指紋解析のための新データとソフトウェア(NIST Helps Fingerprint Examiners With New Data and Software Release)

2026-03-23 米国国立標準技術研究所(NIST)

米国立標準技術研究所(NIST)は、指紋鑑定の精度向上を目的とした新たなデータセットとソフトウェアを公開した。これにより、鑑定者が指紋の一致性をより客観的かつ定量的に評価できるようになる。新ツールは多数の指紋画像と統計的手法を組み合わせ、判断のばらつきを低減し、誤認リスクの軽減に寄与する。さらに、法科学分野における透明性と再現性の向上を促進し、司法判断の信頼性強化にも貢献することが期待されている。

指紋解析のための新データとソフトウェア(NIST Helps Fingerprint Examiners With New Data and Software Release)
NIST’s fingerprint dataset SD 302 includes 10,000 fingerprint images, including this one from the sticky side of a postage stamp. The dataset is now completely annotated with details such as the colorized regions at right. The colors, which represent regions of differing quality, will help train both humans and machine learning algorithms how to distinguish identifying features and weigh their importance as evidence.Credit: B. Hayes/NIST

<関連情報>

NIST特別データベース302:注釈付き遠位指骨 NIST Special Database 302 Annotated Latent Distal Phalanxes

Gregory Fiumara,Matthew Schwarz, Jessica Heising, Jennifer Peterson, Kenneth Ko, Karen Marshal
NIST Technical Note NIST TN 2367  March 2026
DOI:https://doi.org/10.6028/NIST.TN.2367

Abstract

National Institute of Standards and Technology (NIST) Special Database 302 contains 10000 latent impression distal phalanx images. This collection of data came from the Intelligence Advanced Research Projects Activity’s Nail to Nail Fingerprint Challenge. NIST worked with Certified Latent Print Examiners to annotate these images with features (e.g., minutiae, orientation, ridge quality maps) and determine their ground truth finger positions. The fully-annotated collection of data is made freely available for research purposes.

1603情報システム・データ工学
ad
ad
Follow
ad
タイトルとURLをコピーしました