Anthropological Facial Approximation in Three Dimensions (AFA3D): Computer‐Assisted Estimation of the Facial Morphology Using Geometric Morphometrics

Abstract

This study presents Anthropological Facial Approximation in Three Dimensions (AFA3D), a new computerized method for estimating face shape based on computed tomography (CT) scans of 500 French individuals. Facial soft tissue depths are estimated based on age, sex, corpulence, and craniometrics, and projected using reference planes to obtain the global facial appearance. Position and shape of the eyes, nose, mouth, and ears are inferred from cranial landmarks through geometric morphometrics. The 100 estimated cutaneous landmarks are then used to warp a generic face to the target facial approximation. A validation by re-sampling on a subsample demonstrated an average accuracy of c. 4 mm for the overall face. The resulting approximation is an objective probable facial shape, but is also synthetic (i.e., without texture), and therefore needs to be enhanced artistically prior to its use in forensic cases. AFA3D, integrated in the TIVMI software, is available freely for further testing.

Facial Landmarks for Forensic Skull-Based 3D Face Reconstruction: A Literature Review

“…Analysis of FSTT [8] in correspondence to landmark locations showed that males generally had greater values than females, and that Body Mass Index (BMI) had strong influence of FSTT. Similarly, Guyomarc'h et al indicated that corpulence had the major effect on soft-tissue thicknesses, compared to age, sex, and ethnicity [13]. Body mass and weight have also be taken into consideration by Parks et al, who, given the documented prevalence of obesity in U.S.A. area, purposely assembled U.S. population to consider escalating weight of them [10].…”

“…Tissue depths were estimated also by Guyomarc'h et al depending on age, sex, and corpulence by working on 87 hard-tissue and 102 soft-tissue landmarks. Among the skin landmarks adopted for this study, some were manually positioned on face, others were projected on the skin from the corresponding osseous landmark [13]. Parks et al, supported by Federal Bureau of Investigation (FBI) Laboratory Division, aimed at building a detailed tissue depth database among 25 landmarks and compared it with previous relevant studies [10].…”

“…Manual localization methods are digitization [7,29], physical markers placing [15], and software-aided localization [10]. Related supporting tools and software are Microscribe G2X digitizer [29], Mimics 14.11 [10], Amira 5.2.2 [8], MorphoJ [13], VRMesh [16], MakeHuman [13].…”

“…The number of observers positioning landmarks in this context is always single user-based [7,8,13,29]. Thus, no intra-nor inter-individual errors are computed.…”

“…Thus, no intra-nor inter-individual errors are computed. Nonetheless, even if only one user makes the allocation, different observations are registered sometimes: two [16], three [7], five [13,29]. And a time period can pass between different localizations: one week [13] or four weeks [16].…”