DE10043305A1 - Three-dimensional image representation method uses grid of individual image elements representing partial information for scene/object from different directions - Google Patents

Abstract

The image representation method has a large number of individual image elements visualized simultaneously in a grid consisting of columns and lines, each image element representing partial information of a scene/object from several views, with adjacent image elements providing light of different wavelengths or wavelength ranges. Also included are Independent claims for the following: (a) a three-dimensional image representation device; (b) a manufacturing method for a wavelength filter array

Description

The invention relates to autostereoscopic arrangements for several at the same time Viewer in which a spatial scene is multi-channel, i.e. H. with at least four views, preferably perspective views. To enable a spatial Perception, an optical component, preferably a filter array, is provided, which according to the invention corresponds to the eye position of a selected one Is laterally displaced. This allows the selected viewer to see the essential always a certain selection of views, z. B. the middle views, d. H. the spatial impression is created primarily for a constant perspective. Farther is the inertia of the displacement device for the multi-channel arrangement the optical component is substantially balanced, which makes the selected viewer essentially never perceives a pseudoscopic picture - even if it does moved quite quickly. All other viewers see when the selected one moves Viewer an image that changes slightly in perspective.

Such arrangements are suitable for applications in which a special Viewers (e.g. a surgeon) must always see an orthoscopic 3D image while others Viewers (e.g. students) have the same 3D image content, but from slightly changing ones Can see perspectives.

On the one hand, the prior art includes the two-channel methods for spatial representation known. A key disadvantage here, however, is that essentially only one viewer gets a correct, in depth, orthoscopic image can see, while this cannot be guaranteed for other viewers. Versions without allow two-channel processes and arrangements Eye position tracking of the beholder almost no movement of the beholder since the Stereo half images defined can only be perceived from predefined directions. In Versions with eye position tracking of the viewer will be a certain one Freedom of movement granted, although rapid movements due to the mechanical inertia, in particular the displacement device for a required optical Component (e.g. a prism or lens grid) briefly to pseudoscopic  Lead image perceptions. This is e.g. B. the case with arrangements according to DE 197 36 035, if the optical component (and not the image content) corresponds to the position of the Viewer is tracked. In addition, arrangements of this type allow for tracking Lenticular or prismatic grids are not, large-format screens (e.g. plasma displays with 50 " Screen diagonal and larger) sufficient to assemble as a 3D screen, because then the lenticular or prismatic grid i. d. R. have such a mass and thus inertia that a the lens or prism grid is not sufficiently flexible to move.

Furthermore, arrangements and methods are known in which the. Without any tracking scenery to be displayed simultaneously in several views, e.g. B. eight views, is reproduced. Optical components (e.g. wavelength filter arrays, prism or Lenticular grid, barrier screens) ensure that several viewers at the same time can usually see an orthoscopic 3D image from different positions. It is coming however, this usually leads to so-called panaroma jumps, in which the transition from Z. B. the last one takes place to the first view (s). at The corresponding viewer sees viewing positions near these panoramic jumps possibly also a pseudoscopic picture. This can be done by a slight lateral movement Deficiency, however, can be easily cured.

A fairly high-quality example of a multi-channel system based on lenticular screens (Lenticulars) is described in EP 0791847.

Proceeding from this, the object of the invention is an autostereoscopic To enable representation that is always an orthoscopic for at least one viewer 3D image mostly presented from the same perspective while a group of others Viewers mostly see an orthoscopic 3D image, showing the respective perspective that different viewers can change. In a special embodiment, it should also enable large-format screens to meet the aforementioned Convert task.

This object is achieved according to the invention by an arrangement for the spatial representation of a scene / an object, comprising

• - An image reproduction device, in particular a flat screen ( 1 ), for reproducing an image array from a plurality of image elements α _ij , which, in a defined association, contains information from several views A _k (k = 1... n with n <3) of the scene / Represent an object, light being emitted from the picture elements in preferably different wavelength ranges,
• - At least one filter array ( 2 ), which is based on the viewing direction of a viewer ( 5 ) of the image display device upstream or downstream at a predetermined distance z, the filter array comprising a plurality of filter elements that are translucent in certain wavelength ranges, so that for the Image elements α _ij emitted light propagation directions are specified,
• a displacement device ( 6 ) which is able to move the filter array ( 2 ) - preferably in the horizontal direction -
• - a tracking device ( 7 ) for detecting the eye position of a selected viewer ( 5 ),
• - a computing device ( 8 ) which actuates the displacement device ( 6 ) as a function of the eye position of the selected viewer ( 5 ) determined by the tracking device,
• - Wherein at each position of the filter array ( 2 ) the directions of propagation within a viewing space, in a plurality of intersection points, each corresponding to a viewing position, cross, whereby from each viewing position the viewer (s) with one eye predominantly partial information of a first selection and with the other eye predominantly optically perceive partial information from a second selection from the views A _k (k = 1... n), the selected viewer ( 5 ) in particular, almost exclusively the middle one, due to the position of the filter array that is permanently adjusted to his eye position Views A _k (k = 1... N) sees and, due to the use of n <3 views, the inertia of the displacement device is compensated in such a way that the selected viewer ( 5 ) essentially never sees an image that is inverted in depth perceives.

The invention can be embodied in many different ways; in particular, the filter array ( 2 ) can also be designed as a lenticular screen, specifically as a lenticular screen or also as a barrier screen.

For the further explanations, however, the configuration with a Filter array can be considered.  

A flat screen ( 1 ), preferably a color LC display, with separately controllable subpixels is provided as the image display device. One sub-pixel corresponds to one picture element α _ij in columns i and rows j.

The flat screen ( 1 ) is preferably coupled to a control circuit ( 3 ) which generates partial information of the views A _k (k = 1 to n) on the subpixels according to the function

in which

• i corresponds to the index of a picture element α _ij in one line of the grid,
• j corresponds to the index of a picture element α _ij in a column of the grid,
• - k the consecutive number of the view A _k (k = 1... n) from which the partial information originating from a certain picture element α _ij originates,
• - n the total number of views A _{k used} (k = 1... n),
• - c _{ij of} a selectable coefficient matrix for combining or mixing the various partial information on the grid from the views A _k (k = 1... n)
• - IntegerPart of a function for generating the largest integer that the argument in brackets does not exceed.

In other words: the indices i and j denote the positions of picture elements α _ij for which it is to be specified from which of the views A _k (k = 1... N) the partial information to be displayed is to be obtained. Here i stands for the horizontal index (with values from 1 to the horizontal picture element resolution, in the case of displaying the partial information on RGB subpixels that is three times the value of the pixel resolution) and j for the vertical index (with values from 1 to the value of vertical picture element resolution).

For any, but fixed, number n of views A _k (k = 1... N), all of which have the same image resolution or the same format, is the partial information of the views A _k (k = 1 ... n) to be combined, the following must also be taken into account for the combination rule: the coefficient matrix c _ij can have values as entries which correspond to real numbers.

Natural numbers greater than zero for i and j are in the range of values mentioned above possible.

The overall image shown on the grid and combined from the different partial information of the views A _k (k = 1... N) is generated when these parameters are specified in accordance with the above-mentioned function by running through all possible index pairs i, j.

The filter array ( 2 ) contains filters β _pq , which are arranged according to the function in columns p and rows q

With

• p the index of a filter β _pq in a row of the array,
• q the index of a filter β _pq in a column of the array,
• b an integer which defines one of the intended transmission properties λ _b for a filter β _pq at position p, q and can have values between 1 and b _max ,
• n _{m is} an integer value greater than zero, which preferably corresponds to the total number b _{max of} the transmission properties λ _b , the total number b _max again preferably corresponding to the total number n of the views A _k shown in the combination image,
• - d _{pq of} a selectable coefficient matrix for varying the structure generated on the filter array and
• - IntegerPart of a function for generating the largest integer that the argument in brackets does not exceed.

The selectable coefficient matrix d _pq can have values as entries which correspond to real numbers. Natural numbers greater than zero are possible for p and q, which (as already described) describe positions within the filter array.

The generation of the combined overall image from the partial information of the views A _k (k = 1... N) and the generation of the mask image, ie the structure of the filter on the filter array, are therefore based on similar or at least related regulations. The filters β _pq as elements of the mask image have an area similar in size to the image elements α _ij .

The transmission properties λ _b for the filters β _pq preferably correspond to transparency wavelength or transparency wavelength _ranges .

These transparency wavelength ranges are preferably specified such that they each reflect the transparency in one of the three color ranges red (λ ₁ ), green (λ ₂ ) or blue (λ ₃ ) or the transparency in the entire visible spectrum (λ ₄ ) or the lack of transparency in the whole visible spectrum (λ ₅ ).

Alternatively, it is also possible to specify the transmission properties λ _b as transparency wavelength ranges, each of which corresponds to the transparency in one of the three color ranges cyan, magenta or yellow or the transparency in the entire visible spectrum or the non-transparency in the entire visible spectrum.

In a special embodiment, it can also be provided to specify purely wavelength-independent transmittance as transmission properties, ie the filters in this case would be neutral filters for wavelength-independent attenuation of the light intensity with a predetermined transmittance (e.g. 0%, 25%, 50%, 75% and 100%). Further developments are also conceivable, ie the filter array ( 2 ) contains both wavelength filters and neutral filters.

Various arrangements can be used to detect the eye position of a selected viewer ("tracking"). In a simple embodiment, a pivoting camera with a computing device is used as the tracking device ( 7 ). This is in the computing device ( 8 ) by the tracking device ( 7 ), for. B. the camera, delivered signal evaluated by z. B. the eyes are searched as white areas with dark spots that correspond to the pupils. It is also possible to mark the selected viewer preferably between the eyes with a specially colored point or another, technically recognizable feature, which also supports easy tracking of the eye position. In the case of larger screens (e.g. plasma displays with a diagonal of 50 inches), the selected viewer ( 5 ) can also simply be characterized by being in a certain area - precisely the available tracking volume of the tracking device ( 7 ) - stops.

A variety of possible eye position tracking devices is in WO 9618925 described.

A displacement device ( 6 ) is provided for displacing the filter array ( 2 ), which comprises at least one rail and at least one linear motor, which is preferably designed as a stepper motor. The filter array ( 2 ) is preferably displaced in the horizontal direction - in relation to the (usual) viewing direction of an observer.

It is also possible to design the displacement device with the aid of piezo actuators.

In the viewing direction of a viewer, the filter array ( 2 ) can be arranged between the flat screen ( 1 ) and a planar lighting source downstream of it, which emits white light.

The filter array ( 2 ) is preferably arranged in the viewing direction of a viewer in front of the flat screen ( 1 ), which can be followed by a flat, white light-emitting illumination source.

The individual filters of the filter array ( 2 ) each have a polygonal, preferably rectangular outline.

The distance z between the flat screen ( 1 ) and the filter array ( 2 ) can, for. B. be 1 to 10 mm.

Under certain circumstances, it may be advantageous if a lens, preferably a Fresnel lens, is arranged upstream of the flat screen unit ( 1 ) and filter array ( 2 ) in the viewing direction of a viewer ( 5 ), which, depending on the design of the lens, results in a real and / or virtual image the spatial representation of the scene / object for the viewer.

In an advantageous embodiment of the invention, the filters on the array are not only in the Statically, with regard to the transmission characteristics of their individual wavelength filters still as passive, d. H. without any electrical energy filters educated. Filter arrays can be manufactured as thin-walled plates or foils which the desired structure of the individual filters is permanently imprinted.

In a particular embodiment, the filter array (2) is formed by an LCD, wherein the requirement for a shift means (6) associated with calculating means (8) in place of a displacement means (6) said LC display in dependence on the signal of the tracking device ( 7 ) controls. The displacement device ( 6 ) can be omitted. The individual sub-pixels of the LC display act here as filters, which are controlled separately. Black subpixels correspond to e.g. B. opaque filters, while switched optimally transparent, hardware z. B. provided with a green filter, for example, act as a green filter. Equivalent to shifting the filter array, the LC display is controlled here in such a way that the translucent subpixels of the LC display are shifted in response to a shift of a static filter array when required, ie when the viewer moves.

In other words, instead of moving a static filter array, it will Filter array suitably shifted according to the "filter image" shown on the LC display.

In general, a picture element α _{ij is} a self-illuminating or illuminated area of small size with an area of approximately 10,000 µm ² to a few mm ² , on which a small section of one of the views A _k (k = 1... N) is shown below referred to as partial information of such a view A _k (k = 1... n) at which i, j can be reproduced. Advantageously, the index pair i, j can also be used to denote the position in the view A _k from which the partial information originating from the picture element α _ij originates - provided the views A _k (k = 1... N) in Grids of rows j and columns i are structured.

From each viewing position, a viewer perceives predominantly picture elements α _{ij from} a first selection and with the other eye predominantly picture elements α _{ij from} a second selection from the views A _k (k = 1... N).

This also expressly means cases in which the light coming from a picture element α _ij does not reach the eye of the viewer in question completely, but only in part, which is conceivable, for example, when a picture element α _{ij is} partially covered, for example by arrangement parts ,

The directions of propagation are preferably predetermined by a multiplicity of filters β _{pq of} the filter array, one image element α _ij corresponding to several assigned filters β _pq or one filter β _pq corresponding to several assigned image elements α _ij such that the connecting straight line between the center of the visible area Section of a picture element α _ij and the area center of the visible section of a filter β _pq corresponds to a direction of propagation.

It is advantageous if the filters β _{pq are arranged} in arrays with rows q and columns p and one or more such arrays advance and / or the grid with the picture elements α _ij , based on the viewing direction of a viewer, at a fixed distance z. or are subordinate.

The invention is illustrated below with reference to the drawing Embodiments explained in more detail. The drawing shows in:

Fig. 1 is a schematic sketch of a first embodiment of an arrangement for the spatial representation of a scene / an object according to the invention with a filter array lying in the viewing direction of an observer in front of an image display device (color LC displays), which can be moved laterally by a displacement device .

Fig. 2 shows an example of the structure of the filter β _{pq of} the filter array using wavelength filters in a greatly enlarged and not to scale, and in

Fig. 3 shows an example of a total image from the partial information of the views A _k (k = 1... N; n = 8), generated with the sub-pixels RGB of the color LC display used as picture elements α _ij , greatly enlarged and not - shown to scale.

The first embodiment shows an arrangement for three-dimensional (autostereoscopic) representation of scenes and / or objects. Be with this Partial information from different views of the scene or object played. The views represent image information, for example neighboring perspective views.

As an image display device for displaying the combination of partial information of the views Ak (k = 1... N) on picture elements αu, the arrangement comprises a color LC display ( 1 ), such as Batron BT63212, which is currently commercially available. In this way, the arrangement according to the invention can be easily implemented. However, this does not exclude that any other conceivable version is possible for image reproduction. In particular, a plasma screen, for example of the type Pioneer PDP 502-MXE, could also be used instead of the color LC display.

In the viewing direction of one or more viewers, the color LC display ( 1 ) is arranged as an image-forming grid at a predetermined distance z behind a filter array ( 2 ). Behind the color LC display is a planar lighting source, which is connected to the color LC display ( 1 ) to form a structural unit and is not shown in the drawing.

The image-reproducing color LC display ( 1 ) is linked to a control circuit ( 3 ). It has separately controllable sub-pixels of the primary colors red (R), green (G) and blue (B). For better differentiability, the sub-pixels of the color LC display ( 1 ) are referred to below as R, G, B; the wavelength filters of the filter arrays ( 2 ), however, are denoted by R ', G', B ', each of which corresponds to its own transparency wavelength range. The control circuit ( 3 ) is designed such that partial information of the views A _k (k = 1... N) is generated on the individual subpixels R, G, B of the display ( 1 ).

The distance z between the color LC display ( 1 ) and the filter array ( 2 ) is, for example, 2.3 mm, in which case wavelength filters R ', G', B 'of the filter array ( 2 ) with the subpixels R, G , B of the color LC display ( 1 ) correspond in such a way that the directions of propagation, thus determined, of the light emerging from the subpixels R, G, B and passing through the filters are within a viewing space ( 4 ) in which one or more observers , in particular a selected viewer ( 5 ), meet at a plurality of intersection points. These points of intersection of the directions of propagation correspond to observation locations from which the scene or the object can be perceived spatially with one pair of eyes.

The distance z for the design variant according to FIG. 1 was determined from

The average horizontal spacing of the individual subpixels on the color LC display ( 1 ) arranged downstream in the viewing direction was assumed to be 99 µm for s _p . 65 mm was used for the mean pupil distance p _d . 1.5 m was chosen as the mean viewing distance. The resulting distance z is 2.3 mm.

The directions of propagation are in each case through the area centers of the visible sections of the filters and sub-pixels R, G, B in question, the beam paths not spread out only on one level, but often spatially distributed.

Furthermore, the displacement device ( 6 ) for the filter array ( 2 ) is indicated schematically in FIG. 1. In an exemplary embodiment, it is formed from a stepper motor (with a corresponding control) and rails for the filter array ( 2 ). The filter array ( 2 ) is on a support, for. B. laminated on a sheet of glass. This glass pane with the filter array ( 2 ) is guided on its upper edge or lower edge in a rail, whereby additional means for reducing the friction (e.g. rollers) can be provided. The tracking device ( 7 ), not shown in the drawing, for real-time detection of the eye position of a selected viewer ( 5 ) provides information for the control unit of the displacement device ( 6 ) via a computing device ( 8 ). This moves the filter array ( 2 ) continuously with the aim that the selected viewer always perceives the same, pair-wise different views A _x and A _y from the views A _k (k = 1... N). The inertia or hysteresis of this control loop consisting of tracking ( 7 ) and shifting device ( 6 ) is now compensated for by the fact that if the filter array ( 2 ) shifts too slowly or overshoots, compared to the precisely aimed target position in each case, the one besides the middle one mentioned Views A _x and A _y still give a correct 3D impression for the selected viewer ( 5 ). This can be particularly advantageous if large-format screens (e.g. a plasma display of the above type) are to be assembled as 3D screens, because in these cases the mass and thus the inertia of the filter array ( 2 ) and the Inertia or hysteresis of said control loops can be particularly large.

While the selected viewer ( 5 ) then essentially always sees the views A _x and A _{y mentioned} , the other viewers can generally also perceive an orthoscopic autostereoscopic 3D image. In the event that the selected viewer ( 5 ) moves, only the perspective of the other viewers changes. In the event that a viewer thereby enters the transition zone of the views, the so-called jump point, at which the period of the horizontally adjacent views begins anew, ie in the transition area of the last view A _n to the first view A ₁ , such a viewer can return to an orthoscopic viewing area with a slight lateral head movement.

In FIG. 2, an example of the structure of a filter array is β _pq based on individual filters in a plan view of the display surface for the purpose of clarity, greatly enlarged and not drawn to scale. The partial areas each correspond to a filter. As already mentioned, R 'corresponds to a filter which is transparent to red light, G' to a filter which is transparent to green light and B 'to a filter which is transparent to blue light. S denotes opaque filters. This filter structure can be generated with the exemplary filter combination parameter d _pq , which is generated according to the regulation

The filters are preferably rectangular and have, for example, a width of 99 µm and a height of 297 µm.

FIG. 3 shows an example of the combination of partial information from different views A _k (k = 1... 8) in a top view of the grid of the color LC display 1 , which was generated according to the function already described

where c _ij = -1 = const. are selected for all pairs (ij) and n = 8.

Each rectangular partial area corresponds to a sub-pixel R, G, B. The numbers k = 1 specified within the partial areas. , , 8 indicate the respective view A _k (k = 1... N) to which the partial information displayed on a subpixel or a picture element α _ij belongs. Thus, partial information displayed on a subpixel labeled k = 1 belongs to view A ₁ , partial information displayed on a subpixel labeled k = 2 belongs to view A ₂ , and so on. In the selected exemplary embodiment, eight views A ₁ to are therefore for spatial representation A ₈ , preferably perspective views. For reasons of clarity, the "black matrix", which is often incorporated into color LC displays for technical reasons, has been omitted here.

The filter array ( 2 ) is preferably designed as a static, ie temporally unchangeable filter array, here e.g. B. as a thin film, and thus formed as a simple, passive element. Screen printing, in which filter arrays of larger dimensions (several square meters) can also be produced, is particularly suitable as a production method for such films. Such dimensions are advantageous if, for example, a large format screen, e.g. B. a plasma screen with 50-inch screen diagonal to be refined to the 3D screen. Other printing processes, such as B. thermal sublimation printing, or exposure processes can also be used.

The filter array ( 2 ) sets the directions of propagation of the light emitted by the color LC display ( 1 ) in the manner described above, which permits realistic, three-dimensional perception.

The three-dimensional perception then results from the large number of viewing positions resulting from the planar arrangement of the image elements αu and the filter β _pq : both eyes see image elements α _ij or partial information of predominantly different views A _k (k = 1. n), the proportion of the partial information perceivable for each eye being the cause of the three-dimensional perception.

The light rays visible to the eye have a small, however existing cross-sectional area.

The viewing space is to be understood here as the area in which the viewer can stay or move, whereby they perceive the scene or the object spatially in the viewing direction towards the grid. Depending on the structure of the mask image and the number n of views A _k (k = 1... N) used, viewing angles on the grid of more than 45 ° are possible, ie the viewing area can have an opening angle of more than 45 ° starting from the vertical perpendicular of the grid 45 °.

The invention offers many advantages over the prior art. In particular, a selected viewer ( 5 ) can see a correct spatial image from essentially the same perspective without any aids such as glasses. Pseudoscopic phenomena can be avoided for this special viewer even when using very large image display devices and the associated large-sized optical assemblies, which are necessary for spatial perception, while a large number of other viewers can also see the same scene spatially without aids.

Such a device can serve in particular one with an important task entrusted person, such as a surgeon, constantly the same perspective and on top of that always present correct spatial perspective, while an auditorium also a should see a spatial image, but this can change slightly in terms of perspective. This could e.g. B. be a group of medical students attending an operation.  

LIST OF REFERENCE NUMBERS

1

Flat screen with picture elements a _ij

2

Filter array with filters β _pq

3

drive circuit

4

viewing space

5

observer

6

shifter

7

Eye position tracking device

8th

Computing device for the tracking device (

7

)
α _ij

Picture elements in rows j and columns i
β _pq

Filters in rows q and columns p
λ _b

Transmission properties of the filters
i, p columns
j, q lines
z distance
R, G, B sub-pixels of the primary colors (red, green, blue) as picture elements α _ij

R ', G', B 'wavelength _filter β _pq

the basic colors (red, green, blue)
S opaque filter element

Claims (16)

1. Arrangement for the spatial representation of a scene / an object, comprising
an image display device, in particular a flat screen ( 1 ), for displaying an image array comprising a plurality of image elements α _ij , which, in a defined association, contains information from a plurality of views A _k (k = 1... n with n <3) of the scene / object represent, light being emitted from the picture elements in preferably different wavelength ranges,
At least one filter array ( 2 ), which is upstream or downstream of the image display device at a predetermined distance z with respect to the viewing direction of a viewer ( 5 ), the filter array ( 2 ) comprising a plurality of filter elements which are transparent in certain wavelength ranges, so that for the light propagation directions emitted by the picture elements α _ij are specified,
a displacement device ( 6 ) which is able to move the filter array ( 2 ) - preferably in the horizontal direction -
a tracking device ( 7 ) for detecting the eye position of a selected viewer ( 5 ),
a computing device ( 8 ) which actuates the displacement device ( 6 ) as a function of the eye position of the selected viewer ( 5 ) determined by the tracking device,
wherein at each position of the filter array ( 2 ) the directions of propagation within a viewing space ( 4 ) intersect at a plurality of intersection points, each of which corresponds to a viewing position, so that from each viewing position the viewer (s) predominantly has partial information from a first one eye Optically perceive the selection and with the other eye mainly partial information from a second selection from the views A _k (k = 1... N), the selected viewer ( 5 ) in particular almost exclusively because of the position of the filter array that is permanently matched to his eye position middle of the views A _k (k = 1... n) and, due to the use of n <3 views, the inertia of the displacement device is compensated in such a way that the selected viewer ( 5 ) does not enter the depth at any time perceives wrong picture. 2. Arrangement according to claim 1, characterized in that a flat screen ( 1 ), preferably a color LC display, with separately controllable subpixels is provided as the image reproduction device, each subpixel corresponding to a picture element α _ij in columns i and rows j, and that the flat screen ( 1 ) is coupled to a control circuit that generates partial information of the views A _k (k = 1 to n) on the subpixels according to the function
in which

• i corresponds to the index of a picture element α _ij in one line of the grid,
• j corresponds to the index of a picture element α _ij in a column of the grid,
• - k the consecutive number of the view A _k (k = 1... n) from which the partial information originating from a certain picture element α _ij originates,
• - n the total number of views A _{k used} (k = 1... n),
• - c _{ij of} a selectable coefficient matrix for combining or mixing the various partial information on the grid from the views A _k (k = 1... n)
• - IntegerPart of a function to generate the largest integer that does not exceed the argument in square brackets.

3. Arrangement according to claim 1 or 2, characterized in that the filter array contains filters β _pq , which are arranged according to the function in columns p and rows q
With

• p the index of a filter β _pq in a row of the array,
• q the index of a filter β _pq in a column of the array,
• b an integer which defines one of the intended transmission properties λ _b for a filter β _pq at position p, q and can have values between 1 and b _max ,
• n _{m is} an integer value greater than zero, which preferably corresponds to the total number b _{max of} the transmission properties λ _b , the total number b _max again preferably corresponding to the total number n of the views A _k shown in the combination image,
• - d _{pq of} a selectable coefficient matrix for varying the structure generated on the filter array and
• - IntegerPart of a function to generate the largest integer that does not exceed the argument in square brackets.

4. Arrangement according to claim 3, characterized in that the transmission properties λ _b for the filter β _pq preferably correspond to transparency wavelength or transparency wavelength _ranges . 5. Arrangement according to claim 4, characterized in that the transmission properties λ _b transparency wavelength ranges are given, each corresponding to the transparency in one of the three color ranges red, green or blue or the transparency in the entire visible spectrum or the lack of transparency in the entire visible spectrum. 6. Arrangement according to claim 4, characterized in that the transmission properties λ _b transparency wavelength ranges are given, each corresponding to the transparency in one of the three color areas cyan, magenta or yellow or the transparency in the entire visible spectrum or the non-transparency in the entire visible spectrum. 7. Arrangement according to one of the preceding claims, characterized in that a pivotable camera with a computing device ( 8 ) is used as the tracking device ( 7 ). 8. Arrangement according to one of the preceding claims, characterized in that the displacement device ( 6 ) comprises at least one rail and at least one linear motor, which is preferably designed as a stepper motor. 9. Arrangement according to one of claims 1-7, characterized in that the Moving device comprises at least one piezo actuator.   10. Arrangement according to one of the preceding claims, characterized in that in the viewing direction of a viewer ( 5 ) the filter array ( 2 ) between the flat screen ( 1 ) and a subordinate planar lighting source is arranged, which emits white light. 11. Arrangement according to one of the preceding claims, characterized in that in the viewing direction of a viewer, the filter array ( 2 ) is arranged in front of the flat screen ( 1 ), which is followed by a flat, white light-emitting illumination source. 12. Arrangement according to one of claims 1 to 11, characterized in that the individual filters of the filter array each have a polygonal, preferably rectangular Outline. 13. Arrangement according to one of the preceding claims, characterized in that the distance z between the flat screen ( 1 ) and the filter array ( 2 ) is 1 to 10 mm. 14. Arrangement according to claim 11, characterized in that the assembly of the flat screen ( 1 ) and the filter array ( 2 ) in the viewing direction of a viewer ( 5 ) is arranged upstream of a lens, preferably a Fresnel lens, whereby depending on the design of the lens a real and / or virtual representation of the spatial representation of the scene / object for the viewer. 15. Arrangement according to one of the preceding claims, characterized in that the filter array ( 2 ) is designed as a lenticular screen, preferably as a lenticular screen. 16. Arrangement according to one of the preceding claims, characterized in that the filter array ( 2 ) is formed by an LC display, the associated computing device ( 8 ) provided for a displacement device ( 6 ) instead of a displacement device ( 6 ), said LC display Display controls depending on the signal of the tracking device ( 7 ).