AUTOSTEREOSCOPIC SCREEN, SO-CALLED PHOTOGRAPHIC QUALITY

Abstract

AUTOSTEREOSCOPIC SCREEN, REFERRED TO AS PHOTOGRAPHIC QUALITY The invention relates to an autostereoscopic display screen (10), referred to as a photographic quality display, of an autostereoscopic image with MxN viewpoints, N being an odd integer and M being an integer chosen from 1, 2, 3, and 6, said screen comprising a pixel panel (20) arranged in rows and columns, each pixel (26) being composed of a plurality of P active sub-pixels of different colors, and an ordered array of cylindrical lenses (30) inclined with respect to the columns of the screen at an angle equal to arctan(1 / 12), said array having a pitch of N / 2 pixels wide in order to differentiate the even and odd lenses (30) of said array by distinct distributions of the active (26) and black (24) areas of the pixels arranged under the lenses, and, for decoding each of the MxN viewpoints distributed within N vertical blocks, 6 pixels high, derived from M viewpoints,every 12 lines if the panel subpixels are horizontal, and distributed within PxN vertical blocks of 6 / P subpixels in height, from M / P viewpoints every 12 / P lines if the panel subpixels are vertical. Figure for the abbreviation: Figure 2,

Description

Title of the invention: AUTOSTEREOSCOPIC SCREEN, SAID TO BE PHOTOGRAPHIC QUALITY 1. Technical field of the invention

[0001] The invention relates to an autostereoscopic screen and a method of displaying an autostereoscopic image on such an autostereoscopic screen. 2. Technological background

[0002] Autostereoscopy is a technique for displaying relief images without requiring the observer to wear special glasses. This technique is known per se, in particular from patent documents WO2006 / 024764, WO2014 / 041504, WO2013 / 140363, WO2014 / 016768, WO2019 / 063897, WO2019 / 207235, WO2022 / 175053 in the name of the applicant.

[0003] Generally speaking, an auto stereoscopic image is made up of a plurality of bands of nested elementary images, each corresponding to views of the same object or the same scene from different points of view. A selector device, typically made up of a network of cylindrical lenticules or a parallax barrier, is arranged in front of the display screen so as to allow the projection of a pair of elementary images corresponding to two different points of view of the scene, respectively towards the two eyes of the observer, which creates an impression of relief in the brain of the observer.

[0004] The applicant has already proposed an autostereoscopic screen for displaying an image at N viewpoints comprising a matrix of pixels arranged in rows and columns, each pixel being composed of a plurality of sub-pixels of different colors. The screen is further surmounted by a network of identical cylindrical lenticules each having a focal length configured to be able to reflect to infinity the light rays coming from the screen. The pitch of the network of cylindrical lenticules is precisely calculated so that the observer sees, at a predetermined distance from the screen, the images succeed one another thanks to a magnifying effect of the network of lenticules.

[0005] This magnifying effect results from the fact that a lens placed at the correct distance (its focal length) magnifies the sub-pixel(s) which are in alignment with its optical axis and the pupil of the observer's eye. If the lens magnifies N times, the sub-pixel(s) seen through the lens is / are perceived N times larger than it is / are in reality and masks from the eye which receives the light through this lens the Nl other sub-pixels which are not in the alignment described above.

[0006] The very principle of the lenticular network does not allow the viewer to be offered a continuous parallax of the scene represented over an angle of 180°. The angle of observation is therefore subdivided into bands called "lobes", within which the parallax is, however, discretized. In each lobe, the N points of view can be perceived successively. When one leaves a lobe, one finds the succession of N identical points of view in the following lobe, and so on across the entire width of the screen's observation area. The angular width of the lobes is defined both by the pitch of the lenticular network and by its focal length.

[0007] Autostereoscopic screens require subdividing the lobes into discrete, perfectly discernible sub-parts, the individual width of which in the lobe cannot exceed the average inter-pupillary distance, which is 6.5 cm.

[0008] Also, one of the keys to the performance of the autostereoscopic screens currently proposed by the applicant is the control of the separating power of the optical components which are developed specifically for each pixel panel.

[0009] For printed autostereoscopic photographs (and not autostereoscopic screens), it is possible to register a maximum number of printing points or dots, contiguous but well differentiated from each other, on supports offering good dimensional stability. Thanks to the printing techniques currently implemented, the applicant manages to print 100 points per millimeter and sometimes even more, which amounts to visualizing points of 10 microns or less.

[0010] When the viewer moves in front of such a photo, he sees the different elements of the scene moving in relation to each other, continuously and perfectly in accordance with the original physical volume, as it would be perceived if seen from the same distance. The two eyes see two different points of view of the scene simultaneously, which allows the brain to reconstruct a perfectly natural sensation of relief.

[0011] It is a priori impossible to recover the quality of autostereoscopic photographs with screens insofar as the latter are made up of a matrix of pixels with varying resolution, composed of rows and columns orthogonally aligned and separated from each other by a black mask composed of opaque rows and columns of varying thickness depending on the panels and which form a grid around the active areas of the pixels. The surface ratio between the black mask and the colored active sub-pixels impacts the brightness and contrast of the screen, but also its separating power in autostereoscopy.

[0012] Throughout the text, the auto stereoscopic screen is considered to have photographic quality when it allows the elimination of discrete jumps when moving from one stereoscopic pair to another, in the parts of the image where the disparity is greatest.

[0013] In other words, by "photographic quality" we mean a continuity of relief within the lobe, which seems a priori not accessible to an auto stereoscopic screen except equip it with a tracking system.

[0014] In addition to the problem of resolving power, there is also the issue of moiré patterns. Indeed, the lenticular network magnifies both the active areas of the pixel panel and the black mask surrounding them. As a result, certain sub-parts of the observation lobe may appear darker or brighter than others. These variations in brightness are generally perceptible at the transition between two neighboring viewpoints. The light intensity moiré effect varies according to the particular structure of the panel and the orientation of the screen.

[0015] The inventors sought to overcome the technical prejudice of the person skilled in the art according to which it would be a priori impossible to recover the quality of a photographic autostereoscopic image (characterized by a continuity of relief within the lobe) with a screen equipped with a pixel panel. 3. Objectives of the invention

[0016] The invention aims to provide a so-called photographic quality autostereoscopic screen.

[0017] In particular, the invention aims to provide an autostereoscopic screen which makes it possible to obtain a parallax continuum, without perceptible discretization of the viewpoints.

[0018] Thus, the invention aims to provide, in at least one embodiment, a screen which offers a continuous parallax sensation in each lobe.

[0019] The invention also aims to provide an autostereoscopic screen without noticeable moiré or ghosting.

[0020] The invention also aims to provide a so-called photographic quality screen, regardless of the structure of the sub-pixels of the pixel panel which equips the screen (vertical or horizontal sub-pixels).

[0021] The invention finally aims to provide a method for displaying an autostereoscopic image on a screen according to the invention. 4. Statement of the invention

[0022] To do this, the invention relates to an autostereoscopic display screen, said to be of photographic quality, of an autostereoscopic image with MxN points of view ordered from 1 to MxN, N being an odd integer and M being an integer chosen from 1, 2, 3 and 6.

[0023] The screen according to the invention comprises a panel of pixels arranged in rows and columns, each pixel being composed of a plurality of P active sub-pixels of different colors and opaque inactive zones forming a portion of black interlines and / or intercolumns of said panel.

[0024] The screen according to the invention is characterized in that it further comprises an ordered array of cylindrical lenses mounted on said pixel panel, each lens being inclined relative to the screen columns at an angle equal to arctan(l / 12).

[0025] Furthermore, said network has a pitch of N / 2 pixels wide and makes it possible to differentiate the even and odd lenses of said ordered network by distinct distributions of the active pixel zones and the black zones of the panel arranged under each even and odd lens.

[0026] The network also makes it possible to decode each of the MxN viewpoints distributed within N vertical blocks of 6 pixels in height from M viewpoints (6 / M pixels per viewpoint), with a periodic iteration of each block in the axis of a lens every 12 lines if the sub-pixels of the panel are horizontal, and distributed within PxN vertical blocks of 6 / P sub-pixels in height, from M / P viewpoints (6 / M sub-pixels per viewpoint), with a periodic iteration of each block in the axis of a lens, every 12 / P lines if the sub-pixels of the panel are vertical.

[0027] In other words, the screen has the particularity of allowing each of the MxN viewpoints to be coded on the panel extending under the lenses with 6 / M different vertical pixels within N vertical blocks of 6 pixels in height which are repeated, under each lens, every 12 lines if the sub-pixels of the panel are horizontal, and with 6 / M different vertical sub-pixels within PxN vertical blocks of 6 / P sub-pixels in height which are repeated, under each lens, every 12 / P lines if the sub-pixels of the panel are vertical.

[0028] The network is configured to decode by magnifying glass effect each of the MxN coded viewpoints according to the principle set out above.

[0029] The screen according to the invention also has the particularity of being equipped with a network of cylindrical lenticules each inclined relative to the columns of the screen by an angle equal to arctan (1 / 12), that is to say an angle of the order of 4.76°. The function "arctan" designates the arctangent. The arctangent of a real number is the value of an oriented angle whose tangent is equal to this number.

[0030] The screen also has the particularity of having a lenticular network configured to allow the distribution under each lens of the viewpoints over 12 consecutive lines if the sub-pixels are horizontal and 12 / P consecutive lines if the sub-pixels are vertical.

[0031] Throughout the following and to simplify the terminology, the term "horizontal mode" or "horizontal version" of the screen refers to the screen formed from horizontal sub-pixels and the term "vertical mode" or "vertical version" of the screen refers to the screen formed from vertical sub-pixels, it being understood that this does not presage the portrait or landscape orientation of the screen. The horizontal mode and the vertical mode only characterize the orientation of the sub-pixels and not the portrait or landscape orientation of the screen.

[0032] Throughout the following and, unless otherwise stated, to simplify the terminology, the term pixel refers to the active areas of the pixel which allow content to be addressed and by black space, the opaque inactive areas of pixels which form the interlines and / or intercolumns of the screen.

[0033] In horizontal mode, with a slope of 4.76°, each lens of the lenticular network crosses a set of elementary vertical blocks of 6 pixels, then a black space, before meeting a new set of elementary blocks of 6 pixels. It should be noted that all pixel matrices have a black opaque zone between each line of active pixels, but this line, a constituent part of the pixel, allows, depending on its percentage in the pixel, to benefit from good separating power.

[0034] N being an odd integer, the ordered network of the screen has a non-integer pitch N / 2, which makes it possible to differentiate the even lenses from the odd lenses by at least partially contrasting the distribution of the bright areas (active areas of the pixels) with the distribution of the black areas (opaque areas of the pixels) of the screen. In other words, if the lenses of the ordered network are numbered, from the left of the screen to the right of the screen, the pixel panel under the even lenses has the same distribution of the active pixel areas and the black areas of the screen, and the pixel panel under the odd lenses of the screen has the same distribution of the active pixel areas and the black areas of the screen, the distribution of the even lenses being distinct from the distribution of the odd lenses of the screen.If we were to superimpose the portion of the pixel panel under an even lens with the portion of the pixel panel under an odd lens, we would therefore see that at least a portion of the dark areas of the panel under the even lens is opposite at least a portion of the active areas of the panel under the odd lens and at least a portion of the active areas of the panel under the even lens is opposite at least a portion of the dark areas of the panel under the odd lens.

[0035] For a 1xM viewpoint array (N is equal to 11), the lenses have a pitch of 5.5 pixels wide (in the unsimplified definition of the pixel). For a 7xM viewpoint array (N is equal to 7), the lenses have a pitch of 3.5 pixels wide.

[0036] The screen network according to the invention makes it possible to decode by magnifying glass effect each of the MxN viewpoints distributed within N vertical blocks of 6 pixels in height from M viewpoints (6 / M pixels per viewpoint), with a periodic iteration of each block in the axis of a lens every 12 lines if the sub-pixels of the panel are horizontal, and distributed within PxN vertical blocks of 6 / P sub-pixels in height, from M / P viewpoints (6 / M sub-pixels per viewpoint), with a periodic iteration of each block in the axis of a lens, every 12 / P lines if the sub-pixels of the panel are vertical.

[0037] In the horizontal version, the advantage of having vertical blocks (also referred to by the terminology of elementary blocks) of 6 pixels is to offer the alternative to assign each of these pixels the same point of view or different points of view.

[0038] When the scene represented is shallow, it poses little problem to display the same point of view, called the main point of view, on the 6 consecutive pixels of an elementary block.

[0039] On the other hand, the perceived quality deteriorates when the disparity is greater. The perfect focusing of the networks then results in a clear perception of the discretization of the contents in great depth or in strong burst and this harms the photographic experience (continuous parallax) aimed for. When the horizontal translation of a contrasting detail in the image exceeds 2 or 3 lenses during a change of point of view, the discretization is excessive and it then becomes essential to increase the number of points of view. Although this is done to the detriment of the initial resolution of the main points of view displayed, the improvement in the perceived quality is obvious.

[0040] If we assign each of the 6 pixels of each elementary block a slightly different point of view, the discretization disappears in scenes with high disparity.

[0041] The screen according to the invention thus makes it possible to display viewpoints, called intermediate viewpoints, between the N main viewpoints of the image to be displayed. These intermediate viewpoints are slightly offset between each main viewpoint to distribute them within each elementary block. The continuum between the successive main viewpoints creates a photographic depth of field effect in the distant backgrounds and the foregrounds in strong burst instead of an ostensible discretization.

[0042] In other words, the screen according to the invention is remarkable in that it allows the use of groups of 6 pixels high per main pixel (of which it is part) to code slightly different viewpoints to fill with a continuum the discrete succession of the main viewpoints to approach the quality of autostereoscopic photographic images.

[0043] Thanks to this new approach, which we refer to throughout the text as the "tiling effect", the objective of finding on a screen the continuity of the parallax of printed photographic images is achieved.

[0044] In particular, in a relief image, the disparity between viewpoints increases as one moves away from the screen plane (also called the collimation plane). The overlap between adjacent viewpoints, referred to in English as "crosstalk", is identical across the entire screen and affects the entire displayed image but is only apparent in the parts of the image where the disparity between viewpoints is significant.

[0045] The inventors have notably realized that photographic blur is potentially information associated with relief, but also the materialization of the depth of field. The inventors therefore sought, contrary to the prejudice of the technical field, to ensure that the depth or the burst in an image is materialized by a progressive fade and not by a discretization between points of view. The overall impression sought is that of a deep and continuous image, in which the fade between points of view becomes more pronounced as the objects move away from the plane of collimation, assimilated as the plane of focus.

[0046] Discretization between viewpoints occurs when identifiable details in the image are repeated several times. When the disparity in a scene is significant and the number of viewpoints insufficient to distribute it harmoniously between each, it is common for contrasting objects far from the collimation plane to be degraded by the perception of ghosts.

[0047] Conversely, the deep details of several viewpoints can overlap simultaneously, with an attenuation of the intensity and contrast of the perceived redundancy as one moves away from the central viewpoint, thus, a single object is visually re-formed even if its image remains a little blurred.

[0048] The screen according to the invention offers the possibility of adapting the number of viewpoints displayed (by modifying the value of M) according to the disparity in the image, in depth as well as in projection, with the ghosts of the details at the limits of the volume as arbiter. This new screen offers the viewer a continuous parallax within the lobe, without apparent discretization, but instead with a progressive blurring of the details as one reaches the limits of the volume.

[0049] This approach goes against the screens proposed until now where the search for a perfect discretization of the points of view was aimed at and considered as the way to follow to offer a good quality screen.

[0050] Advantageously and according to the invention, each pixel of said pixel panel is composed of a plurality of P horizontal sub-pixels of different colors juxtaposed on top of each other in the direction of the columns, so as to be able to code each of the MxN viewpoints with 6 / M different vertical pixels, which are repeated, under each lens, every 12 lines, M being equal to 1, 2, 3 or 6.

[0051] In other words, the screen according to the invention allows, in the horizontal version, to display as desired: - N main viewpoints, each coded with 6 different and contiguous vertical pixels (case where M is equal to 1), - 2xN viewpoints each coded with 3 different vertical pixels (case where M is equal to 2), - 3xN viewpoints each coded with 2 distinct vertical pixels (case where M is equal to 3), - 6xN viewpoints each coded with a single pixel (case where M is equal to 6).

[0052] Advantageously and according to the invention, each pixel of said pixel panel is composed of a plurality of P vertical sub-pixels of different colors juxtaposed next to each other in the direction of the lines so as to be able to code each of the MxN viewpoints with 6 / M different vertical sub-pixels, which are repeated, under each lens, every 12 / P lines, M being equal to 3 or 6.

[0053] In other words, the screen according to the invention allows, in the vertical version, to display as desired: - 3xN coded viewpoints each with 2 distinct vertical sub-pixels (case where M is equal to 3), - 6xN viewpoints each coded with a single sub-pixel (case where M is equal to 6).

[0054] Furthermore, the screen according to the invention has narrower lobes than the screens previously proposed by the applicant. To this end, the optical components have relatively long focal lengths in order to narrow the lobe as much as possible without degrading the viewer's positioning comfort. This makes it possible to increase the resolution power and angular continuity. Narrower lobes multiply the transition zones but they also divide the observation space in front of the screen into narrower bands, where several viewers can position themselves simultaneously and very comfortably, at different distances.

[0055] The invention thus goes against a prejudice of the person skilled in the art who considers that a larger lobe width is inseparable from the quality of an autostereoscopic screen. Contrary to what one might think, the image quality perceived when the 6 pixels of an elementary block belong to the same high-resolution viewpoint is comparable to that obtained when one "tiles" in this elementary block intermediate viewpoints of lower resolution instead of spreading the main viewpoint. In other words, a display with very high theoretical resolving power does not prove to be better than a display having a partial overlap of intermediate viewpoints by a "tiling" effect. On the other hand, instead of the deep details being discretized when the resolving power is optimized, they spread out in the manner of a depth of field effect into a blur of circular appearance.The intermediate viewpoints spread both horizontally and vertically, which allows for this quasi-isotropic improvement of the resulting image.

[0056] Humans are accustomed to depth-of-field blurs and it is physiologically inevitable to perceive blur in the most distant planes of a relief image. The brain readily accepts that the blur results from the distance of these planes and this does not alter the feeling of overall sharpness that we otherwise perceive in the image.

[0057] According to a variant of the invention, the screen has the particularity of being equipped with a pixel panel, which when it is oriented in horizontal mode (respectively in vertical mode), has columns (respectively lines) separated from each other by black intercolumns (respectively black interlines) whose width (respectively height) is close to that of the active areas of the pixels of the panel. These black surfaces form a black mask whose surface is close to the active surface of the screen. This makes it possible to have excellent levels of contrast.

[0058] Furthermore, according to this variant and in the case of horizontal sub-pixels, the active and dark areas of the portions of the panel arranged under the even and odd lenses are opposite each other. In other words, the active areas under the even lenses correspond to the dark areas under the odd lenses and the dark areas under the odd lenses correspond to the dark areas under the odd lenses.

[0059] Advantageously and according to the invention, N is equal to 7 or 11.

[0060] The inventors have determined that finer lenticular networks, which allow only 7 or 11 main viewpoints to be displayed, allow the structure effect to be further marginalized. That being said, there is of course nothing to prevent a different number of viewpoints from being used.

[0061] Advantageously and according to the invention, P is equal to 3.

[0062] According to this variant, each pixel of the pixel matrix is ​​composed, when said screen is oriented in horizontal mode, of three horizontal sub-pixels of different colors juxtaposed on top of each other in the direction of the columns and said columns are separated from each other by a black inter-column space. Of course, when this screen is oriented in a vertical mode, each pixel of the matrix is ​​composed of three vertical sub-pixels of different colors juxtaposed next to each other in the direction of the lines and said lines are separated from each other by a black inter-line space.

[0063] However, nothing prevents the use of pixel panels in which each pixel is made up of four sub-pixels. In particular, there are panels which, in addition to the three usual colors, red, green and blue, add a white sub-pixel.

[0064] Advantageously and according to the invention, said pixel matrix is ​​an 8K pixel panel.

[0065] Such a pixel panel makes it possible to exploit the highest resolutions currently available. Such a panel is, for example, a 31.5” (80.01 cm) panel with a resolution of 7680x4320 pixels, which, when oriented in portrait mode, has a matrix of pixels, each formed of horizontal RGB sticks.

[0066] Indeed, the higher the screen resolution, the more it is possible to display a large number of viewpoints simultaneously. It is then possible either to use these additional viewpoints to widen the lobe by decreasing the focal length at a constant step, or to reduce the width occupied by each viewpoint in the lobe by devoting a narrower sub-part of the lobe to it, in order to approach a photographic effect.

[0067] The invention also relates to a method for the photographic quality autostereoscopic display of an autostereoscopic image with MxN viewpoints ordered from 1 to MxN, N being an odd integer and M being an integer chosen from 1, 2, 3 and 6, characterized in that it comprises: • the selection of an autostereoscopic display screen according to the invention, • the distribution under each lens of the MxN viewpoints distributed within N vertical blocks of 6 pixels in height from M viewpoints (6 / M pixels per viewpoint), with a periodic iteration of each block in the axis of a lens every 12 lines if the sub-pixels of the panel are horizontal, and distributed within PxN vertical blocks of 6 / P sub-pixels in height, from M / P viewpoints (6 / M sub-pixels per viewpoint), with a periodic iteration of each block in the axis of a lens, every 12 / P lines if the sub-pixels of the panel are vertical.

[0068] In other words, the autostereoscopic display method according to the invention allows the distribution under each lens of the MxN viewpoints of the image within N elementary blocks which are repeated every 12 consecutive lines if the pixels of the screen are formed of horizontal sub-pixels or within NxP elementary blocks which are repeated every 12 / P consecutive lines if the pixels of the screen are formed of vertical sub-pixels.

[0069] The advantages and technical effects of the screen according to the invention apply mutatis mutandis to a display method according to the invention.

[0070] Advantageously and according to the invention, each of the MxN viewpoints of the image are coded, for a panel formed of horizontal sub-pixels, with 6 / M different vertical pixels within the N elementary blocks of 6 pixels in height, M being equal to 1, 2, 3 or 6.

[0071] Advantageously and according to the invention, each of the MxN viewpoints of the image are coded, for a panel formed of vertical sub-pixels, with 6 / M different vertical sub-pixels within PxN elementary blocks of 6 / P sub-pixels in height, M being equal to 3 or 6.

[0072] When the screen according to the invention is oriented in vertical sub-pixel mode, the pixels are in a configuration in which the sub-pixels extend vertically and are P times higher than they are wide. In the case where each pixel is formed of three RGB sub-pixels (case where P is equal to 3), each square pixel is formed of three vertical sub-pixels juxtaposed horizontally.

[0073] The content is no longer addressed by pixel as for the screen according to the horizontal variant, but by sub-pixel. The distribution of the viewpoints (also designated by the term "mix") can then be done on a set of 4 consecutive lines (12 / 3) instead of 12. With an equivalent number of viewpoints, the viewpoints then spread horizontally in proportions three times greater than the mix of the horizontal version.

[0074] Each viewpoint is encoded on at least two lines, which again offers the possibility of either using fewer higher resolution viewpoints, or of “tiling” intermediate viewpoints depending on the depth of the relief in the scene.

[0075] In the vertical subpixel version, the subpixel of the same viewpoint under the adjacent lens is located 4 lines lower and on the following color. In the case where N is equal to 11, we thus obtain two encoding possibilities: 33 viewpoints each coded with 2 different vertical subpixels (case where M is equal to 3) or 66 viewpoints each coded with a single subpixel (case where M is equal to 6).

[0076] Whatever the version of the screen (horizontal or vertical), the method according to the invention makes it possible to adapt the number of points of view displayed according to the disparity in the image, in depth as well as in projection, with the ghosts of the details at the limits of the volume as arbiter.

[0077] One can either use fewer higher resolution viewpoints, or "tile" intermediate viewpoints based on the depth of relief in the scene.

[0078] The invention also relates to a display screen and a display method characterized in combination by all or part of the characteristics mentioned above or below. 5. List of figures

[0079] Other aims, characteristics and advantages of the invention will appear on reading the following description given solely for non-limiting purposes and which refers to the appended figures in which: • [Fig.l] is a schematic view of a display screen according to one embodiment of the invention, • [Fig.2] is a schematic view of a display screen according to a mode of realization lization of the invention, from eleven main points of view, • [Fig.3] is a schematic view of a display screen according to another mode embodiment of the invention, with seven main viewpoints and oriented in portrait mode, • [Fig.4] is a schematic view of a display screen according to another mode embodiment of the invention, with seven main viewpoints and oriented in landscape mode, • [Fig.5] is a schematic view of a display screen according to another mode embodiment of the invention, from forty-two points of view and oriented in portrait mode

[0080] 6. Detailed description of an embodiment of the invention

[0081] In the figures, the scales and proportions are not strictly respected, for the purposes of illustration and clarity.

[0082] [Fig.l] schematically illustrates a screen 10 for displaying an autostereoscopic image with 6xN viewpoints comprising a matrix of pixels 20 arranged in rows and columns. In other words, in the embodiment described, M is equal to 6.

[0083] The pixel matrix is ​​surmounted by a network of cylindrical lenses 30 (also referred to by the terminology of "lenticular network") each inclined relative to the columns by an angle equal to arctan(l / 12), that is to say by an angle of 4.76°. In [Fig.l], the inclination shown is accentuated solely for illustration purposes.

[0084] This array of cylindrical lenses 30 has a pitch of N / 2 pixels wide, such that each lens of the array extends above a vertical block of six active pixel zones of a column of the screen, called an elementary block, then a black intercolumn space and a new elementary block of six active pixel zones of the following column and so on to the bottom of the screen.

[0085] As indicated previously, each pixel is composed of a plurality of 3 active sub-pixels (P is equal to 3) of different colors and opaque inactive zones forming a portion of black interlines and / or intercolumns of the panel. That being said, the term "pixel" (and for the sake of simplification) designates the active zones of the pixel which make it possible to address content and the terms "black space" designate the opaque inactive zones of the pixels which form the interlines and / or intercolumns of the screen.

[0086] Reference 26 illustrates a pixel within the meaning of the invention, in its non-simplified version, that is to say the assembly formed by the active zone of the pixel and the black zone which surrounds this active zone of the pixel.

[0087] The lenticular network is for example a network manufactured according to the method described in application WO2020 / 178506 in the name of the applicant.

[0088] [Fig.2] schematically illustrates a portion of a pixel panel of a screen according to an embodiment of the invention comprising horizontal sub-pixels and on which a lens 30 of the lenticular network has been shown. This screen is designed to display an image at eleven main viewpoints (i.e. N is equal to H).

[0089] To do this, the pixel panel (or pixel matrix) comprises a plurality of pixels 26 arranged in rows and columns. Each pixel 26 is formed of three horizontal sub-pixels of different colors and corresponding to the R, G, B components.

[0090] The columns of pixels are separated from each other by a black inter-column space 24 of a width substantially equal to the width of an active area of ​​the pixel 26 of the screen. In [Fig.2], these black columns are shown in white for clarity.

[0091] Each lens 30 of the cylindrical lens array is inclined relative to the columns of the screen by an angle equal to arctan(l / 12), i.e. 4.76°.

[0092] The network is designed so that its pitch is 5.5 pixels wide (11 / 2). Each lens of the network 30 extends above a set of vertical blocks 22 of six pixels of a column of the screen, called elementary block 22, then a set of black inter-column spaces of equivalent height and a new set of elementary blocks of six pixels of the following column and so on to the bottom of the screen.

[0093] Such a screen allows 1xM viewpoints to be displayed, M being an integer which can take the value 1, 2, 3 or 6.

[0094] In [Fig.2], the eleven main viewpoints are represented by Roman numerals in italics. Thus, if we take the elementary block referenced 22, the main viewpoint of this block is viewpoint 1. The next elementary block in the same column encodes viewpoint 11.

[0095] It can be seen that under the lens 30 shown, the first line of elementary blocks encodes the odd viewpoints (7, 3, 5, 7, 9 and 77) and the 2nd line of elementary blocks still the even viewpoints (2, 4, 6, 8 and 10).

[0096] The eleven main viewpoints are therefore interlaced, under the lens, on twelve consecutive lines in the horizontal version. If the screen is designed in vertical mode, the eleven viewpoints are then interlaced on four consecutive lines (insofar as it is then possible to address the sub-pixels and not the entire pixels).

[0097] As can be seen in [Fig.2], to find the same viewpoint under the lens shown, it is necessary to shift one column to the right and twelve lines downwards. It is therefore clear that if the observer views viewpoint 1 with one eye and if he shifts slightly to the left (the observer's movement is the opposite of the horizontal shift of the viewpoints addressed to the eye due to the presence of the lens), he will then see viewpoint 2, then viewpoint 3, and so on up to viewpoint 77. These eleven viewpoints are interlaced on twelve lines of the pixel panel.

[0098] For each elementary block, it is also possible to assign to the other pixels of this elementary block, intermediate viewpoints between said main viewpoint of the block and the main viewpoint of the following block (in other words, it is possible to display an image with 1 IxM viewpoints, with M chosen from 1, 2, 3 and 6).

[0099] The optical axis of the lenticular network 30 intersects the pixels crossed at an angle 4.76°. This allows each of the six pixels of an elementary block crossed to be successively aligned with sub-parts of the solid angle occupied by each main viewpoint.

[0100] A pixel being formed of three sub-pixels, each elementary block 22 of six pixels is therefore made up of eighteen stacked horizontal sub-pixels. The crossing of an elementary block of six pixels is therefore in reality the progressive crossing of these eighteen sub-pixels.

[0101] When a distinct viewpoint is assigned to each of the six pixels of an elementary block 22, eighteen cases of partial overlap therefore ultimately occur.

[0102] The crossed sub-pixels must therefore be assimilated to theoretical points, located at the barycenter of the surface considered.

[0103] When the six pixels of an elementary block are assigned to six different viewpoints, overlaps are evident and blending is inevitable. However, the perceived quality is superior to that obtained with higher resolution main viewpoints.

[0104] The screen according to the embodiment of [Fig.2] therefore allows, as desired, the exploitation of contents at 11, 22, 33 or 66 points of view: - 11 main viewpoints each coded with six different vertical pixels and contiguous (M is equal to 1), - 22 viewpoints each coded with three different vertical pixels (M is equal to 2), - 33 viewpoints each coded with two different vertical pixels (M is equal to 3), - 66 viewpoints each coded with a single pixel (M is equal to 6).

[0105] [Fig.3] schematically illustrates a portion of a pixel panel of a screen according to another embodiment of the invention, oriented in horizontal sub-pixel mode and intended to display seven main viewpoints (N is equal to 7).

[0106] The difference compared to the screen of [Fig.2] is the lenticular network which has a pitch of 3.5 pixels instead of 5.5 pixels.

[0107] Each lens of the array 30 of cylindrical lenses is also each inclined relative to the columns of the screen by an angle equal to arctan(l / 12), i.e. 4.76°.

[0108] The seven main viewpoints are distributed under each lens on twelve consecutive lines, following the same principle as that explained in connection with [Fig.2]

[0109] [Fig. 3] shows the different distributions of the active zones of the opaque zones under the first two lenses, representative of the odd lenses and the even lenses. In particular, the portion of the pixel slab under viewpoint 1 of the first lens corresponds to an opaque zone under the second lens, and the adjacent opaque portion of viewpoint 1 under the first lens corresponding to viewpoint 2 under the second lens.

[0110] It is clearly seen that if the portion of the pixel panel under the first lens (characteristic of an odd lens) were superimposed with the portion of the pixel panel under the second lens (characteristic of an even lens), the active zones of the even lens correspond to the opaque zones of the odd lens and the opaque zones of the even lens correspond to the active zones of the odd lens.

[0111] [Fig.5] illustrates the same screen as that of [Fig.3], except that each pixel of the screen has been assigned a distinct image viewpoint. The screen therefore allows an image to be displayed at 42 viewpoints (N is equal to 7 and M is equal to 6).

[0112] It is observed that in the first elementary block of the first column represented, the viewpoints 42, 41, 40, 39, 38 and 37 are assigned to the six pixels of the elementary block. In the adjacent elementary block of the second column, the viewpoints 12, 11, 10, 09, 08 and 07 are assigned to the six pixels of the elementary block. In the following elementary block of the third column, the viewpoints 24, 23, 22, 21, 20 and 19 are assigned to the six pixels of the elementary block, and so on for the 7 elementary blocks arranged under the first lens on the first 12 lines of the pixel panel. The 7 elementary blocks are represented schematically by dotted rectangles each surrounding six consecutive pixels in the direction of the columns.

[0113] In other words, for each of the 7 elementary blocks, one of the 42 viewpoints was assigned to one of the pixels of the elementary block and the other pixels of this elementary block were assigned one of the 5 following viewpoints in the series of 42 viewpoints. For the first elementary block of the first column, viewpoint 37 can be considered as the main viewpoint and viewpoints 38, 39, 40, 41 and 42 can be considered as the intermediate viewpoints, formed by the following viewpoints in the series of 42 viewpoints.

[0114] Under the first lens shown, the first twelve lines of the screen thus allow the 42 viewpoints of the image to be displayed. It can be seen that viewpoint 42 is found under the first lens, in the second column and in the 13th line. Viewpoint 12 of the first line and the second column is found in the third column in the 13th line, and so on for each of the viewpoints. The same pattern is repeated for each lens and across the entire screen (not shown in the figure for clarity).

[0115] [Fig.4] illustrates a screen designed to display a maximum of 42 viewpoints, but with a vertical sub-pixel orientation. This screen therefore also has a lenticular network inclined relative to the screen columns at an angle corresponding to 4.76°. The seven main viewpoints are distributed, under the lenses, over four consecutive lines (instead of 12 for the horizontal sub-pixel version), in the since we can then address sub-pixels and not entire pixels.

[0116] Thus, we see that viewpoint 1 is encoded by the white sub-pixel of the first pixel at the top left of the figure. We find viewpoint 1 at the 5th row and 2nd column of the screen, under this same lens, encoded by the light gray sub-pixel of the pixel and at the 9th row and 3rd column of the screen, encoded by the black sub-pixel of the pixel. It is the same for all seven viewpoints of the image.

[0117] Like the screen in horizontal subpixel mode, we have the possibility of either using fewer higher resolution viewpoints, or of "tiling" intermediate viewpoints according to the depth of the relief in the scene.

[0118] Whatever the embodiment, the lenticular network has an angle which makes it possible to distribute each viewpoint both vertically and horizontally. These viewpoints are called "main" when a single viewpoint can occupy all the pixels of a vertical elementary block alone and they are called "intermediate" when several slightly offset viewpoints are "tiled" in this same block and complete the main viewpoint. The intermediate viewpoints can be considered as filler viewpoints, which fill the discrete parallax jumps between main viewpoints, so as to offer the viewer a sensation of continuous parallax. They serve to smooth any parallax jumps that may be perceptible between main viewpoints.

[0119] The invention has been described with a fixed M within the image. In other words, the number of viewpoints is determined uniformly as a function of the disparity targeted in the image. That being said, it is possible to use the screen with a dynamic variation of the number of viewpoints (and therefore of the integer M) within the same image as a function of the disparity. For example, it can be provided that M is fixed at 1 or 2 for little relief and that M is fixed at 3 or 6 for more relief. This makes it possible to limit the need to generate 3D information with a large number of different viewpoints to only the areas of high relief of an image.

Claims

Claims

1. Autostereoscopic display screen (10), said to be of photographic quality, of an autostereoscopic image with MxN points of view ordered from 1 to MxN, N being an odd integer and M being an integer chosen from 1, 2, 3 and 6, said screen comprising a panel of pixels (20) arranged in rows and columns, each pixel (26) being composed of a plurality of P active sub-pixels of different colors and opaque inactive zones forming a portion of black inter-rows and / or inter-columns of said panel, characterized in that it further comprises an ordered array of cylindrical lenses (30) mounted on said panel of pixels, each lens being inclined relative to the columns of the screen by an angle equal to arctan(l / 12),said network having a pitch of N / 2 pixels wide so as to be able on the one hand to differentiate the even and odd lenses (30) of said network ordered by distinct distributions of the active pixel zones (26) and the black zones (24) of the panel arranged under each even and odd lens, and, on the other hand to decode each of the MxN viewpoints distributed within N vertical blocks of 6 pixels in height from M viewpoints (6 / M pixels per viewpoint), with a periodic iteration of each block in the axis of a lens every 12 lines if the sub-pixels of the panel are horizontal, and distributed within PxN vertical blocks of 6 / P sub-pixels in height, from M / P viewpoints (6 / M sub-pixels per viewpoint), with a periodic iteration of each block in the axis of a lens, every 12 / P lines if the subpixels of the panel are vertical.,

2. Autostereoscopic display screen (10) according to claim 1, characterized in that each pixel (26) of said pixel panel is composed of a plurality of P horizontal sub-pixels of different colors juxtaposed on top of each other in the direction of the columns, so as to be able to code each of the MxN viewpoints with 6 / M different vertical pixels, which are repeated, under each lens, every 12 lines, M being equal to 1, 2, 3 or 6.

3. Autostereoscopic display screen (10) according to claim 1, characterized in that each pixel (26) of said pixel panel is composed of a plurality of P vertical sub-pixels of different colors juxtaposed next to each other in the direction of the lines so as to be able to code each of the MxN viewpoints with 6 / M sub- different vertical pixels, which are repeated, under each lens, every 12 / P lines, M being equal to 3 or 6.

4. Screen according to one of claims 1 to 3, characterized in that N is equal to 11.

5. Screen according to one of claims 1 to 4, characterized in that P is equal to 3.

6. Screen according to one of claims 1 to 5, characterized in that said pixel panel (20) is an 8K pixel panel.

7. A method for the photographic quality autostereoscopic display of an autostereoscopic image with MxN viewpoints ordered from 1 to MxN, N being an odd integer and M being an integer chosen from 1, 2, 3 and 6, characterized in that it comprises: • the selection of an autostereoscopic display screen (10) according to one of claims 1 to 6, • the distribution under each lens (30) of the MxN viewpoints distributed within N vertical blocks of 6 pixels in height from M viewpoints (6 / M pixels per viewpoint), with a periodic iteration of each block in the axis of a lens every 12 lines if the sub-pixels of the panel are horizontal, and distributed within PxN vertical blocks of 6 / P sub-pixels in height, from M / P viewpoints (6 / M sub-pixels per point-of-view), with a periodic iteration of each block in the axis of a lens, every 12 / P lines if the sub-pixels of the panel are vertical.

8. Method according to claim 7, characterized in that each of the MxN viewpoints of the image are coded, for a panel formed of horizontal sub-pixels, with 6 / M different vertical pixels within the N elementary blocks of 6 pixels in height, M being equal to 1, 2, 3 or 6.

9. Method according to claim 7, characterized in that each of the MxN viewpoints of the image are coded, for a panel formed of vertical sub-pixels, with 6 / M different vertical sub-pixels within PxN elementary blocks of 6 / P sub-pixels in height, M being equal to 3 or 6.

Images