G. Iliev DMD*, J. Pavlova DMD, PhD**, D. Filtchev DMD, PhD**,

K. Haralampiev**, B. Yankov PhD***, S. Rangelov***, B. Kocheva***,

M. Dimova***, S. Nasiv-Macci***, A. Filchev DMD, PhD, DSc****

Purpose: To analyse main features of the four facial types – strong, dynamic, delicate and calm (according to the adap­ tated classification of Hippocrates) and by mathematical modeling to be generated digital face maps for facial type determination.

Material nad Method: 91 Full face pictures with maximum smile were made in standardized conditions from them 42 male and 49 women of age between 18 and 30 years old.

The faces were categorized in four main types: strong, dy­ namic, delicate and calm (according to the adaptated clas­ sification of Hippocrates). Four basic elements which car­ acterize the face were determined: Face contour, eyes, nose/ eyebrows, mouth and each of them determine 25 % of the main type. By the principal of kefaloscopy each element of the face was classified to one of the four types.

Each picture was analysed by a software „VisagiSMile“. By specially developed module (Analysis of the face – map se- lection) by marking of 27 facial landmark points were gen­ erated digital facial maps for each face. 12 basic facial sec­ tions were defined. Based on all these data were generated mathematical models which describes each type.

Results: Facial characteristic of each individual is a combi­ nation of features typical for many types with different level of dominance over one another. In 64,3 % of the men and 59,2 % of the women were established features of two types,

while in 35,7 % of the men and 40,8 % of the women – three types. Faces only with one type or with features of four types wasn’t established.

There were generated averaged digital facial masks, de­ scribing each of four main types

Conclusions:

1. By mathematical modeling were generated four digital fa­ cial maps for each of the main facila types: strong, dynamic, delicate and calm
2. Generated average digital facial masks could be used for digital facial analysis with the algorithm C5.1 – classifica­ tion tree with confidence to 84,337 % of determined by them facial type.
3. Created software compares each feature to the reference maps and calculates which shape fits best. By using the re­ sults for all features, it classifies the type of the whole face.

Key words: digital face map, mathematical modeling, facial type

Download the full paper.