Premium
Personal Identification by the Comparison of Facial Profiles: Testing the Reliability of a High‐Resolution 3D–2D Comparison Model
Author(s) -
Cattaneo Cristina,
Cantatore Angela,
Ciaffi Romina,
Gibelli Daniele,
Cigada Alfredo,
De Angelis Danilo,
Sala Remo
Publication year - 2012
Publication title -
journal of forensic sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.715
H-Index - 96
eISSN - 1556-4029
pISSN - 0022-1198
DOI - 10.1111/j.1556-4029.2011.01944.x
Subject(s) - glabella , significant difference , identification (biology) , reliability (semiconductor) , maxima and minima , point (geometry) , mathematics , computer science , artificial intelligence , statistics , computer vision , pattern recognition (psychology) , orthodontics , forehead , medicine , geometry , surgery , biology , physics , mathematical analysis , power (physics) , botany , quantum mechanics
Identification from video surveillance systems is frequently requested in forensic practice. The “3D–2D” comparison has proven to be reliable in assessing identification but still requires standardization; this study concerns the validation of the 3D–2D profile comparison. The 3D models of the faces of five individuals were compared with photographs from the same subjects as well as from another 45 individuals. The difference in area and distance between maxima (glabella, tip of nose, fore point of upper and lower lips, pogonion) and minima points (selion, subnasale, stomion, suprapogonion) were measured. The highest difference in area between the 3D model and the 2D image was between 43 and 133 mm 2 in the five matches, always greater than 157 mm 2 in mismatches; the mean distance between the points was greater than 1.96 mm in mismatches, <1.9 mm in five matches ( p < 0.05). These results indicate that this difference in areas may point toward a manner of distinguishing “correct” from “incorrect” matches.