Методи, моделі та система формування тривимірних зображень облич людей для реконструктивної та пластичної медицини

Abstract

Facial injuries are the most common injuries with a constant tendency to grow. This leads to an increase in the role of plastic and reconstructive medicine. Given the persistence of the dynamics of traumatism, the issues of improving diagnosis, improving the efficiency of plastic and reconstructive operations are urgent. Increasing the level of implementation and planning such operations is associated with the use of three-dimensional modeling. Three-dimensional models of human face are the most realistic, reflect the anatomical structure, accurately convey the relief and color features of the object, subject to modification to change the appearance. The 3D face model is a multifactorial source of patient information, allowing you to significantly reduce the amount of user interaction required compared to the existing methods. In medical practice, a three-dimensional model of the face image is distinctive from other applications because it is the object of study for diagnosis, and therefore must be accurate and allow for different metrics. In this case, the combination of anthropometric data with other methods of research allows to clarify the diagnosis to a great extent. Today, many diseases can be diagnosed on the basis of the analysis of the ratios of different areas of the face. When shaping colors for different areas of the face, it is important to realistically reflect the diffuse, specular components of the intensity of the color. It is necessary to take into account the dependence of the reflection coefficients on age-related changes in the person, offset structure of the face surface. An important requirement for 3D rendering techniques is high performance, which should be sufficient to support dynamic and interactive modes. Therefore, the development of high-performance methods, models and tools for forming three-dimensional images of people's faces for planning and conducting plastic and reconstructive operations is an urgent task. The thesis develop the theory for the construction of medical devices and systems through the development of methods, models and means of forming threedimensional images of human faces for plastic and reconstructive medicine. There had been developed physically correct distributive functions of the surface reflectivity, the peculiarity of which is the observance of the law of conservation of energy in the formation of face images. This allowed to increase an accuracy of determining the intensities of colors in the formation of face images. New modifications of the Schlick model of surface reflectivity for forming the image of a person's face have been developed, which made it possible to increase realism by more accurate determining the specular component of color. Further development was envisaged for a cosine model of reflectivity of the surface of a person's face, in which, unlike the classical, elevation was used to a lesser degree, which allowed to increase the productivity of forming the image of the face with high accuracy of reproduction of both the epicenter and bluming of flare. An adaptive method of coloring a person's face has been developed, the peculiarity of which is an adaptive use for determining the intensities of color in different parts of the surface of the face image of polynomials of the second and third degrees. This made it more realistic to create 3D images of people's faces in highperformance systems. There had been suggested the method of reconstruction of the human face according to the data of the stereo-pair, in which, unlike the existing ones, analytical perturbation functions were used, which allowed to reduce the amount of memory required to store the geometric model, to simplify its scaling and to use the Ray tracing method for visualization. Analytical dependences of the speculative coefficient of the skin of a person's face for different age groups were obtained, which allowed to increase the realistic color reproduction and to change the details of the polygonal network of the model in the problem areas. To reproduce the offset properties of the skin, new modifications of the Cook- Torrence and Ward models had been suggested. Compared to analogues, only one function and smaller polynomials are used for the calculation. This made it possible to improve the imaging performance of human faces. It is suggested to use the morphing of 3D images for planning the plastic operations in order to determine the optimal shape of the selected areas of the face, as well as to establish the age-related changes in human development to the established standards. On the basis of the suggested methods and models there had been suggested the means and systems for the formation of 3D-image of faces for plastic and reconstructive medicine.