Milestones:AFIS NEC

Title

First Operational Large-Scale Latent Fingerprint Identification System, 1982

Citation

NEC, formerly known as Nippon Electric Company, introduced the world's first large-scale automated fingerprint identification system (NEC AFIS) equipped with a latent fingerprint matching function in 1982. This was a powerful crime-solving tool capable of matching even fragmented latent fingerprints against a large database, a task that previously had been impossible. It enabled the world's police agencies to expedite searches for suspects, an efficiency that many public-safety experts valued.

Street address(es) and GPS coordinates of the Milestone Plaque Sites

5-7-1 Shiba, Minato-ku, Tokyo 108-8001, JapanLongitude: 139.748056 Latitude: 35.649432, 5-7-1 Shiba, Minato-ku, Tokyo 108-8001, JapanLongitude: 139.748056 Latitude: 35.649432

Details of the physical location of the plaque

The milestone plaque will be attached to the wall of the main entrance of the headquarters of NEC Corporation. The location will be suitable as every visitor will be able to see the milestone plaques

How the intended plaque site is protected/secured

The intended plaque site is in the site of the headquarters of NEC Corporation. Security guards watch but visitors can freely observe the milestone plaque.

Historical significance of the work

This work is historically significant because it was the world's first large-scale automated fingerprint identification system (AFIS) capable of matching latent fingerprint against a large database of 6 million subjects to be developed and successfully installed. After installation and commencement of operation at the San Francisco Police Department (SFPD) in December 1983 and in the State of California (CAL-ID) in August 1985, the NEC AFIS identified many previously unidentified latent prints and proved that this computerized system could match even small fragmental latent prints, which were impossible for human examiners to search for, to a large database very quickly. The success of these early installations prompted other law enforcement agencies to seriously consider procuring an AFIS and also pushed many AFIS vendors to enhance and improve their AFIS functions and matching accuracy. Other notable facts are given below.

1) The NEC AFIS achieved high latent print matching accuracy (far better than competitors) and quick response time (several minutes to several tens of minutes) with the large database (up to 6 million subjects). It identified the famous serial killer known as the “Night Stalker” in Los Angeles in 1985. [N1], [N2], [N3], [N4].

2) The NEC AFIS helped reduce crime rates. For example, the number of burglaries reduced 26% at the SFPD in the first 5 years of the NEC AFIS’s operation [N1], [N2].

3) The NEC AFIS was the first AFIS online connected with a digital image retrieval system (DIRS) of fingerprint images, and it was able to retrieve candidate fingerprint images quickly and helped to find suspects more easily and swiftly. The quick response time of the latent matching and image retrieval led to reduced recidivism.

Features that set this work apart from similar achievements

NEC implemented many innovations in order to improve latent matching accuracy and response time. Some of them are given below:

1) NEC implemented double scanning resolution (500-ppi scanning instead of other vendors’ 250-ppi scanning) in order to read narrow fingerprint ridges from small fingerprints (typically those of women and juveniles) and to read unclear ridges. [A3]

2) NEC implemented a new feature called “relation,” consisting of an inter-minutia ridge counting feature that other vendors did not have at that time. The relation feature improved matching accuracy for latent prints with severe plastic distortion. [A1], [A2], [A3]

3) NEC implemented a new feature called “zoning” or “local quality,” which gives the quality of a specific region of data, a feature that other vendors did not have at that time. The zoning feature improved matching accuracy by reducing the matching score of non-matching tenprints in the database (gallery). [A1]

4) NEC adopted new feature called “curvature,” which gives local ridge curving characteristics, a feature that other vendors did not have at that time. The curvature feature improved matching accuracy for small latent prints by adding additional features for comparison. [A1]

5) NEC developed a specialized fingerprint reader (FR) that scans fingerprint at 500-ppi resolution. FR has parallel-processing architecture and is equipped with a customized image processer and memory board so that it can process tenprint cards very quickly (15 seconds per tenprint card). [A3]

6) NEC developed a specialized fingerprint matching processor (FMP). FMP has parallel-processing architecture and is equipped with a customized LSI so that it can process complex (computation intensive) matching algorithms very quickly. FMP can match latent prints to fingerprints of tenprint cards at a rate of 700 comparisons per second. [A3]

7) Because of newly implemented features, the template data size became 4 times larger than that of other vendors (1000 bytes per tenprint instead of 250 bytes). The data transfer speed of these templates became a serious obstacle for high-speed matching. NEC implemented a new system architecture so that two main processors (the AFIS central controller and each FMP controller) could directly access the same hard disk drive of the database independently and simultaneously. FMP direct access to the database reduced the time required for template data transfers. [A3]

8) In DIRS, NEC implemented an optical disk system, which was new hardware in the early 1980s, to fulfill the need for quick fingerprint image retrieval without increasing hardware costs too much. The NEC AFIS was the first AFIS with a digital image retrieval function. [A3]

Significant references

News N1: Moses, K. R. Consumer’s Guide to Fingerprint Systems. Ident. News 1986, 36 (6), 5–7, 10. (Ident_News_June1986_Moses.jpg)

N2: Moses, Automated Fingerprint Identification System (AFIS). The Fingerprint Sourcebook, Chapter 6. (fingerprintsourcebook_Chap6_AFIS_Moses.pdf) https://www.ncjrs.gov/pdffiles1/nij/225320.pdf

N3: TIME No. 41 October 14, 1985 (p53) (TIME1985Oct P53.pdf)

Papers and Patents or Others A1: Asai, et al. Automatic fingerprint identification; SPIE Vol. 182 1979: (SPIE1979_AFIS_Asai.pdf)

A2: Asai, et al. Method and device for matching fingerprints with precise minutia pairs selected from coarse pairs; USP 4,646,352 (USP4646352Asai.pdf) http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=4,646,352.PN.&OS=PN/4,646,352&RS=PN/4,646,352

Supporting materials

N4: CAL-DOJ Status Report 1986 (CAL-DOJ Status Report 1986.pdf)

A3: NEC AFIS Brochure, NEC 1986