The invention relates to an assembly for the selective three-dimensional or two-dimensional representation of images. The inventive assembly comprises an image reproduction unit (1) with a plurality of image elements, which in a predetermined allocation represent information from one or more views of a scene, object or text, a filter array (2) that is located behind the image reproduction unit (1) in the line of vision (B) of an observer, a first scattering layer (3) that is located behind the image reproduction unit (1) and in front of the filter array (2) in the line of vision of the observer, said layer being switched back and forth between a transparent condition and a scattering condition and a second scattering layer (4) that is located in front of the image reproduction unit (1) in the line of vision (B) of the observer and lies directly on said unit, the layer corresponding preferably to an anti-glare matt layer. The filter elements are arranged in such a way that defined expansion directions are predetermined for the light that is emitted by the image reproduction unit (1) in the transparent condition of the first scattering layer (3), said directions remaining essentially unaffected by the second scattering layer (4) and the structuring or the light that passes through the filter array (2) in the scattering condition of the first scattering layer (3) is reduced in relation to the first condition.

An autostereoscopic projection arrangement, comprising at least one projector and at least one filter array which has a multitude of filter elements arranged in columns and rows, in which arrangement bits of partial information from views of a scene or object are projected by the projector(s) onto a projection screen, where there bits of partial information are rendered on image rendering elements and, having passed one or several of the filter arrays, are made visible to at least one observer, and in which, as regards to propagation direction of the bits of partial information, the image rendering elements correspond with correlated filter elements in such a way that an observer will see predominantly bits of partial information from a first selection of views with one eye and predominantly bits of partial information from a second selection of views with the other eye, and thus will have a spatial impression.

A method of spatial visualization in which several views are decomposed into bits of partial information, which are made optically visible by image rendering elements. Neighboring image rendering elements emit light of different wavelengths or wavelength ranges. Propagation directions for the light are given by wavelength filters such that an observer will see predominantly bits of partial information of a first selection of views with one eye and predominantly bits of partial information of a second selection with the other eye. At least one image rendering element is simultaneously allocated bits of partial information from at least two different views. The allocation is made in such a way that the wavelength of the partial information is always equal to the wavelength, or lies in the wavelength range, of the light emitted by the allocated image rendering element.