-V Applied to Candidate Evaluation

ACE-V stands for Analysis, Comparison, Evaluation, and Verification. When working through a candidate list, you apply this process to the top candidates in order.

Analysis — Before looking at any candidate, analyze the unknown latent on its own. Note its quality, area of clarity, the number of discernible minutiae, their types, and their spatial relationships. Document your observations.

Comparison — Place the unknown latent alongside the candidate print (on screen, side by side, or using overlays). Work systematically — typically from a landmark outward — and note where features agree, where they differ, or where the latent is too unclear to assess a given area.

Evaluation — Based on your comparison, reach one of three conclusions: identification (the prints share the same source), exclusion (they do not), or inconclusive (insufficient quality or quantity of corresponding detail to reach either conclusion). In a training context, you will record "preliminary identification" or "no match" and flag for verification.

Verification — A second qualified examiner, working independently, must reach the same conclusion before any identification is reported. As a technician-level student, your role ends with the recommendation that a certified examiner verify your findings. Never report a final identification without that step.

Entering a Latent: Step-by-Step Overview

1. Load the latent image into the AFIS workstation and adjust brightness and contrast as needed — do not alter the image file itself; use only display adjustments.

2. Set the correct finger position code if known (or mark as unknown).

3. Mark the core and delta where they are visible and reliable.

4. Place minutiae one at a time, assigning each a type (ending or bifurcation) and confirming the system has captured the correct ridge direction.

5. Review your encoding against the print before submitting. Remove any doubtful minutiae.

6. Submit the search and record the search parameters and time in your notes.

Documentation Requirements

Each AFIS session must be documented in your case notes: the search date and time, the database searched, the number of candidates returned, the scores of the candidates you evaluated, your ACE-V conclusion for each evaluated candidate, and your verification recommendation. Incomplete documentation is grounds for a report to be rejected and can compromise a case in court.

Safety and Professional Responsibility

AFIS training terminals contain simulated or sanitized data, but treat all fingerprint data as sensitive. Do not share candidate images outside the classroom. If you are ever working in a real agency system during an internship or practicum, follow all agency data-security protocols without exception. Unauthorized access to law enforcement databases carries serious legal consequences. Always log out of the workstation when you step away.