Abstract: The method/device reconstructs/represents multidimensional objects from one- or two-dimensional image data based on recordings of one- or two-dimensional partial image areas, wherein spatial/positions of individual partial image areas are used along with one- or two-dimensional image information of the individual partial image areas to generate one- or two-dimensional image data, wherein a first group of space elements is generated in a multidimensional voxel space from first space elements which contain multidimensional image information and touch/intersect planes or lines of partial image areas by means of the one- or two-dimensional image data, and wherein a second group of space elements is generated in the multidimensional voxel space from second space elements by means of information transformation from the multidimensional image information of the first group of space elements.

Abstract: The invention relates to a method for sensing ultrasound images of an object (1). The method is characterized in that an ultra-sound head (3) is moved along the object so as to sense individual image subdomains (13) of said object (1). An evaluation subdomain (14) is analyzed in order to determine the extent of the object (1) or the spacing (&dgr;x) between the individual image subdomains (13). This evaluation subdomain (14) is sensed perpendicularly to the sensed image subdomains (13); that is, it represents a domain of the object (1) which overlaps at least two image subdomains (13). In this way, allocation within the individual image subdomains (13) can be determined by using the structures of the object (1) thus obtained. To this end, in particular a vertical column structure (11) of the object (1) (evaluation subdomain (14)) is analyzed, thus enabling allocation of a plurality of individual horizontal line structures (10), i.e.