Most of the proposals to compute the Euler number of a binary image have been designed to work with images composed of squared cells. Only a few of these methods (in the case of images composed of hexagonal cells) have been reported in literature, although it is known that images composed of hexagonal cells do not suffer from the problems of connectivity frequently found in the case of images composed of squared cells. In this paper, a new way to compute the Euler number (E) of a binary image composed of hexagonal cells is presented. For this, the perimeter P of the isolated regions in the image, their contact perimeter Pc and the type T of a cell are used to obtain this important invariant. The proposal can be used alone or in combination with other features to describe any binary planar shape composed of hexagonal pixels for its further recognition.
展开▼
