Facial modeling and measurement based upon homologous topographical features

Area of Science:

• Anthropometry
• Biometrics
• Medical Imaging

Background:

• Human face measurement is crucial for applications like recognition and genetic phenotyping.
• Conventional anthropometric landmarks face challenges in homology with digital scans.
• Existing methods lack information density and struggle with precise facial topography analysis.

Purpose of the Study:

• To introduce a novel, topography-based parametric model for human facial measurement.
• To provide a richer information density and derive homology from shared facial features.
• To quantify local morphological variation and compare it with conventional methods.

Main Methods:

• Developed a 71-parameter parametric model to capture facial variations.
• Modeled a sample of 80 adult human faces using photogrammetric surface meshes.
• Compared the parametric model's analysis with landmark-based geometric morphometrics (GMM).

Main Results:

• The parametric model accurately matched facial scans to within 1 mm.
• Multivariate analysis of 71 attribute values reliably separated geographic ancestry and sex groups.
• Results showed broad similarity to GMM but with key differences in quantifying facial shape and variation.

Conclusions:

• The topography-based parametric model offers an efficient and information-dense approach to facial shape encoding.
• This method demonstrates potential for phenotypic and dysmorphology studies.
• It provides a robust alternative to conventional landmark-based and non-parametric methods.