ASSOCIABLE WITH SAID SCREEN

DESCRIPTION

An object of the present invention is a screen (or whatever "monitor" or "display") that can be used for projecting images and/or videos and/or multimedia contents, as well as applicable to different civil or industrial fields such as, for example, the projection of real- time high-quality images during surgery or gaming activity; such screen will be of the auto-stereoscopic type, i.e. it will be capable of recreating a static or moving image clearly having a nature or ability to be perceived in a "three-dimensional" way by a user/observer placed at a given (observation) distance from the screen itself.

At the same time, an object of the present invention is a so-called "optical barrier" associable with the above-mentioned screen, and such optical barrier will be adapted to generate auto-stereoscopic optical phenomena being the basis of generating, displaying, and perceiving a three-dimensional image by the above-mentioned user/observer .

As is known, the neurophysiological ability of "stereoscopic" or three-dimensional vision in human beings results from specific physiological characteristics of the optical system and of how nervous system reads and processes visual signals perceived by each of the two eyes human beings normally have: on the other hand, it is known that image "artificial" generation and display devices which have been historically created in human history (technological but also artistic) were mainly limited to depicting, in a static or "dynamic" form (i.e. in the form of a flow of sequential images thereby reproducing scenes dynamically evolving over time) because of the difficulty of replicating the generation and perceiving of a sufficiently accurate and realistic three-dimensional image by a human being.

However, it is also known that there are different technologies adapted to artificially recreate an image being perceptible as "three-dimensional" by human beings: for example, the so-called stereoscopic glasses, to be used in combination with particular images (static or "dynamic") referred to as "anaglyph" and actually consisting of two overlapping images taken with suitably different angle from each other, are known: the structure of such stereoscopic glasses, being functionally coupled with the pair of partial images forming an anaglyph and suitably interfaced with human being eyes, makes it possible for each eye to see only the image relative to a camera angle, then the two different images separately reach the brain, which neurologically processes them in a three-dimensional object (NB: in other words, stereoscopic glasses make the three- dimensional vision possible by suitably using chromatically different lenses worn by a user/observer looking at an image being specifically decomposed according to the "anaglyph" mode reported above).

Stereoscopic glasses have remarkable operation and application limits, often resulting in inconvenience for users and that can generate an image quality which can be affected by chromatic aberrations and/or shape distortions or other vision/perceiving faults (blurring, etc.), obviously apart from the fact of necessarily needing to be "optically pointed" towards anaglyph images.

In order to overcome at least partially the drawbacks of this three- dimensional display technology, so-called auto-stereoscopic screens, which are functionally capable of decomposing an image projected on the screens themselves by a suitable structural association with a so-called "parallax barrier" have been developed: this device makes an optical decomposition of the image using the optical principle of parallax, and such decomposition is generated by a series of parallel slots placed side by side or by a suitable lenticular structure without needing to use secondary optical devices, as the support is provided with a system that provides to direct to each eye the image intended for it.

Therefore, in modern display technologies based on LED (light emitting diode) matrixes, parallax barriers are associated such that the image generated by selective lighting of pixels forming such LED matrixes is decomposed and suitably perceived by the user/observer: such association can be considered as a layering or overlapping of substantially planar members ordered along an ideal axis (which can be defined as an ideal "vision axis") originating from a first layer consisting of the LED matrix to pass through the optical barrier towards the point where the eyes of the user/observer are: along this axis, the LED matrix and the optical barrier are usually in a relation of mutual adjacency (or otherwise of close vicinity/proximity), while the position of the user/observer can be at a remarkably larger distance, as a function of the various fields of use of the screen comprising such matrix and such optical barrier mutually associated.

On the other hand, auto-stereoscopic screens are also characterized by drawbacks due to the deterioration of the quality of the image generated and perceived in a three-dimensional way, specifically because of possible aberrations and/or distortions of the colours and/or edges of the image: generally, such adverse effects occur upon moving aside not only the ideal distance, along the above- mentioned vision axis where the user/observer is with respect to the screen, but also moving away transversally and/or angularly with respect to the vision axis itself: these faults are essentially due to both the geometry of the constitutive members of the optical barrier and reflective, refractive and diffractive phenomena to which the light, generated by various pixels of the LED matrix when passing through the optical barrier itself, is subjected.

The above-mentioned image generation faults are particularly serious in case of using auto-stereoscopic screens in fields with high efficiency and quality requirements, such as for example real-time imaging required during "augmented reality" surgery or otherwise if a team of skilled operators requires a display of the operating area, usually subcutaneous and/or involving complex and crucial organs or apparatuses, in order to guide robotic surgical instruments or otherwise to direct instruments inside a human body without having a direct vision thereof: in such fields, incorrect perceiving of a three-dimensional structure of an object (which can be an organ of the patient !) could generate even very serious surgery errors or lengthen intervention time in an extremely inconvenient manner, as well as generate an additional neuromuscular fatigue load for operators performing surgery on the patient.

In view of the State of the Art above, an object of the present invention is to make an optical barrier, and accordingly an auto- stereoscopic screen provided with such optical barrier, which are capable of overcoming the above-mentioned drawbacks.

Specifically, the present invention is aimed to create an auto- stereoscopic screen which allows to define/generate a high-quality three-dimensional image (intended as definition of its lines, colour consistency of surfaces thereof, etc.) and especially being perceptible by a user/observer with high clarity, low neuro-optical fatigue with maximum positioning freedom (both in terms of linear distance and angular or positional displacement with respect to the ideal vision axis) with respect to the screen itself.

Even more generally, the present invention is intended to provide an auto-stereoscopic screen which can also be made in a very high dimension variety, also achieving remarkable values (for example, screens with a diagonal equal to or even greater than 50 inches, according to the unit of measurement currently used in such technical field) and without incurring into deterioration problems of the already low general quality of a three-dimensional image generated by auto-stereoscopic screens made with known technologies (NB: a further disadvantage of the above-mentioned Known Art is that applying known production technologies to parallax barriers generates undesirable curvatures of the diffraction microcracks or lenticules forming the barriers themselves, and the wider such geometric distortions are, the more superficially extended the barrier is: this generates further distortions and/or aberrations of the image upon increasing of the distance from the centre of the screen, and such further distortions and/or aberrations are added to those above-mentioned).

At the same time, an object of the invention is to provide an optical barrier which can be made according to scalable criteria and being repeatable on a wide range of surface extensions, and which can further be made with extremely reliable technologies, and which ensure very high repeatability results maintaining a constant production quality.

Also, an object of the invention is to create an innovative and original method for assembling and/or producing an auto-stereoscopic screen, combining the intrinsic functional advantages of the optical barrier described and claimed below and further advantages in terms of construction facility, structural simplification and possibility to achieve a high structural cohesion screen with a very high degree of insulation of its internal components from adverse environmental factors (the latter characteristic can be crucial if the auto- stereoscopic screens should be employed in surgical fields, since such fields of use require very high level hygienic-sanitary certifications) .

These and other objects are made by an auto-stereoscopic screen and an optical barrier associable with such auto-stereoscopic screen in accordance with the present invention, having the characteristics illustrated in the attached claims and illustrated below according to an exemplary (but not limiting) embodiment thereof, as well as in the enclosed drawings, wherein:

— Fig. 1 shows a partially exploded schematic perspective view of an auto-stereoscopic screen according to the invention and aligned according to a vision axis with a user/observer;

— Figs. 2 to 4 show schematic top views of a structural component of the screen of figure 2 on which, at different relative tilt angles, geometric lines corresponding to the projections of geometric directrixes of another structural component of the screen of figure 1 are projected; and

— Fig. 5 shows a schematic view of an alternative embodiment of a structural component of the auto-stereoscopic screen according to the invention.

With reference to figures, the auto-stereoscopic screen according to the invention is generally referred to as number (1) and firstly comprises a generation matrix (2) comprising a plurality of pixels (3) mutually adjacent arranged in the generation matrix (2) itself according to a plurality of mutually transversal rows and columns, and typically perpendicular. Pixels (3) comprise in turn some sub-pixels, which are represented by way of example (but not limiting) in a number equal to 3 and referred to as the alphanumeric references (3a; 3b; 3c).

The screen (1) further comprises an optical barrier (4) optically associated to the generation matrix (2) and adapted to receive as well as to decompose an image generated by the generation matrix (2): from a structural and geometric point of view, such optical barrier (4), which functionally makes possible the auto-stereoscopic effect such that the user/observer (0) of the screen (1) can perceive a three-dimensional image, comprises a plurality of plano-convex lenticular members (6a) arranged (or, in other words, developing) along parallel directrixes (6b).

For the purpose of the present invention, it should be noted that the above-mentioned directrixes (6b) are geometrically associated to the rows and/or columns of pixels (3) of the generation matrix (2) according to a predetermined slant angle (a), i.e. they are arranged such that their geometric projection on the lying plane of the generation matrix (2) is not parallel to the rows or columns of the matrix itself, but it is tilted with respect to the latter according to a predetermined angle (which is in fact the so-called slant angle well known in the technical field of pertinence of the invention). Advantageously, the invention involves that the directrixes (6b) of the plano-convex lenticular members (6a) are geometrically overlapped to the sub-pixels according to a so-called interlacing relation such that the (above-mentioned) geometric projections of the directrixes (6b) on the generation matrix (2) are substantially overlapped to following diagonals underlain at least between "corresponding points" of:

- a first sub-pixel pertaining to a pixel (3) located in a "first" row and a "first" column in the generation matrix (2); and

- a second sub-pixel located in a second row corresponding to or below or above the first row of the first sub-pixel and located in a second column corresponding or next or previous to the first column (in the generation matrix (2)); and/or

- a third sub-pixel located in a second row corresponding to or below or above the second row of the second sub-pixel and located in a second column corresponding or previous or next to the second column (in the generation matrix (2)).

The geometric/structural nature of the above-mentioned interlacing relation can be construed, in the present invention, according to the attached figures (and specifically according to what can be seen in figures 2 to 4), in which there is displayed the fact that the projections of the directrixes (6b) on the (plane of the) generation matrix (2) are arranged such that they pass, intersecting the edges or contours of the sub-pixels, still through equal points of each edge/contour (and meaning as "equal" or "corresponding" points, points of such edges of the sub-pixels which can be described by the same features, such as, for example, corresponding vertexes of edges of different sub-pixels or geometric centres of the geometric figures defined by the edges of the sub-pixels or any other typology of topological localization, such as for example middle points of long or short sides of the edges of the sub-pixels, etc.). It should be noted that "first", "second", "third" (and so on, in discrete ordinal progression) values attributable to the rows and columns of the generation matrix (2) should be intended, in the language of the present invention, as values of rows or columns defined by sub-pixels (3a; 3b; 3c) mutually aligned according to the two usual "cartesian" reference directions which can be ideally defined in the matrix (2) itself: obviously, since a pixel (3) usually consists of a series of sub-pixels mutually placed side by side, a series of rows and columns composed of pixels (3) could also be defined in the generation matrix (2) (and in such case, the number of columns of pixels (3) will be a fraction of the number of columns of sub-pixels, and obviously such fraction will be proportional to the number of sub-pixels forming the pixels themselves).

Therefore, the interlacing relation at the basis of the present invention can be defined as a series of geometric intersections between lines: in this regard, it should be noted that such intersections were illustrated in a "rough" manner in figure 1, while more accurate examples of the interlacing relation are shown in detail in figures 2 to 4.

Also, it should be noted that, in order to clarify the language of the present invention, the projections of the directrixes (6b) on the generation matrix (2) are represented in enclosed figures by lines "lying" on the generation matrix (2): in enclosed figures, such lines have the same numeric reference (6b) assigned to the real directrixes (6b), which are instead "physically located" in the transparent body forming the optical barrier (4) as clearly seen in figure 1.

In order to further explain the expressions used in the present invention, it should be noted that the above-illustrated interlacing relation refers to a "first row" and a "first column" of the generation matrix (2) meaning as "first" row or column not necessarily a row or column referred to as number "1" in a reference system of cartesian axes ideally lying on the generation matrix (2), but a general nth row or column in such matrix (2): therefore, in the expressions used for the invention, it can also be understood that the second row or column and the third row or column are previous or next (and thus next or previous by at least one ordinal value) to the general nth row or column of sub-pixels which can be defined in the generation matrix (2).

By way of example of the above illustrated, it can be considered that in figure 2, which illustrates (at least partially or even totally, as a function of the overall dimensions to be defined for the auto-stereoscopic screen (1) in accordance with the invention) a generation matrix (2) wherein the different sub-pixels (3a; 3b; 3c) are arranged on respective rows and columns: in such figure, the projections of the directrixes (6b) overlap/intersect the different sub-pixels (3a; 3b; 3c) such that:

- a first intersection/projection/overlapping point touches the centre of the first sub-pixel (3a) of the pixel (3) positioned on the row considerable as "first" (i.e., on the row placed higher and more to the left in figure 2 itself) as well as on the column considerable as "first";

- a second intersection/projection/overlapping point touches the centre of the second sub-pixel (3b) positioned on a "second" row below the first row and on a "second" column next to the first column;

- a third intersection/projection/overlapping point touches the centre of the third sub-pixel (3c) of the pixel (3) positioned on the "third" row and on the "third" column respectively below and next to the "second" row and the "second" column; and

- further following the upward and rightward trend in the scheme of the matrix (2), a series of nth intersection and/or projection and/or overlapping points touching the centre of respective nth sub- pixels positioned on columns "reducing" with respect to the previous rows and columns (and so on until reaching sub-pixels located at the end/perimetral edges of the generation matrix (2)).

The subsistence, in the present invention, of the interlacing relation thereby characterized makes possible an ideal and optimized optical coupling between the plano-convex lenticular members (6a) and the sub-pixels (3a; 3b; 3c) pertaining to the generation matrix (2): such coupling is such that the light projected by each sub- pixel (3a; 3b; 3c) "optically aligned" under a given lenticular member (6a) is correctly collected and redirected perfectly and exactly knowing which the sub-pixels to be "driven" selectively are (that is, to be turned on and/or supplied in order to allow an overall composition of a part or "view" of the overall image) such that the next decomposition of the image generated by the matrix (2) is performed minimizing the chromatic or geometric aberrations and/or distortions.

Still with reference to enclosed figures (and specifically to figures 1 to 4), it can be observed that sub-pixels (3a; 3b; 3c) can advantageously be superficially shaped according to polygons (for example, rectangles) and respectively define a width (s) (measured along one of the rows of the generation matrix (2)) and a height (h) (measured along one of the columns of the generation matrix (2)): the height (h) can be in turn a (integer) multiple of the width (s) such that it can be defined according to the formula h = ns where "n" can be defined by the expression "elongation ratio" of the sub-pixel (3a; 3b: 3c) (for example, in an embodiment of the invention, such elongation ratio can be equal to 3).

Still at the level of constitutive characteristics of the sub-pixels (3a; 3b; 3c), it can be observed that the latter can be arranged in a mutual side by side relation within a given pixel (3) and can respectively emit a predetermined light (which in the case of modern colour screens can differ from each other, but which could also be chromatically corresponding in the case of monochromatic auto- stereoscopic screens are to be made).

Returning now to the geometric concept of "corresponding points" in accordance with the language of the invention, it can be noted that the corresponding points of a first sub-pixel (3a) and/or of a second sub-pixel (3b) and/or of a third sub-pixel (3c) can for example consist of (at least):

- a vertex of the above-mentioned polygons (such vertex can be positioned in turn in the same position for each of the polygons defining the edges of the sub-pixel, such as for example the "top left" or "bottom right" vertex of each sub-pixel touched/crossed by the projection of the directrix (6b) as can be seen in figures 2 to 4); or

- a geometric centre of the polygons forming the edges/perimeters of the sub-pixels.

In order to be able to determine the auto-stereoscopic effect, which is based in turn on perceiving from different angles (by the user/observer (0)) of different views or "parts of overall image", which are then assembled at a neuro-perceptive level of the user/observer (0) in a three-dimensional "overall" or whatever "resulting" image, the generation matrix (2) is adapted to generate a predetermined number "N" of views (which can be for example an odd number of views and/or which can be between 2 and 18, depending on different possible positioning parameters of the user/observer (0) with respect to the screen (1) and/or depending on the complexity or resolution characteristics to give to the overall/resulting image): such number "N" of views defines, in a cooperative manner, a three-dimensional image perceptible by a user/observer (0) positioned along a vision axis (A) with respect to the screen (1). According to an aspect of the invention, each of said "N" views can be associated to a view index "N _View (k, i) ", which thus results a fundamental parameter for "selectively driving" the sub-pixels (3a; 3b; 3c): such view index "N _View (k, i) " actually correlates, in the generation matrix (2), a determined sub-pixel (lying at a relative coordinate, with respect to the scheme of ideal/cartesian axes of the matrix (2) described above as well as illustrated in attached figures) to a particular/specific view "N _View " among all those to be generated in order to create the auto-stereoscopic effect: such correlation is synergically established considering also the optical/geometric characteristics of the screen (1), so that in accordance with the present invention it is calculated/determined in a mode such that the interlacing relation described above and claimed below is always ensured between that determined sub-pixel and the optical barrier (4).

Advantageously, in accordance with the invention, the view index "N _View (k, i) " is determined according to the formula

In the above-reported formula, there are different parameters, which can be in turn calculated/defined through suitable dedicated "sub- formulas", and specifically:

(k) is determined according to the formula

(i) is determined according to the formula

(X) is determined according to the formula

Moreover, in the formula for calculating the view index there are parameters directly correlated to the structure/geometry/topology of the auto-stereoscopic screen (1) and to its relative position with respect to a user/observer (0), such as for example:

- (x), which is a position of a sub-pixel (3a; 3b; 3c) along an ideal reference axis parallel to at least one row of the generation matrix (2) (such ideal reference axis can be considered as the reference horizontal cartesian axis which can be seen in figures 2 to 4);

- (y), which is a position of a sub-pixel (3a; 3b; 3c) along an ideal reference axis parallel to at least one column of the generation matrix (2) (such ideal reference axis can be considered as the reference vertical cartesian axis which can be seen in figures 2 to 4);

- (s), which is the width of a sub-pixel (3a; 3b; 3c);

- (n), which is the above-mentioned elongation ratio of a sub-pixel (3a; 3b; 3c);

- (L), which is a distance between two directrixes (6b) of corresponding plano-convex lenticular members (6a) adjacent and advantageously forming part of the optical barrier (4) (in industry technical jargon, such distance can also be defined by the expression "lens pitch");

- (a), which is the above-mentioned slant angle;

- (D), which is the so-called "viewing distance", measured along a vision axis (A) which can be seen in attached figures, between the above-mentioned user/observer (0) of the screen (1) and the optical barrier (4); and finally

- (f), which is the so-called "focal length" of a plano-convex lenticular member (6a) pertaining to the optical barrier (4) and optically associated to at least one of the sub-pixels (3a; 3b; 3c) (for the purpose of the present invention, such focal distance actually coincides with a distance, measured along the vision axis (A), between the optical barrier (4) and the generation matrix (2)). Still with reference to the formula for the view index, it should be observed that it contains some peculiar mathematical operators, and specifically the operator "mod", which is a mathematical operator having as a result a remainder of a so-called "Euclidean division" between two numbers (referred to in the so-called "subject" of the operator itself) as well as the operator "int", which is a mathematical operator having as a result a smaller integer numeric value (with respect to the number or size referred to in the so- called "subject" of the operator itself).

In the field of the present invention, now it should be noted that the above-mentioned formula for calculating the view index can be used in synergy with the interlacing relation (of a geometric/topological nature) in order to maximize the "qualitative results" of an auto-stereoscopic screen in terms of reduction or absence of geometric or chromatic aberrations and/or distortions of the resulting three-dimensional image, but depending on current requirements it can be used also in preliminary dimensioning (and thus in "driving" modes for turning on the pixels and/or sub-pixels of a generation matrix) of auto-stereoscopic screens wherein the optical barrier is not necessarily coupled, from an optical point of view, to the generation matrix itself according to the interlacing relation of the present invention (for example, it can be possible that the view index is determined by the above-mentioned formula but that the generation matrix "driven" according to such view index calculation is then coupled to an optical barrier in which the projections of the directrixes of the lenticular members have a relative tilt different from that illustrated in the present invention).

However, because of the particular synergic effect between the subsistence of the interlacing relation and the result of the formula for calculating the view index mentioned above (and claimed below), the structural/construction parameters of the auto-stereoscopic screen (1) can be defined in an innovative and original manner, in order to obtain optical effects on the resulting three-dimensional image (in the eyes of the user/observer (O)) of very high quality: as a result of this synergic effect, it should be noted that the invention is capable of determining a slant angle (a) according to the following formula

In the above-mentioned formula the following parameters can be read:

- (s), which is the above-mentioned width of a sub-pixel (3a; 3b; 3c);

- (K), which can be advantageously defined as a "horizontal pitch of repetition" of the interlacing relation needed between the projections of the directrixes (6b) and between sub-pixels (3a; 3b; 3c) mutually arranged in the interlacing relation according to the invention itself (such sub-pixels (3a; 3b; 3c) are then positioned in columns of the generation matrix (2) mutually spaced according to such horizontal pitch of repetition); and finally

- (P), which can be defined as a "vertical pitch of repetition" of the interlacing relation needed between the projections of the directrixes (6b) and between sub-pixels (3a; 3b; 3c) mutually arranged in the interlacing relation according to the invention (such sub-pixels (3a; 3b; 3c) are then positioned in rows of the generation matrix (2) mutually spaced according to such vertical pitch of repetition) .

From a merely numerical point of view, the horizontal and vertical pitches (of repetition) (K), (P) are defined by an integer number at least equal to 1 (and for example between 1 and 7), while for the purpose of clarification of their "physical" nature, reference can be made to figures 2 to 4, wherein some indicative values of such horizontal and vertical pitches of repetition (K) and (P) are reported.

With reference to enclosed figures, it should be noted that these pitches (P) and (K) denote (respectively) the number of columns or rows at which the interlacing relation in accordance with the invention repeats itself, or in other words they denote the number of rows or columns to "skip" in a count (along the "horizontal" or "vertical" scanning directions of the generation matrix (2)) to find the geometric correspondence between the projections of the directrixes (6b) and the above-mentioned "corresponding points" of the (edges of the) sub-pixels (3a; 3b; 3c) which determines the interlacing relation in accordance with the invention.

By a suitable utilization of the above-mentioned formulas and considering the above-mentioned optimization criteria of the quality of the three-dimensional image, it is advantageously possible to determine an embodiment of the invention such that the slant angle (a) - which can be seen in other words as a relative rotation angle set between the optical barrier (4) and the generation matrix (2) - is between 18° and 19° and is for example equal to 18.435°.

The above-mentioned slant angle (a) value advantageously occurs at a value of (K) equal to 1 and at a value of (P) equal to 1 (as well as at the usual elongation ratio of 3 - in height - to 1 - in width - typical of the sub-pixels available on the market nowadays) and is advantageously displayed in attached figure 2.

Moreover, in such figure 2 it can be noted that the interlacing relation at the basis of the invention is such that the projections of the directrixes (6b) pass at least through the following "corresponding points":

- a "top left" vertex of a first sub-pixel positioned in a first row and a first column of the generation matrix (2);

- a "bottom right" vertex of such first sub-pixel;

- a "top left" vertex of a second sub-pixel positioned in a second row immediately below the first row and in a second column immediately next to the first column of the generation matrix (2); a "bottom right" vertex of such second sub-pixel; - a "top left" vertex of a third sub-pixel positioned in a third row immediately below the second row and in a third column immediately next to the second column of the generation matrix (2); and

- a "bottom right" vertex of such third sub-pixel.

Moreover, in a further embodiment of the invention, it is possible that the slant angle (a) is between 9° and 10° and for example is equal to 9.4326° at a value of (K) equal to 1 and at a value of (P) equal to 2 (still at the usual elongation ratio of 3 - in height - to 1 - in width - typical of the sub-pixels available on the market nowadays).

In this further embodiment of the invention, schematically illustrated in figure 3, the subsistence of the interlacing relation between the projections of the directrixes (6b) and the corresponding points of the sub-pixels is such that a directrix (6b) touches/intersects/passes through at least:

- a "top left" vertex of a first sub-pixel (3a) positioned in a first row and in a first column of the generation matrix (2);

- a "bottom right" vertex of a first sub-pixel (3a) positioned in a second row (below the first row) and in the same column of the first sub-pixel (3a) listed above;

- a "top left" vertex of a second sub-pixel (3b) positioned in a third row below the second row and in a second column next to the above-mentioned first column; - a "bottom right" vertex of a second sub-pixel (3b) positioned in a fourth row (below the third row) and in the same second column of the previous second sub-pixel (3b);

- a "top left" vertex of a third sub-pixel (3c) positioned in a fifth row (below the previous fourth row) and in a third column next to the above-mentioned second column; and

- a "bottom right" vertex of a third sub-pixel (3c) positioned in a sixth row (below the fifth row) and in the same third column mentioned above.

In a further embodiment of the invention, illustrated in figure 4 (where the slant angle (a) is equal to 33.69° at the usual elongation ratio of 3 - in height - to 1 - in width - typical of the sub-pixels available on the market nowadays), assigning a value of (K) equal to 2 and a value of (P) equal to 1 the interlacing relation between the projections of the directrixes (6b) and the "corresponding points" of the edges of the sub-pixels (3a; 3b; 3c) forming the generation matrix (2) will be advantageously such that a directrix (6b) touches/intersects/passes through:

- a "top left" vertex of a first sub-pixel (3a) positioned in a first row and in a first column of the generation matrix (2);

- a "bottom right" vertex of a second sub-pixel (3a) positioned in the first row and in the next "second" column of the generation matrix (2);

- a "top left" vertex of a third sub-pixel (3c) positioned in a second" row below the second row and in a "third" column next to the second column; - a "bottom right" vertex of a first sub-pixel (3a) positioned in the second row and in a "fourth" column next to the third column;

- a "top left" vertex of a second sub-pixel (3b) positioned in a third row (below the previous second row) and in a fifth column (next to the fourth column); and

- a "bottom right" vertex of a third sub-pixel (3c) positioned in the third row and in a "sixth" column next to the fifth column.

The above-mentioned interlacing relation and/or peculiar mode for defining (by mathematical formula) the view index can be applied, in accordance with the present invention, both to generation matrixes (2) wherein sub-pixels (3a; 3b; 3c) have a traditionally rectangular shape and to other typologies and/or shapes of sub-pixel: for example, it is possible that at least one pixel (3) (but typically all pixels (3) of the generation matrix (2)) comprises at least one sub-pixel (3a or 3b or 3c) shaped according to at least one pair of juxtaposed parallelograms as illustrated in figure 5: moreover, in such figure it should be noted that the above-mentioned pair of juxtaposed parallelograms is placed in a mutual contact condition at least along one side of the parallelograms themselves and thus defines an "arrow" or "V" shape of the sub-pixel (3a or 3b or 3c). As a result of the particular shape of the sub-pixels illustrated in figure 5, all sub-pixels of the pixels (3) of the generation matrix (2) comprise a predetermined number of sub-pixels (3a; 3b; 3c) defining an "arrow" or "V" shape: such sub-pixels are positioned, in the single pixel (3), in a mutual approaching relation having the same alignment of a cusp of said "arrow" or said "V". The invention allows to achieve important advantages.

Firstly, it should be noted that the peculiar construction architecture of the invention which, as seen, involves a substantial inversion of orientation (along the vision axis) of the main components with respect to known-type screens, allows to achieve a remarkable improvement of quality of the three-dimensional image generated and then perceived by the user/observer: such improved quality can occur both upon varying of the distance of the user/observer and upon varying of the angle of incidence of the vision axis with respect to the "plane" of the screen and also upon varying of the overall dimensions of the screen itself.

Moreover, it should be noted that the improvement of quality of the three-dimensional image generated by the auto-stereoscopic screen according to the invention is such also for images of remarkable dimension or otherwise being created/projected on peripheral screen areas, avoiding distortions of shapes and/or aberrations of chromatism and even reducing (if not totally suppressing, at least in relation to the human eye ability in terms of minimum resolution distinguishable or perceptible) the so-called "cross-talk" phenomenon which, as is known, involves the vision/perceiving of edges and lines according to stair-step/broken lines rather than according to continuous geometric or otherwise "regular" trends.

Furthermore, it should be noted that peculiar definition modes of the optical barrier and of its both optical and structural interfacing with the LED matrix below allow a very efficient decomposition of the image generated (by the matrix itself) with an almost constant qualitative level on the entire screen area: this further increases the vision quality by the user/observer and makes the present auto-stereoscopic screen capable of projecting both static and dynamic images, moreover reaching very high update accruals of the image and thus actually allowing to project/display three-dimensional videos almost in real time (to the advantage of the applications requiring, for example, an accurate "live" representation of three-dimensional objects dynamically movable over time with respect to the observer itself).

Regarding the methodological part of the present invention, it should be noted that the structure described above and/or claimed below allows, apart from the above-mentioned structural compaction, a remarkable overall lightening of the screens which can be made according to the present invention (and such lightening is crucial by increasing of the dimension of the screens, which can have also diagonal dimensions around 80 inches or more nowadays), a cost reduction (due to usable materials, adopted forming methods and assembling modes of the different components) and a very high degree of insulation of the internal components with respect to the environment in which the screens can be operated (and this results, still by way of example, in the possibility to achieve very high level hygienic (sanitary certifications).