3D ear recognition for forensics using 2D images

2018 Fiscal Year Final Research Report

• Project/Area Number: 15K00191
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Information security
• Research Institution: Saitama Institute of Technology
• Principal Investigator: WATABE DAISHI 埼玉工業大学, 工学部, 教授 (80337609)
• Research Collaborator: SAI Hideyasu ; MINAMIDANI Takanri
• Project Period (FY): 2015-04-01 – 2019-03-31
• Keywords: 生体認証 / 耳介認証 / 3D認証 / 犯罪操作支援システム

Outline of Final Research Achievements

The prominent feature of this research is to use "ear shape" which has not been researched despite its usefulness for biometrics. The applicant has conducted research on ear biometrics that is robust against changes in the angle of the ear that occurs when shaking or tilting the neck using Gabor Jet or discriminant analysis. Extending this research, we conducted the following research study:"Estimation method of images with different shooting angles","Algorithms that estimate the shooting angle of the ear","Improved ear detection and feature point detection algorithm","Algorithms that determines the absence of ear feature points, due to accidental injury, abnormal occurrence, occlusion by hair or occlusion ear itself" Summarising above, we presented 10 domestic conference presentations andpublished 6 international conference papers and 4 academic papers.

Free Research Field

画像認識、生体認証

Academic Significance and Societal Importance of the Research Achievements

本研究の特色は，生体認証にその有用性にもかかわらず研究が進んでいない「耳介」を用いることにある．耳介のパターンは個人により異なりかつ指紋や虹彩より大きいため，機器への接触なしに遠くからでも個人を識別できる可能性がある．申請者は首を左右に振るないしは傾けたときにおこる耳介の角度変化にロバストな耳介認証の研究を行ってきた．この研究を発展させ防犯カメラの耳介画像から捜査対象者リストを作成する捜査支援システムの実用化を研究目的とする．現在既に実用化されている顔画像を利用した捜査対象者リスト作成システムは横顔に弱いとされるので，耳介を利用した捜査支援システムが実用化されれば相補的に利用でき有益である．