Autostereoscopic screen considered to have photographic quality

Abstract

The present disclosure relates to a screen (10) for autostereoscopic display of an autostereoscopic image considered to have photographic quality, the autostereoscopic image having M * N viewpoints of M * N, N being an odd number, M being an integer selected from 1, 2, 3, and 6, the screen comprising a panel of pixels (20) arranged in rows and columns, each pixel (26) consisting of a plurality of P active sub-pixels of different colors, an ordered array (30) of cylindrical lenses is inclined relative to the columns of the screen at an angle equal to arctan (1 / 12), said array having a pitch of N / 2 pixels in width, so as to be able to distinguish between even and odd lenses (30) of said array by different distributions of active regions (26) and black regions (24) of pixels arranged below the lenses, and if the sub-pixels of the panel are horizontal, if the sub-pixels of the panel are not horizontal, even and odd lenses (30) of said array are not horizontal. Each of M * N viewpoints distributed within N vertical blocks at a height of 6 pixels can be decoded from M viewpoints per 12 rows if sub-pixels of the panel are vertical, and each of M * N viewpoints distributed within P * N vertical blocks at a height of 6 / P can be decoded from M / P viewpoints per 12 / P rows if sub-pixels of the panel are vertical.

Description

Technical Field

[0001] The invention relates to an autostereoscopic screen and a method for displaying autostereoscopic images on such an autostereoscopic screen. Background Art

[0002] Autostereoscopic technology is a technique that enables the three-dimensional display of images without requiring the viewer to wear special glasses. This technology is known per se, in particular from patents WO2006 / 024764, WO2014 / 041504, WO2013 / 140363, WO2014 / 016768, WO2019 / 063897, WO2019 / 207235, and WO2022 / 175053 in the name of the applicant.

[0003] An autostereoscopic image is typically formed from a number of interleaved elementary images, each corresponding to a view of the same object or scene from a different viewpoint. A selection mechanism, typically formed by a cylindrical microlens array or a parallax barrier, is positioned in front of the display screen. This allows a pair of elementary images, corresponding to two different viewpoints of the scene, to be projected towards each eye of the viewer, thereby creating a three-dimensional impression in their brain.

[0004] Applicants have proposed an autostereoscopic screen for displaying images with N viewpoints. The screen comprises a matrix of pixels arranged in rows and columns, with each pixel composed of multiple sub-pixels of different colors. Furthermore, above the screen is an array of identical cylindrical microlenses, each with a focal length configured to reflect light from the screen to infinity. The pitch of the array of cylindrical microlenses is precisely calculated so that, at a predetermined distance from the screen, a viewer continuously sees the image through a magnifying glass effect in the microlens array.

[0005] This magnifying glass effect is due to the fact that a lens placed at the correct distance (the focal length of the lens) magnifies the sub-pixels that are aligned with its optical axis and the viewer's eye pupil. If the lens magnifies N times, the sub-pixels seen through the lens are perceived as N times larger than they actually are, and N-1 other sub-pixels that are not in this alignment are obscured from the eye receiving light through the lens.

[0006] Even the principle of a lenticular array cannot provide viewers with continuous parallax of a scene represented through a 180° angle. Therefore, the viewing angle is subdivided into strips called "lobes," and parallax is discretized within each lobe. Within each lobe, N viewpoints are continuously visible. When exiting a lobe, the same sequence of N viewpoints is found in the next lobe, and so on, spanning the entire width of the screen's viewing area. The angular width of the lobe is determined by the pitch of the lenticular array and its focal length.

[0007] Autostereoscopic screens need to subdivide the lobes into clearly visible discrete subsections, with the width of a single lobe not exceeding the average interpupillary distance of 6.5 cm.

[0008] Therefore, one of the keys to the performance of the autostereoscopic screen currently proposed by the applicant is to control the resolution of the optical elements developed specifically for each pixel panel.

[0009] For printed autostereoscopic photographs (as opposed to autostereoscopic screens), the maximum number of continuous but well-distinguishable printed dots or points that can be written on a medium that provides good dimensional stability is achieved. With currently used printing technology, the applicant has managed to print 100 dots per millimeter, and sometimes even more, which is equivalent to seeing dots of 10 microns or less.

[0010] When a viewer moves in front of such a photograph, he sees the different elements of the scene moving relative to each other in a continuous manner and in exactly the same physical volume as if they were seen from the same distance. Each eye simultaneously sees two different perspectives of the scene, which enables the brain to reconstruct a perfectly natural three-dimensional perception.

[0011] A priori, it is impossible to determine the quality of autostereoscopic images with screens that are formed by a matrix of pixels of higher or lower resolution, consisting of rows and columns arranged orthogonally and separated from each other by a black surround, which is made up of opaque rows and columns of greater or lesser thickness depending on the panel, forming a grid pattern around the active area of the pixels. The surface ratio between the black surround and the colored active sub-pixels affects not only the brightness and contrast of the screen, but also its autostereoscopic resolution.

[0012] Throughout this article, an autostereoscopic screen is considered to be of photographic quality when it can eliminate the discrete jumps from one stereo pair to another in the parts of the image where the differences are greatest.

[0013] In other words, “photographic quality” is understood to refer to the continuity of the three-dimensional space within the lobe, which an autostereoscopic screen seems a priori to be unable to achieve without a tracking system.

[0014] In addition to issues related to resolution, there are also issues related to motion. The lenticular array actually expands the active area of the pixel panel and the surrounding black surround. This causes certain subsections of the viewing lobe to appear darker or brighter than others. These changes in brightness are often noticeable at the transition between two adjacent viewpoints. The moiré effect on light intensity varies depending on the specific panel structure and the orientation of the screen.

[0015] The inventors have attempted to overcome the technical prejudice of those skilled in the art, according to which it is a priori impossible to find the quality of a photographic autostereoscopic image of a screen with a panel of pixels (characterized by the continuity of the three dimensions within the lobes). Summary of the Invention

[0016] The present invention aims to provide an autostereoscopic screen that is considered to be of photographic quality.

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

[0018] Thus, the present invention aims, in at least one embodiment, to provide a screen that provides a continuous parallax perception in each lobe.

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

[0020] The present invention also aims to provide a screen that is considered to have photographic quality regardless of the structure of the sub-pixels of the screen's pixel panel (vertical or horizontal sub-pixels).

[0021] Finally, the present invention aims to provide a method for displaying autostereoscopic images on a screen according to the invention.

[0022] To this end, the invention relates to an autostereoscopic display screen for autostereoscopic images, which is considered to be of photographic quality, the autostereoscopic images having M×N viewpoints ordered from 1 to M×N, N being an odd number strictly greater than 1 and M being an integer selected from 1, 2, 3 and 6.

[0023] A screen according to the present disclosure comprises a panel of pixels arranged in rows and columns, each pixel (26) consisting of a plurality of P active sub-pixels of different colors and an opaque inactive area forming part of the black inter-row and / or inter-column areas of the panel.

[0024] The screen according to the present disclosure is characterized in that it further comprises an ordered array of cylindrical lenses mounted on said pixel panel, each lens being inclined with respect to a column of the screen by an angle equal to arctan(1 / 12).

[0025] Furthermore, the array has a pitch of N / 2 pixels in width, and the even and odd lenses of the ordered array can be distinguished by the different distributions of the active areas of the pixels arranged under each even and odd lens and the black areas of the panel.

[0026] In addition, if the sub-pixels of the panel are horizontal, the array can decode each of the M×N viewpoints distributed in N vertical blocks with a height of 6 pixels from M viewpoints (6 / M pixels per viewpoint), and each block is periodically iterated every 12 rows on the axis of the lens, and if the sub-pixels of the panel are vertical, the array can decode each of the M×N viewpoints distributed in P×N vertical blocks with a height of 6 / P sub-pixels from M / P viewpoints (6 / M sub-pixels per viewpoint), and each block is periodically iterated every 12 / P rows on the axis of the lens.

[0027] In other words, the screen has the following characteristics: each of the M×N viewpoints can be encoded on the panel extending below the lens, and if the sub-pixels of the panel are horizontal, then the 6 / M different vertical pixels within N vertical blocks of 6 pixels in height are repeated every 12 rows below each lens, and if the sub-pixels of the panel are vertical, then the 6 / M different vertical sub-pixels within P×N vertical blocks of 6 / P sub-pixels in height are repeated every 12 / P rows below each lens.

[0028] The array is constructed to decode each of the M×N viewpoints encoded according to the above principle by means of a magnifying glass effect.

[0029] The screen according to the invention is also characterized by an array of cylindrical microlenses, each of which is tilted relative to the columns of the screen by an angle equal to arctan(1 / 12), i.e., an angle of the order of 4.76°. The "arctan" function refers to the inverse tangent. The inverse tangent of a real number is the value of an angle whose tangent is equal to this number.

[0030] The screen also features a lens array that is constructed to distribute the viewpoints under each lens over 12 consecutive rows if the sub-pixels are horizontal, and 12 / P consecutive rows if the sub-pixels are vertical.

[0031] To simplify terminology, the "horizontal mode" or "horizontal version" of a screen refers to a screen composed of horizontal sub-pixels, and the "vertical mode" or "vertical version" of a screen refers to a screen composed of vertical sub-pixels. It should be understood that this does not indicate the vertical or horizontal orientation of the screen. Horizontal and vertical modes only characterize the orientation of the sub-pixels, not the vertical or horizontal orientation of the screen.

[0032] In the following, unless otherwise specified, to simplify the terminology, "pixel" refers to the active area of the pixel that can process content, and "black space" refers to the opaque, non-active area of the pixel that forms the inter-row and / or inter-column areas of the screen.

[0033] In horizontal mode, each lens of the lens array is tilted 4.76 degrees, passes through a set of 6 pixel basic vertical blocks, then passes through a black space, and then encounters a new set of 6 pixel basic blocks. It should be noted that all pixel matrices have a black opaque area between each row of active pixels, but this row (as a component of the pixel) can benefit from good resolution according to its percentage in the pixel.

[0034] Since N is an odd number, the ordered array of the screen has a non-integer pitch of N / 2, which makes it possible to distinguish between even lenses and odd lenses by at least partially setting the distribution of the light area of the screen (the active area of the pixels) relative to the distribution of the black area (the opaque area of the pixels). In other words, if the lenses of the ordered array are ordered so that, from the left to the right side of the screen, the pixel panel below the even lenses has the same distribution of the active pixel area and the black area of the screen, while the pixel panel below the odd lenses has the same distribution of the active pixel area and the black area of the screen, the distribution of the even lenses is different from the distribution of the odd lenses. If a portion of the pixel panel below the even lenses overlaps a portion of the pixel panel below the odd lenses, it should be noted that at least a portion of the dark area of the panel below the even lenses faces at least a portion of the active area of the panel below the odd lenses, and at least a portion of the active area of the panel below the even lenses faces at least a portion of the dark area of the panel below the odd lenses.

[0035] For an array of 11×M viewpoints (N equals 11), the pitch of the lenses has a width of 5.5 pixels (according to the non-simplified definition of a pixel). For an array of 7×M viewpoints (N equals 7), the pitch of the lenses has a width of 3.5 pixels.

[0036] The screen array according to the present invention is such that: if the sub-pixels of the panel are horizontal, each of the M×N viewpoints distributed in N vertical blocks with a height of 6 pixels can be decoded from M viewpoints (6 / M pixels per viewpoint) through a magnifying glass effect, and each block is periodically iterated every 12 rows on the axis of the lens; if the sub-pixels of the panel are vertical, each of the M×N viewpoints distributed in P×N vertical blocks with a height of 6 / P sub-pixels can be decoded from M / P viewpoints (6 / M sub-pixels per viewpoint) through a magnifying glass effect, and each block is periodically iterated every 12 / P rows on the axis of the lens.

[0037] In the horizontal version, the advantage of having vertical blocks of 6 pixels (also referred to in the term basic blocks) is that it provides an alternative to assigning the same viewpoint or even different viewpoints to each of these pixels.

[0038] When the scene being represented is shallow, displaying the same viewpoint (called the primary viewpoint) on 6 consecutive pixels of a basic block rarely causes a problem.

[0039] On the other hand, when the disparity is large, the perceived quality deteriorates. Perfect focus of the array results in a clear discretization of content at great depth or with strong prominence, which compromises the intended photographic experience (continuous parallax). When the horizontal shift of contrasting details in the image during a viewpoint change exceeds two or three lenses, the discretization becomes excessive, and increasing the number of viewpoints becomes necessary. While this compromises the initial resolution of the displayed primary viewpoint, the improvement in perceived quality is noticeable.

[0040] If a slightly different viewpoint is assigned to each of the 6 pixels of each basic block, the discretization disappears in scenes with strong disparity.

[0041] The screen according to the present invention thus makes it possible to display viewpoints between the N primary viewpoints of the displayed image, referred to as intermediate viewpoints. These intermediate viewpoints are slightly shifted between each primary viewpoint to distribute them within each elementary block. The continuity between successive primary viewpoints creates an effect of photographic depth in the distant background and strongly highlighted foreground, rather than a superficial discretization.

[0042] In other words, the remarkable thing about the screen according to the invention is that it can encode slightly different viewpoints using groups of 6 pixels in height for each primary pixel (of which it is part) so as to fill in a discrete sequence of primary viewpoints with a continuum, approaching the quality of an autostereoscopic image.

[0043] Through this novel approach, which we refer to as the “tiling effect” throughout the article, our goal of finding parallax continuity of printed photographic images on screen was achieved.

[0044] In particular, in 3D images, the differences between viewpoints are exacerbated as one moves away from the plane of the screen (also known as the collimation plane). The overlap between common viewpoints (called "crosstalk") is the same across the entire screen, affecting the entire displayed image but only showing up in those parts of the image where the differences between viewpoints are greater.

[0045] In particular, the inventors have recognized that photographic blur can be information related to three dimensions and a manifestation of depth of field. Therefore, contrary to preconceptions in the technical field, the inventors have attempted to achieve depth or relief in images through gradual fades rather than through discretization between viewpoints. The overall impression aimed at is a deep, continuous image, where the fade between viewpoints becomes more pronounced as the object moves away from the plane of collimation, perceived as the plane of focus.

[0046] Thus, discretization between viewpoints is achieved when recognizable details in the image are repeated multiple times. When the differences in the scene are large and the number of viewpoints is insufficient to distribute them harmoniously between each viewpoint, contrasting objects far from the collimation plane are typically degraded by the perception of ghosting.

[0047] Instead, deep details from multiple viewpoints can overlap simultaneously, and the intensity and contrast in the perceived redundancy diminish as one moves away from the central viewpoint, so that even if the image of a single object remains somewhat blurry, it is visually reshaped.

[0048] The screen according to the invention offers the possibility of adjusting the number of displayed viewpoints according to the differences in both depth and prominence in the image (by modifying the value of M), where the ghosting of details at the limits of the volume is a decisive factor. This novel screen offers the viewer continuous parallax within the lobe, without a noticeable discretization, but rather a gradual fading of details as the limits of the volume are reached.

[0049] This approach is in contrast to screens proposed so far, where the goal is to seek a perfect discretization of the viewpoints, which is considered the path to providing high-quality screens.

[0050] Advantageously, according to the present invention, each pixel of the pixel panel is composed of a plurality of P horizontal sub-pixels of different colors, which are arranged parallel to each other in the column direction so that each of the M×N viewpoints can be encoded with 6 / M different vertical pixels, and the 6 / M different vertical pixels are repeated every 12 rows under each lens, where M is equal to 1, 2, 3 or 6.

[0051] In other words, the screen according to the invention can be displayed as desired in the horizontal version:

[0052] - N main viewpoints, each of which is encoded by 6 different consecutive vertical pixels (M is equal to 1 in this case),

[0053] - 2×N views, each coded with 3 different vertical pixels (M equals 2 in this case),

[0054] - 3×N views, each coded with 2 different vertical pixels (M equals 3 in this case),

[0055] - 6×N views, each coded with a single pixel (M equals 6 in this case).

[0056] Advantageously, according to the present invention, each pixel of the pixel panel is composed of a plurality of P vertical sub-pixels of different colors arranged parallel to each other in the row direction, so that each of the M×N viewpoints can be encoded with 6 / M different vertical sub-pixels, and the 6 / M different vertical sub-pixels are repeated every 12 / P rows under each lens, where M is equal to 3 or 6.

[0057] In other words, the screen according to the invention can display, in the vertical version, as required:

[0058] - 3×N views, each coded with 2 different vertical sub-pixels (M equals 3 in this case),

[0059] - 6×N views, each sub-encoded with a single pixel (M equals 6 in this case).

[0060] Furthermore, the screen according to the present invention has narrower lobes than screens previously proposed by the applicant. To achieve this, the optical elements have a relatively long focal length, so that the lobes are as tight as possible without compromising the ease with which viewers can position themselves. This improves resolution and angular continuity. While narrower lobes increase the transition zone, they also divide the viewing space in front of the screen into narrower strips, within which many viewers can comfortably position themselves simultaneously at varying distances.

[0061] The present invention thus contradicts the prejudice of those skilled in the art who believe that larger lobe widths are inextricably linked to the quality of autostereoscopic screens. Contrary to what one might assume, when all six pixels of a basic block belong to the same high-resolution viewpoint, the perceived image quality is comparable to that achieved when lower-resolution intermediate viewpoints are "tiled" within this basic block rather than expanding the primary viewpoint. In other words, displays with very high theoretical resolutions do not prove to be better than displays with partially overlapping intermediate viewpoints, resulting in a "tiled" effect. On the other hand, when resolution is optimized, rather than dispersing into a circular-looking blur in the manner of a depth-of-field effect, deep details are discretized. The simultaneous expansion of intermediate viewpoints in both the horizontal and vertical directions makes it possible to achieve this quasi-isotropic improvement in the final image.

[0062] Humans are accustomed to depth of field blur, and perceiving blur in the most distant planes of a 3D image is physiologically inevitable. The brain easily accepts that blur is caused by the distance of these planes, and this does not adversely affect the overall sharpness we perceive in the image.

[0063] According to one variant of the invention, the screen is characterized by being provided with a pixel panel which, when oriented in horizontal mode (or respectively in vertical mode), has columns (or respectively rows), each separated from another by black inter-column areas (or respectively black inter-row areas) whose width (or height) is close to the width of the active area of the panel pixels. These black surfaces form a black surround whose surface is close to the active surface of the screen. This makes it possible to obtain an excellent contrast.

[0064] Furthermore, according to this variation, in the case of horizontal sub-pixels, the active and dark areas of the panel portions arranged below the even and odd lenses are arranged opposite each other. In other words, the active areas below the even lenses correspond to the dark areas below the odd lenses, and the dark areas below the even lenses correspond to the active areas below the odd lenses.

[0065] Advantageously, according to the invention, N is equal to 7 or 11.

[0066] The inventors have determined that a finer lens array that can display only 7 or 11 main viewpoints can further marginalize the structural effect. In this case, there is naturally nothing to prevent using a different number of viewpoints.

[0067] Advantageously, according to the invention, P is equal to 3.

[0068] According to this variant, when the screen is oriented in horizontal mode, each pixel of the pixel matrix consists of three horizontal sub-pixels of different colors juxtaposed in the column direction, with the columns separated by black inter-column spaces. Of course, when the screen is oriented in vertical mode, each pixel of the matrix consists of three vertical sub-pixels of different colors juxtaposed in the row direction, with the rows separated by black inter-row spaces.

[0069] However, there is nothing to prevent the use of pixel panels where each pixel consists of four subpixels. In particular, in some panels, a white subpixel is added to the three common colors of red, green, and blue.

[0070] Advantageously, according to the invention, said pixel matrix is an 8K pixel panel.

[0071] Such pixel panels make it possible to use the highest resolution currently available, for example a 31.5-inch panel with a resolution of 7680×4320 pixels, oriented in portrait mode, with a matrix of pixels each formed by horizontal RGB stripes.

[0072] In fact, the higher the resolution of the screen, the greater the possibility of displaying a large number of viewpoints on it simultaneously. These additional viewpoints can therefore be used to widen the lobe by reducing the focal length at a constant pitch, or to reduce the width of the lobe occupied by each viewpoint by dedicating a subsection of a narrower lobe to it, in order to achieve a closer photographic effect.

[0073] The present invention also relates to a method for the automatic stereoscopic display of photographic quality of an autostereoscopic image having M×N viewpoints ordered from 1 to M×N, N being an odd number strictly greater than 1 and M being an integer selected from 1, 2, 3 and 6, characterized in that the method comprises:

[0074] - selecting an autostereoscopic display screen according to the invention,

[0075] If the sub-pixels of the panel are horizontal, M×N viewpoints are distributed under each lens, and the M×N viewpoints are distributed from M viewpoints (6 / M pixels per viewpoint) in N vertical blocks with a height of 6 pixels, and each block is periodically iterated every 12 rows on the axis of the lens, and, if the sub-pixels of the panel are vertical, M×N viewpoints are distributed under each lens, and the M×N viewpoints are distributed from M / P viewpoints (6 / M sub-pixels per viewpoint) in P×N vertical blocks with a height of 6 / P sub-pixels, and each block is periodically iterated every 12 / P rows on the axis of the lens.

[0076] In other words, the automatic stereoscopic display method according to the present invention makes it possible to distribute M×N viewpoints of the image under each lens within N basic blocks if the pixels of the screen are formed by horizontal sub-pixels, and the N basic blocks are repeated every 12 consecutive rows. If the pixels of the screen are composed of vertical sub-pixels, it is possible to distribute M×N viewpoints of the image under each lens within N×P basic blocks, and the N×P basic blocks are repeated every 12 / P consecutive rows.

[0077] The advantages and technical effects of the screen according to the invention apply mutatis mutandis to the display method according to the invention.

[0078] Advantageously, according to the invention, for a panel formed by horizontal sub-pixels, encoding each of the M×N views of the image, there are 6 / M different vertical pixels within N basic blocks of height 6 pixels, M being equal to 1, 2, 3 or 6.

[0079] Advantageously, according to the invention, for a panel formed by vertical sub-pixels, for each of the M×N viewpoints of the image, there are 6 / M different vertical sub-pixels within P×N basic blocks of a height of 6 / P sub-pixels, M being equal to 1, 2, 3 or 6.

[0080] When the screen according to the present invention is oriented in vertical sub-pixel mode, the pixels are arranged in a configuration where the sub-pixels extend vertically and are P times taller than they are wide. With each pixel formed from three RGB sub-pixels (in this case P equals 3), each square pixel is formed from three vertical sub-pixels juxtaposed horizontally.

[0081] According to the horizontal variant, content is no longer addressed at the pixel level, as in the case of the screen, but at the sub-pixel level. Therefore, the distribution of viewpoints (also known as "mixing") can be achieved over a group of four consecutive lines (12 / 3) instead of 12 lines. Given an equal number of viewpoints, the viewpoints are distributed horizontally at a ratio three times greater than the mixing ratio of the horizontal version.

[0082] Each viewpoint is encoded on at least two lines, in which case this also offers the possibility of using fewer high-resolution viewpoints or "tiling" intermediate viewpoints depending on the depth of the three dimensions in the scene.

[0083] In the vertical sub-pixel version, the sub-pixels of the same viewpoint under adjacent lenses are located 4 rows below and on the subsequent colors. With N equal to 11, two encoding possibilities are thus obtained: 33 coded views, each with 2 different vertical sub-pixels (in this case M equals 3), or 66 views, each coded with a single sub-pixel (in this case M equals 6).

[0084] Whatever the version of the screen (horizontal or vertical), the method according to the invention makes it possible to adapt the number of displayed viewpoints according to the disparity, depth and prominence in the image, with ghosting of details at volume limits being the determining factor.

[0085] You can use fewer high-resolution viewpoints, or you can "tile" intermediate viewpoints based on the depth of the three dimensions in the scene.

[0086] The present invention also relates to a display screen and a display method, which are characterized by a combination of all or some of the features mentioned above or below. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Other objects, features and advantages of the present invention will become apparent on reading the following description, given in a non-limiting manner only and with reference to the accompanying drawings, in which:

[0088] Figure 1 is a schematic diagram of a display screen according to an embodiment of the present invention,

[0089] Figure 2 is a schematic diagram of a display screen having eleven main viewpoints according to an embodiment of the present invention,

[0090] Figure 3 is a schematic diagram of a display screen having seven viewpoints and oriented in portrait mode according to another embodiment of the present invention,

[0091] Figure 4 is a schematic diagram of a display screen having seven primary viewpoints and oriented in landscape mode according to another embodiment of the present invention,

[0092] Figure 5 is a schematic diagram of a display screen having forty-two viewpoints and oriented in portrait mode according to another embodiment of the present invention. DETAILED DESCRIPTION

[0093] In the drawings, for the sake of illustration and clarity, the drawing and proportions have not been strictly adhered to.

[0094] Figure 1 Schematically shown is a screen 10 for displaying an autostereoscopic image with 6×N viewpoints comprising a matrix of pixels 20 arranged in rows and columns. In other words, M is equal to 6 in the embodiment described.

[0095] On top of the pixel matrix is an array of cylindrical lenses 30 (also called a "lens array"), each of which is tilted relative to the column at an angle equal to arctan(1 / 12), or 4.76°. Figure 1 , the inclination shown is emphasized for illustration purposes only.

[0096] The pitch of the cylindrical lens array 30 is N / 2 pixels in width, so each lens of the array extends over a vertical block consisting of the six pixel active areas of one column of the screen, called a basic block, followed by a black inter-column space and a new basic block consisting of the six pixel active areas of the next column, and so on, to the bottom of the screen.

[0097] As mentioned above, each pixel consists of a plurality of active sub-pixels of three different colors (P equals 3) and an opaque, non-active area that forms part of the black inter-row and / or inter-column areas of the panel. In this case (for simplicity) the term "pixel" refers to the active area of the pixel that can address content, while the term "black space" refers to the opaque, non-active area of the pixel that forms the inter-row and / or inter-column areas of the screen.

[0098] Reference numeral 26 denotes a pixel in a non-simplified version of the present invention, ie a group formed by an active area of the pixel and a black area surrounding the active pixel area.

[0099] For example, the lens array is an array produced according to the method described in application WO2020 / 178506 in the name of the applicant.

[0100] Figure 2 Schematically shows a portion of a pixel panel of a screen according to an embodiment of the invention, the screen comprising horizontal sub-pixels and on which lenses of a lens array 30 are shown. The screen is intended to display images with 11 primary viewpoints (ie N equals 11).

[0101] To this end, the pixel panel (or pixel matrix) comprises a plurality of pixels 26 arranged in rows and columns. Each pixel 26 is composed of three horizontal sub-pixels of different colors, corresponding to the R, G, B components.

[0102] The columns of pixels are separated from each other by black inter-column spaces 24, the width of the inter-column spaces being substantially equal to the width of the active area of the pixels 26 of the screen. Figure 2, these black columns are shown in white for clarity.

[0103] Each lens 30 of the cylindrical lens array is tilted relative to the columns of the screen at an angle equal to arctan(1 / 12), ie 4.76°.

[0104] The array is designed so that its pitch is 5.5 (11 / 2) pixels across. Each lens of the array 30 extends over a set of vertical blocks 22 of six pixels for one column of the screen, called basic blocks 22, followed by a set of black inter-column spaces of equal height and a new set of basic blocks of six pixels for the next column, and so on to the bottom of the screen.

[0105] Such a screen can display 11×M viewpoints, where M is an integer whose value can be 1, 2, 3 or 6.

[0106] Figure 2 The 11 primary views are shown using italic Roman numerals. Thus, for a basic block numbered 22, the primary view for that block is view 1. The subsequent basic block in the same column encodes view 11.

[0107] It will be noted that under the illustrated lens 30, the first row of basis blocks encodes the odd views (1, 3, 5, 7, 9 and 11) and the second row of basis blocks encodes the even views (2, 4, 6, 8 and 10).

[0108] Thus, in the horizontal version, the 11 main viewpoints are interleaved over 12 consecutive rows below the lens. If the screen is designed for vertical mode, the 11 viewpoints are therefore interleaved over four consecutive rows (as long as this allows addressing sub-pixels rather than whole pixels).

[0109] like Figure 2 As shown, in order to find the same viewpoint under the lens shown, one must shift one column to the right and twelve rows down. It is therefore easy to understand that if a viewer sees viewpoint 1 with one eye, and if he moves slightly to the left (due to the presence of the lens, the viewer's movement is opposite to the horizontal shift of the viewpoint toward the eye), then he will see viewpoint 2, then viewpoint 3, and so on, up to viewpoint 11. These 11 viewpoints are interlaced on the 12 rows of the pixel panel.

[0110] For each basic block, intermediate views between the main view of the block and the main view of the next block can also be allocated to other pixels of the basic block (in other words, an image with 11×M viewpoints can be displayed, where M is selected from 1, 2, and 6).

[0111] The optical axis of the lens array 30 cuts the traversed pixels at an angle of 4.76°. This makes it possible to continuously align each of the six traversed pixels of a basic block with a sub-portion of the solid angle occupied by each primary viewpoint.

[0112] Since a pixel is composed of three sub-pixels, each 6-pixel basic block 22 is composed of 18 stacked horizontal sub-pixels. Therefore, traversing the 6-pixel basic block is actually a step-by-step traversal of these 18 sub-pixels.

[0113] When a different viewpoint is assigned to each of the six pixels of the basic block 22, 18 cases of partial coverage actually occur.

[0114] Therefore, the traversed sub-pixels must be assimilated to a theoretical point located at the centroid of the relevant surface.

[0115] When the 6 pixels of a basic block are assigned to 6 different viewpoints, overlap is noticeable and blending is inevitable. However, the perceptual quality is higher than that obtained using the primary viewpoint with a better resolution.

[0116] Therefore, according to Figure 2 The screen of an embodiment can utilize content with 11, 22, 33, or 66 viewpoints as needed:

[0117] - 11 main viewpoints, each of which is encoded by six different consecutive vertical pixels (M is equal to 1),

[0118] - 22 main viewpoints, each of which is encoded by three different vertical pixels (M is equal to 2),

[0119] - 33 main viewpoints, each of which is encoded by two different vertical pixels (M is equal to 3),

[0120] - 66 primary viewpoints, each encoded by a single pixel (M equals 6).

[0121] Figure 3 There is schematically shown a portion of a pixel panel of a screen according to another embodiment of the invention, the screen being oriented in a horizontal sub-pixel pattern and intended to display seven primary viewpoints (N equals 7).

[0122] and Figure 2 The screen differs in that the lenticular array has a 3.5-pixel pitch instead of 5.5 pixels.

[0123] Each lens of the cylindrical lens array 30 is also tilted, with the angle at which each lens is tilted relative to the columns of the screen being equal to arctan(1 / 12), ie 4.76°.

[0124] According to Figure 2Using the same principle explained in

[15] , the seven primary viewpoints are distributed on 12 consecutive rows below each lens.

[0125] Figure 3 The figure shows the different distributions of active and opaque areas under the first two lenses, representing odd-numbered and even-numbered lenses, respectively. In particular, the portion of the pixel panel under viewpoint 1 of the first lens corresponds to the opaque area under the second lens, while the adjacent opaque portion of viewpoint 1 under the first lens corresponds to viewpoint 2 under the second lens.

[0126] It will be clearly observed that if the portion of the pixel panel below the first lens (characteristic of the odd lenses) is superimposed on the portion of the pixel panel below the second lens (characteristic of the even lenses), the active area of the even lenses corresponds to the opaque area of the odd lenses, and the opaque area of the even lenses corresponds to the active area of the odd lenses.

[0127] Figure 5 Shown with Figure 3 The screen is the same screen, the only difference is that each pixel of the screen has been assigned a different image viewpoint. Therefore, the screen can display images with 42 viewpoints (N is equal to 7 and M is equal to 6).

[0128] It will be noted that in the first basic block in the first column shown, viewpoints 42, 41, 40, 39, 38, and 37 are assigned to six pixels of the basic block. In the adjacent basic block in the second column, viewpoints 12, 11, 10, 09, 08, and 07 are assigned to six pixels of the basic block. In the subsequent basic block in the third column, viewpoints 24, 23, 22, 21, 20, and 19 are assigned to six pixels of the basic block, and so on for the seven basic blocks arranged on the first 12 rows of the pixel panel under the first lens. These seven basic blocks are schematically represented by dashed rectangles, with each dashed line enclosing six consecutive pixels in the column direction.

[0129] In other words, for each of the seven basic blocks, one of the 42 viewpoints is assigned to a pixel of the basic block, and one of the next five viewpoints in the sequence of 42 viewpoints is assigned to the other pixels of the basic block. For the first basic block in the first column, viewpoint 37 can be considered the main viewpoint, and views 38, 39, 40, 41, and 42 can be considered intermediate views formed by the subsequent views in the sequence of 42 viewpoints.

[0130] Under the first lens shown, the first 12 rows of the screen can display 42 viewpoints of the image. Note that in the second column and thirteenth row, viewpoint 42 is located below the first lens. Viewpoint 12 in the first row and second column is located in the third column of the thirteenth row, and so on for each viewpoint. The same pattern is repeated for each lens and across the entire screen (not shown for clarity).

[0131] Figure 4 The figure shows a screen designed to display up to 42 viewpoints, but with a vertical sub-pixel orientation. Consequently, this screen also exhibits a lens array tilted at an angle corresponding to 4.76° relative to the screen's fissures. Below the lenses, the seven primary viewpoints are distributed over four consecutive rows (instead of the 12 rows of the horizontal sub-pixel version), allowing for processing of sub-pixels rather than entire pixels.

[0132] Thus, it will be noted that viewpoint 1 is encoded by the white sub-pixel of the first pixel at the top left corner of the diagram. Under the same lens, viewpoint 1 is located at row 5 and column 2 of the screen, encoded by the light gray sub-pixel of the pixel, and is located at row 9 and column 3 of the screen, encoded by the black sub-pixel of the pixel. The same applies to all 7 viewpoints of the image.

[0133] Following the example of a screen in horizontal sub-pixel mode, it is possible to use fewer high-resolution viewpoints, or to "tile" intermediate viewpoints based on the depth of the three dimensions in the scene.

[0134] Regardless of the embodiment, the lens array has an angle that allows each viewpoint to be distributed vertically and horizontally. When the same viewpoint can only occupy a group of pixels in a vertical basic block, these viewpoints are called "primary" viewpoints. When multiple slightly offset viewpoints are "tiled" in the same block and complete the primary viewpoint, these viewpoints are called "intermediate" viewpoints. Intermediate viewpoints can be thought of as filler viewpoints that fill in the discrete parallax jumps between the primary viewpoints, thereby providing a continuous parallax perception to the viewer. They are used to smooth out the parallax jumps that may be perceived between the primary viewpoints.

[0135] The present invention describes the use of a fixed M within an image. In other words, the number of viewpoints is uniformly determined based on the target disparity within the image. In this case, the screen can be used with a dynamically varying number of viewpoints (hence the integer M) within the same image, depending on the disparity. For example, M can be fixed to 1 or 2 for low levels of three-dimensionality, while M can be fixed to 3 or 6 for greater three-dimensionality. This makes it possible to restrict only the highly three-dimensional regions of the image, requiring the generation of 3D information with a large number of different viewpoints.

Claims

1. A screen (10) for autostereoscopic display of autostereoscopic images, considered to be of photographic quality, said autostereoscopic images having M×N viewpoints ordered from 1 to M×N, N being an odd number strictly greater than 1 and M being an integer selected from 1, 2, 3, and 6, said screen comprising a panel of pixels (20) arranged in rows and columns, each pixel (26) consisting of a plurality of P active sub-pixels of different colors and an opaque non-active area forming part of the black inter-row and / or inter-column areas of said panel, It is characterized by Also included is an ordered array (30) of cylindrical lenses mounted on the pixel panel, each lens being tilted relative to the columns of the screen by an angle equal to arctan (1 / 12), the array having a pitch of N / 2 pixels in width so as to be able to distinguish the even and odd lenses (30) of the ordered array by the different distribution of the active areas of the pixels (26) arranged below each even and odd lens and the black areas (24) of the panel, and on the other hand: if the sub-pixels of the panel is horizontal, it is possible to decode each of the M×N viewpoints distributed in N vertical blocks of 6 pixels in height from M viewpoints (6 / M pixels per viewpoint), each block periodically iterating every 12 rows on the axis of the lens, and if the sub-pixels of the panel are vertical, it is possible to decode each of the M×N viewpoints distributed in P×N vertical blocks of 6 / P sub-pixels in height from M / P viewpoints (6 / M sub-pixels per viewpoint), each block periodically iterating every 12 / P rows on the axis of the lens.

2. The autostereoscopic display screen (10) according to claim 1, characterized in that: Each pixel (26) of the pixel panel is composed of a plurality of P horizontal sub-pixels of different colors arranged parallel to each other in the column direction so that each of the M×N viewpoints can be encoded with 6 / M different vertical pixels, and the 6 / M different vertical pixels are repeated every 12 rows under each lens, where M is equal to 1, 2, 3 or 6.

3. The autostereoscopic display screen (10) according to claim 1, characterized in that: Each pixel (26) of the pixel panel is composed of a plurality of P vertical sub-pixels of different colors arranged parallel to each other in the row direction so that each of the M×N viewpoints can be encoded with 6 / M different vertical sub-pixels, and the 6 / M different vertical sub-pixels are repeated every 12 / P rows under each lens, where M is equal to 3 or 6.

4. The screen according to any one of claims 1 to 3, characterized in that: N is equal to 11.

5. The screen according to any one of claims 1 to 4, characterized in that: P is equal to 3.

6. The screen according to any one of claims 1 to 5, characterized in that: The pixel panel (20) is an 8K pixel panel.

7. A method for automatically displaying photographic quality of an autostereoscopic image having M×N viewpoints ordered from 1 to M×N, N being an odd number strictly greater than 1, and M being an integer selected from 1, 2, 3, and 6, characterized in that The method comprises: Selecting an autostereoscopic display screen (10) according to any one of claims 1 to 6, If the sub-pixels of the panel are horizontal, M×N viewpoints are distributed below each lens (30) from M viewpoints (6 / M pixels per viewpoint), and the M×N viewpoints are distributed in N vertical blocks with a height of 6 pixels, each block periodically iterating every 12 rows on the axis of the lens, and, if the sub-pixels of the panel are vertical, M×N viewpoints are distributed below each lens (30) from M / P viewpoints (6 / M sub-pixels per viewpoint), and the M×N viewpoints are distributed in P×N vertical blocks with a height of 6 / P sub-pixels, each block periodically iterating every 12 / P rows on the axis of the lens.

8. The method according to claim 7, wherein: For a panel formed by horizontal sub-pixels, each of the M×N viewpoints of the image is encoded with 6 / M different vertical pixels within N basic blocks of 6 pixels in height, where M is equal to 1, 2, 3 or 6.

9. The method according to claim 7, wherein: For a panel formed by vertical sub-pixels, each of the M×N viewpoints of the image is encoded with 6 / M different vertical sub-pixels within P×N basic blocks of height 6 / P sub-pixels, where M is equal to 1, 2, 3 or 6.

Images