Geometric Analysis On Craniofacial Morphology

The facial characters of human being's are the most direct basis to be recognised for each other and the characters of the skull are the intrinsic biometrics of the face. Researching on the growing laws of the face and skull, extrinsic shapes and their relationship have formed the field of craniofacial morphology. The thesis is based on the the national natural science foundation key project named "on craniofacial morphology" and focused on the geometric analysis of morphology characteristics between the skull and face,which is the most basic problem in the domain. Through related digital topology concepts being introduced into image segmentation algorithms which can obtain the surface data of face and skull with a higher accuracy, several geometric analysis methods of craniofacial morphology are presented and a prototype system of geometric analysis on craniofacial morphology is implemented. The main contributions of the thesis are as following:1) Segmentation algorithms under topological constraints of image are presented. Introduced the algebraic topology into the segmentation algorithms of image and proposed algorithm of the number of topological correlation and based on which image segmentation algorithm under topology constraints are presented and the experimental results of the algorithms are listed with their comparisons of runing time;2)Surface data of the face and skull are translated into their two dimensional parameter space and geometric analysis methods are presented in the space. Aaccording to three dimensional surface of face and skull data been seen as two dimensional smooth manifold of embedding in the three dimensional Euclid space, algorithms of translating the three dimensional surface data of face and skull into their two dimensional parameter space are presented and the methods of geometric analysis relationship on the face and skull are also presented in space with their corresponding error analysis;3)Geometric analysis methods of the face and skull are proposed under the landmark on them. Based on the three dimensional surface data of the face and skull, methods of geometric analysis on them are presented through these parameters with their numerical calculating methods under the landmarks of the anthropologist and anthropometrist are proposed,which is the foundation of the statistical model that will be used in the field;4)Geometric analysis methods that based on the approximation coefficients of spherical harmonic functions are proposed. Against the facts that the surface of the face and skull in three dimension Euclid space are more complex, the representations of face and skull model and methods of geometric analysis on them are presented which are based on the theories of spherical harmonic functions by their approximation coefficients and the experimental results are also showed;5) Prototype system of geometric analysis on the craniofacial morphology is accomplished. Combined the two open source software packages,VTK(Visualization ToolKit) with CGAL(Computational Geometry Algorithm Library),use visual C++ 6.0 programming language as integration tool to complete the design of prototype system and algorithms that are presented are implemented in the thesis and the results are verified to be correct,effective and more excellent.