Kermi, Adel (2008) Reconstructions faciales à partir d’images tridimensionnelles de crânes humains par recalage et modèle déformable pour l’identification de personnes. PhD thesis signal et images, Département Traitement du Signal et des Images, ENST.

Full text available as: - Thesis_KERMI_Adel.pdf ( 7503 Kb )

Licence: Copyright

Abstract

Facial reconstruction from a cranial skeleton is an important technique in several scientific areas, especially in forensic sciences, archaeology, and palaeontology. It is used for skull identification and recognition of people. It can be classified, among reconstructive identification methods. It is most often used as a last resort, when all other techniques fail to produce a person presumed identity.

This thesis addresses the facial reconstruction problem from three-dimensional (3D) images of human skulls, considered as unknown. We present a 3D computerized facial reconstruction method based on recent techniques of medical imaging, where the main objectives are the processing speed and the elimination of the subjectivity by leaning particularly on mathematical criteria for evaluating the results.

Our method is based on a model deformable approach constrained by the knowledge of soft tissues thicknesses, at a certain number of characteristic landmarks. It uses, for each facial reconstruction, a 3D image of a reference head from which skull and skin are extracted, and a 3D image of the skull of the unknown head for which the skin should be reconstructed. The facial reconstruction process is divided into two main stages. An initialization step of the deformable model, based on a non-linear registration technique guided by B-Spline Free-Form Deformations (FFD) model. We therefore, propose an automatic initialization that is achieved only from a reference skull and skin set as well as an unknown skull. At first, a transformation of the reference skull image into the unknown skull image is computed. Then, we apply the same transformation to deform the reference skin image in order to obtain a new skin, which is considered close to the unknown skin and that will be used as initialization of the final facial reconstruction. In the second step, the initial skin, resulting from the transformation computed previously, is refined by means of a 3D deformable model based on simplexes meshes that is attracted by some characteristic landmarks previously computed by a calculation of Mean and Gaussian curvatures. The landmarks are fixed according the positions of the anthropological landmarks of Rhine and Campbell [Rhine and Campbell, 1980]. The evolution of our deformable model is done according to different internal and external forces among which the Gradient Vectors Flow force (GVF) and a balloon force.

This method has been tested on thirteen sets of skull/skin data, extracted from 3D-MRI of individual heads of children and adults. For each facial reconstruction, two skull/skin sets corresponding to a reference head and a head with unknown skin are selected according to similar anthropological characteristics. By this method, encouraging results are obtained. The reconstructed shapes remain faces, visibly acceptable, and are relatively close to the real faces.

Statistiques de consultation

Repository Staff Only: edit this item