Forensic 3D facial identication software (FIDENTIS)

CHALÁS, Igor, Petra URBANOVÁ, Zuzana KOTULANOVÁ, Marie JANDOVÁ, Miroslav KRÁLÍK, Barbora KOZLÍKOVÁ and Jiří SOCHOR. Forensic 3D facial identication software (FIDENTIS). In Proceedings of the 20th World Meeting of the International Association of Forensic Sciences,. 2014.

Basic information

Original name

Forensic 3D facial identication software (FIDENTIS)

Authors

CHALÁS, Igor (703 Slovakia, guarantor, belonging to the institution), Petra URBANOVÁ (203 Czech Republic, belonging to the institution), Zuzana KOTULANOVÁ (203 Czech Republic, belonging to the institution), Marie JANDOVÁ (203 Czech Republic, belonging to the institution), Miroslav KRÁLÍK (203 Czech Republic, belonging to the institution), Barbora KOZLÍKOVÁ (203 Czech Republic, belonging to the institution) and Jiří SOCHOR (203 Czech Republic, belonging to the institution)

Edition

Proceedings of the 20th World Meeting of the International Association of Forensic Sciences, 2014

---

Other information

Language

English

Type of outcome

Konferenční abstrakt

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

Czech Republic

Confidentiality degree

není předmětem státního či obchodního tajemství

RIV identification code

RIV/00216224:14330/14:00082105

Organization unit

Faculty of Informatics

Keywords in English

forensic facial analysis; software tool; 3d face database

Tags

International impact, Reviewed

---

Abstract

V originále

This presentation introduces a novel software tool for forensic 3D facial identification, called FIDENTIS. FIDENTIS is multi-platform open-source software written in Java for which source code, binaries and documentation are freely available at http://fidentis.cz. It serves for the analysis of human faces while enabling to perform a variety of expertize- and/or research-oriented tasks which can become challenging for forensic analysts dealing with facial portrait identification in the threedimensional digital environment (e.g., forensic anthropologists, forensic pathologists). These tasks include comparing two or multiple faces, quantifying the magnitude of dis/similarities between faces, creating 3D facial composites from selected parts of the faces or analyzing sets of faces in order to detect specific biological or other dependencies (e.g., inter-population variations, sex-related characteristics etc.). The integral part of FIDENTIS is formed by our own database of 3D virtual models of human faces (3D Virtual Model Database of Human Faces) which contains high resolution models of approximately 2,000 individuals acquired by optical surface scanners. FIDENTIS integrates the state-of-the-art and novel computer graphics methods and algorithms utilizing, in particular, the latest 3D technologies. The comparison of 3D faces in FIDENTIS is grounded on standard facial landmarks (up to 40 points included) in conjunction with principals of geometric morphometrics and on more complex facial surface alignments and computations. This provides the users with a high degree of freedom and flexibility, from manual to semi-automatic and automatic processes, required while examining forensic evidence. More importantly, FIDENTIS meets legal requirements that demand the use of quantifiable scientific methods with known error rates. FIDENTIS offers the users with both numerical and visual representations of outputs. In combination with other features, this gives the program the potential to become a valuable tool for forensic analysts.