For example,Бобцов

SHAPE RESTORATION OF THREE-DIMENSIONAL OBJECT FROM TWO-DIMENSIONAL IMAGES

Annotation

The paper presents a new method of the object shape restoration based on the tomography approach in signal processing. Most existing methods of solving the problem require either an active surveillance system that besides images registration includes special illuminating of the object, or it is necessary to identify corresponding points in the case of stereo images, sometimes also with illuminating. The proposed method requires neither of that. Instead, it is enough to allocate the area of the object image on each image plane. It is also necessary to know the image planes positions. The method essence is the following. We assume that the observed space area that includes the object is known. We divide it into a set of small volumes, resolution elements (ReE), and number them. The observed space area is described by the vector-original with the amount of components equal to the amount of ReE and the values of components equal to 1 if the ReE with the same number is inside the object and equal to 0, if  it is not inside it. Obviously, if we know such a vector, it is quite simple to receive a three-dimensional object, forming it from ReE with numbers equal to the numbers of the vector-original elements equal to one. Thus, the main stage of solving the problem is the restoration of the vector-original by its mapping. The problem is solved with the aid of   logic algebra. The mapping is being formed from binary images. The logical expressions have been obtained that give the possibility to reconstruct the vector-original; computing simulation of the method realization has been performed, and as a result the object shape has been obtained as the set of ReE inside it, that makes it possible to consider the restored object in all angles.

Keywords

Articles in current issue