Projective computer tomography and image reconstruction of discrete structure of materials

Projective computer tomography and image reconstruction of discrete structure of materials Computer tomography is efficiently used in different applied areas, particularly for investigation of the material's structure. Traditional approach requires scanning the object with a lot of angles and reconstruction of images, using sufficiently difficult mathematical operations as, for example, the inversion of the Radon transform in X-rays tomography. It requires sufficiently large volume of calculations, but it is necessary for object with difficult structures. We consider in this paper the object with the simple " discrete " structure, i.e., the structure that consists of some separate elements inside of the homogeneous (or quasi homogeneous) substance in the considered domain. These types of structures appear at investigation of defectoscopy of materials, particularly at production of semiconductor's plate and composite materials, so as in other applied areas. We propose for these types of structures the " Projective Computer Tomography " (PCT), which consists in localization of intersections of prolongation of some known projections, obtained as images by some concrete tomography. To realize numerically PCT we develop here the three dimensional generalization of the proposed before the plane scheme of Rotating Projection Algorithm (RPA). It simplifies the image reconstruction of discrete structures, because RPA does not require numerical realization of complicate formulas, such as, for example, the inverse Radon transform in X-rays tomography. Constructed algorithms are realized as program package in MATLAB system, which quality is demonstrated on simulated numerical examples. aplicadas. Proponemos para este tipo de estructuras la " To-mografía Computacional Proyectiva " (TCP), la cual consiste en la localización de intersecciones de prolongaciones de algunas proyecciones conocidas. A realizar numéricamente TCP desarrollamos aquí la generalización tres dimensional del esquema plano del Algoritmo Rotatorio Proyectivo