SYSTEM AND METHOD FOR DISPLAYING AN AUTOSTEREOSCOPE 2-PERSPECTIVE IMAGE ON AN AUTOSTEREOSCOPE N-PERSPECTIVE DISPLAY SCREEN AND METHOD FOR MONITORING THE DISPLAY ON SUCH A DISPLAY SCREEN

Description

1. Technical field of the invention

[0001] The invention relates to a system and method for autostereoscopic display. 2. Technological background

[0002] Autostereoscopy is a technique that allows the display of three-dimensional images without requiring the observer to wear special glasses. This technique is known in itself, particularly from patent documents WO2006 / 024764, WO2014 / 041504, WO2013 / 140363, and WO2014 / 016768 in the name of the applicant.

[0003] In general, an autostereoscopic image consists of a plurality of nested elementary image strips, each corresponding to views of the same object or scene from different viewpoints. A selector device, typically consisting of a cylindrical lenticular array or a parallax barrier, is positioned in front of the display screen to allow the projection of a pair of elementary images corresponding to two different viewpoints of the scene, respectively, towards the two eyes of the observer, thus creating an impression of depth in the observer's brain.

[0004] One of the drawbacks of known autostereoscopic techniques lies in the large number of viewpoints that must be simultaneously displayed on the screen. Indeed, unlike glasses-based technologies where only two viewpoints are needed, autostereoscopy requires a large number of viewpoints to free the observer from positioning constraints.

[0005] In the context of autostereoscopy, it can be considered that the screen is wearing the glasses. It is known that a good compromise combining good screen resolution, image quality, and freedom of movement for the observer in front of the screen involves using eight different viewpoints.

[0006] The applicant has already proposed an autostereoscopic screen 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 an array of identical cylindrical lenticules, each with a focal length configured to reflect light rays from the screen to infinity. The width of the lenses is approximately equal to the width of 8 sub-pixels (since the applicant advocates using only one sub-pixel per viewpoint per lens and per horizontal line instead of one pixel). However, other configurations are possible without significantly altering the effect.The pitch of the cylindrical lenticule array is precisely calculated so that the observer sees, at a predetermined distance from the screen (called the flat tint distance, which is for example fixed at 85cm), the images succeeding one another every 6.5cm (average distance between the eyes, chosen as the basis for calculation), thanks to the magnifying effect of the lenticule array.

[0007] This magnifying effect results from the fact that a lens placed at the correct distance (its focal length) magnifies the sub-pixel that is aligned with its optical axis and the pupil of the observer's eye. If the lens magnifies 8 times, the sub-pixel seen through the lens is perceived as 8 times larger than it actually is, and obscures, for the eye receiving light through that lens, the 7 other sub-pixels that are not aligned as described above.

[0008] Thus, the observer perceives only one-eighth of the screen's resolution with each eye. These points, combined to form an image or viewpoint, are all magnified 8 times horizontally. Seven-eighths of the resolution remains to display seven other viewpoints using the same method.

[0009] The space in front of the screen is then divided into 52cm zones, called lobes, in which the 8 viewpoints can be perceived successively every 6.5cm. When leaving a lobe, the succession of the 8 identical viewpoints is repeated.

[0010] Thus, an observer who moves parallel to the plane of the screen at 85cm from it (in the case where the flat tint distance is fixed at 85 cm), from the far right position to the far left position sees with one eye the viewpoints succeeding each other every 6.5cm in a series of 8 forming one lobe and again the same 8 forming the second lobe and so on.

[0011] With the observer's two eyes spaced 6.5 cm apart on average, when the right eye sees a viewpoint n (n (being between 1 and 8 in the case of an 8-viewpoint screen), the left eye sees the complementary viewpoint in n-1 Or n + 1 and he perceives a three-dimensional image. When he moves to the left or to the right, his eyes change position and therefore his points of view simultaneously, and the sensation remains as if he were in front of a real scene.

[0012] The constraint on the observer's movement is explained by the fact that if the observer has one eye in one lobe and the other eye in a neighboring lobe, not only are the images perceived by the two eyes not consecutive, as is the case when both eyes perceive images from the same lobe (image pairs 1-2; 2-3; 3-4; 4-5; 5-6; 6-7; 7-8), but also the image perceived by the right eye is from a viewpoint further to the left than that perceived by the left eye, or vice versa. In other words, upon moving from one lobe to the other, it is the image pair 1-8 that is perceived. The difference between these two viewpoints is then far too great for the observer to reconstruct the depth perception in their brain. The resulting sensation is so uncomfortable that it is impossible to remain in this position without an immediate and severe headache. The relief is totally incoherent and reversed, this is what is called pseudo-scopy.It is then necessary to shift slightly to the right or to the left to find a pair of successive views.

[0013] The inventors therefore sought a solution that would completely free a privileged observer from movement constraints, both in their movements parallel to the screen and in their movements perpendicular to the screen.

[0014] Documents US 2006 / 170764 A1, US 2017 / 127050 A1, EP 2 797 328 A1, US 2011 / 216171 A1 represent prior art. 3. Objectives of the invention

[0015] The invention aims to provide a system and method for autostereoscopic display that eliminates the occasional appearance of pseudoscopic images.

[0016] The invention also aims to provide, in at least one embodiment, a system and method for autostereoscopic display that avoids the appearance of ghost images.

[0017] The invention also aims to provide, in at least one embodiment of the invention, an autostereoscopic display system and method which completely frees the movements of the observer, both parallel to the screen and perpendicular to the screen, while maintaining the quality of the image perceived by the observer.

[0018] The invention also aims to provide a method for controlling the display on an autostereoscopic display screen equipped with a device for detecting the position of the observer. 4. Description of the invention

[0019] To this end, the invention relates to a method for displaying an autostereoscopic image from 2 viewpoints, called image viewpoints, each formed of a plurality of sub-pixel bands, called image bands, on an autostereoscopic display screen configured to display N viewpoints, called screen viewpoints, N being greater than or equal to 3, and comprising a matrix of pixels arranged by rows and columns, each pixel being composed of a plurality of sub-pixels of different colors, said display screen comprising an optical component, of the type of cylindrical lenticle array or parallax barrier, configured to allow the projection of two predetermined screen viewpoints respectively to the right eye and the left eye of a privileged observer located at a nominal distance from the screen, called the flat tint distance.

[0020] The method according to the invention further comprises the following steps: a detection step, at each instant, of the position of said observer relative to a predetermined point on the display screen, allowing the X, Y, Z coordinates of the observer to be defined with a predetermined precision, in an orthonormal coordinate system of which said predetermined point on the screen is the origin of said coordinate system, said screen defining the X and Y axes of this coordinate system and the direction perpendicular to the screen defining the Z axis of this coordinate system, a determination step, for each eye of the observer at said detected position, of the sub-pixels of the screen forming bands visible to that eye and corresponding to 1 / Nth of the total screen area, a determination step, of the sub-pixels of the screen forming bands of the screen not visible to either eye of the observer at said detected position and corresponding to 1-2 / Nth of the total screen area, a determination step, of the assignment to said sub-pixels of each visible band of the screen,For each eye of the observer at said detected position, subpixels of the corresponding image band for that eye, a step of assigning to said subpixels of the bands not visible on the screen by either of the observer's eyes, subpixels corresponding to viewpoints, called safety viewpoints, constructed from the subpixels of the image bands and intended to adapt the display to the X, Y, Z movements of the observer and to compensate for said detection accuracy of said position of the privileged observer.

[0021] A method according to the invention thus provides for displaying an autostereoscopic image consisting of only two elementary images to form just two viewpoints (respectively for the right and left eyes of the observer) on a screen configured to display at least three viewpoints (hereinafter referred to as a multi-viewpoint screen). In other words, the invention provides for repurposing a multi-viewpoint screen (i.e., configured to display at least three viewpoints) with a known predetermined configuration (in particular, the physical flat tint distance of the screen) and displaying only two viewpoints on it so as to be able to use the pixels initially dedicated to the unused viewpoints of the screen to display safety viewpoints constructed from the two viewpoints of the image.These safety viewpoints eliminate lobe transitions and improve observation quality when the observer moves relative to the screen (both parallel and perpendicular to it). Therefore, using a multi-viewpoint screen to display an image from only two viewpoints masks visible transitions at areas of the screen where pseudoscopy or orthoscopy occur.

[0022] The method also involves detecting the observer's position using a dedicated device. This dedicated device is referred to in the text as an observer position detection device or a tracking device, following the English terminology commonly used in the relevant technical field. Safety considerations allow for compensation of errors in the observer's position detection by this dedicated device (related both to the time lag between the observer's actual position and its detection, and to the accuracy of the detection device itself).

[0023] Given the observer's position relative to the screen, the projected screen viewpoints towards each of the observer's eyes at that detected position can be determined. It is therefore possible to assign these known screen viewpoints the viewpoints of the image to be projected, and to use the screen viewpoints not visible to the privileged observer to display safety viewpoints that compensate for detection errors and mask transitions visible to the observer.

[0024] In particular, the portions of the screen visible depending on the observer's position change not only when the observer moves parallel to the screen, but also when the observer moves perpendicular to the screen. When moving perpendicular to the screen, the portions of the screen that allow viewing the different screen viewpoints appear to increase or decrease in width depending on the direction of movement. Before and after the flat tint distance, the order in which the different screen viewpoints appear is reversed, and at the flat tint, only one viewpoint is perceived, as explained previously.

[0025] The invention therefore makes it possible to take into account all the movements of the observer in order to constantly adapt the display on the screen and eliminate the disadvantages of prior art display methods.

[0026] Safety viewpoints can be formed by copying one of the image viewpoints, that is, the viewpoint dedicated to the right eye or the viewpoint dedicated to the left eye. For example, with a flat tint and in the case of a multi-viewpoint screen configured to display 8 viewpoints, once the two image viewpoints are assigned to two screen viewpoints that depend on the observer's position, the process can ensure that the 6 remaining safety viewpoints display the image viewpoint dedicated to the right eye three times and the image viewpoint dedicated to the left eye three times.

[0027] Advantageously and according to the invention, the method includes a calculation step, for each detected position and for at least one eye of the observer, called the leading eye - in particular for each eye of the observer - of the moiré bands formed at this detected position so as to be able to substitute each sub-pixel contributing to form each moiré band at this position, by the corresponding sub-pixel of the image band dedicated to this eye, so as to form a simulated flat tint, called a false flat tint, at said detected position.

[0028] This advantageous variant is particularly useful for quickly adapting the display to observer movements perpendicular to the screen. Indeed, as soon as the observer deviates from the flat color distance (that is, moves closer to or further from the screen relative to the nominal viewing distance at which a single screen viewpoint is projected towards an eye of the observer), moiré patterns appear on the screen. These moiré patterns result from the superposition of the lenticular lens array and the pixel matrix forming the display screen. To compensate for this phenomenon, the method according to this variant of the invention provides for replacing, for each eye, each moiré band with the image band dedicated to that eye.This makes it possible to create a false flat tint at the detected distance, that is to say to simulate that the display screen and the cylindrical lenticule array are configured to allow the projection of two predetermined views of the simulated screen respectively to the right eye and left eye of a privileged observer located at this detected distance.

[0029] For example, if we consider a screen with 8 viewpoints (in other words, N equals 8), and we display a test pattern consisting of 8 different colors on the screen connected to a lenticular lens array configured so that viewpoints 1-4, 2-5, etc., are projected towards the observer's two eyes at the nominal distance of 85 cm, then as soon as the observer changes their distance from the screen, increasingly numerous and closely spaced bands of color appear on the screen. It is then possible, knowing the observer's position relative to the screen, to replace these color bands with the image bands intended for the observer's eyes, and this for each eye. Thus, not only is the relative transition position in the initial order of viewpoints modified, but a false flat color is also reconstructed to compensate for the observer's Z-axis movement and X and / or Y-axis phase shift.

[0030] When the images to be displayed are pre-captured or calculated and cannot be modified, the observer's position is accounted for by moving the same initial pixels. With a real-time 3D stereoscopic rendering system, it becomes possible, after identifying the observer's movement, to modify the displayed content in real time and create the illusion of being within a three-dimensional volume that incorporates the observer's X, Y, and Z movements for a perfect illusion.

[0031] The invention according to this embodiment therefore includes a step of determining the moiré bands visible to each eye of the observer at each detected position. As previously stated, these moiré bands result from the superposition of the lenticular array and the pixel matrix forming the display screen and therefore depend on the display screen and the cylindrical lenticular array in question. Thus, knowledge of the pitch of the cylindrical lenticular array, the drilling pitch (more commonly known by its English name) pitchThe calculation of the moiré bands visible at each viewing distance depends on several factors, including the display screen, screen size, the flat tint distance of the screen and cylindrical lenticle array, and the angle of the cylindrical lenticles relative to the screen columns. Therefore, when the observer moves along the X, Y, and Z axes, only the moiré band calculation is needed to compensate for this movement, and this calculation applies across the entire height and width of the image.

[0032] Furthermore, the method applies equally to all screens of the same size, resolution, number of viewpoints, and flat tint. In other words, it is sufficient to determine the moiré equation associated with a specific screen to implement the method according to the invention with all screens, regardless of the screen's serial number within its category. No ray tracing calculations are required with a method according to the invention, which significantly simplifies the implementation steps.

[0033] Advantageously and according to the invention, the method includes, for each position, a step of calculating a phase shift between the two eyes of the observer so as to be able to determine said moiré bands visible by the complementary eye to the leading eye, from the calculation of the moiré bands of the leading eye and the calculated phase shift.

[0034] This advantageous variant allows us to account for the phase difference between the observer's two eyes. Specifically, if we consider one of the observer's eyes (for example, their right eye) moving along a Z-axis while remaining centered on the midpoint of a viewpoint on the screen, then their left eye, depending on the observer's distance from the screen, will successively pass through other viewpoints on the screen, in both phase and opposite-phase positions. The phase difference between the observer's two eyes depends on the observer's distance from the screen.

[0035] According to another variant, each of the observer's eyes is tracked by the tracking device (or by a device dedicated to each eye) so that the moiré bands are determined from the detected position of each of the two eyes.

[0036] According to one variant of the invention, at least one safety viewpoint is formed from a mixture of image bands which is constituted by adding to the first image viewpoint a negative percentage of the second image viewpoint.

[0037] In other words, at least one security viewpoint is formed by subtracting from the first image viewpoint a predetermined negative percentage of the other image viewpoint.

[0038] This allows for the cancellation of ghosting. Specifically, taking advantage of the existence of safety viewpoints—in principle invisible to the user, but nevertheless projected into their field of vision in practice, given that position detectors cannot provide instantaneous and completely precise position measurements—the method, according to this variant, modifies the image displayed at these viewpoints by subtracting a certain percentage (in negative) from the other viewpoint. This cancels out the ghosting image, which is always a certain percentage (positive) of the juxtaposed viewpoint. The two images—the ghosting due to imperfections in the physical system (primarily refractive optics and residual diffusion) and the adjusted anti-ghosting image (computer calculation, after calibration, of the negative image with an equivalent intensity to be subtracted from the main image)—thus cancel each other out.

[0039] It is important to understand that the management of safety viewpoints (composed of a single screen viewpoint at the distance of the flat tint and bands of varying widths everywhere else) depends on the observer's position relative to the screen, since it is the observer's position that defines which screen viewpoints are visible to the observer, and therefore, it is the observer's position that conditions the construction of safety viewpoints according to the objectives pursued (suppression of ghost images, suppression of pseudoscopic images, etc.).

[0040] According to one embodiment of the invention, the security viewpoints can also be used to display image viewpoints dedicated to a second privileged observer whose position relative to the screen is determined by another dedicated tracking device. This embodiment is only possible if N is greater than or equal to 4.

[0041] Advantageously and according to the invention, the display screen is a seven-viewpoint screen configured so that the viewpoints visible to the eyes of the privileged observer at said distance from the flat tint are viewpoints of the screen separated from each other by 2 viewpoints, so that the visible pairs are views 1-4, 2-5, 3-7 for viewpoints numbered successively from 1 to 7.

[0042] This variant allows for two viewpoints to be available between the observer's two eyes at the distance of the flat tint, thus facilitating display management when the observer moves perpendicularly to the screen. Indeed, since the privileged observer sees stereoscopic bases that vary according to their Z-axis movements, it is essential to have at least two viewpoints available at the distance of the flat tint to offer the privileged observer a wide range of movement without compromising image quality. Of course, other configurations are also possible without significantly altering the effect provided by the invention.

[0043] The invention also relates to a method for controlling the display on an autostereoscopic display screen (E) with N viewpoints, N being greater than or equal to 3, associated with a device for detecting the position of a privileged observer in front of this screen, comprising: a step of detecting the position of the observer by said detection device allowing to define a presence of the observer in a presence zone among at least three predetermined presence zones delimiting the space in front of the screen, a step of displaying an autostereoscopic image at 2 viewpoints by implementing the display method according to the invention, if said observer is in a first predetermined presence zone, a step of displaying an autostereoscopic image at N viewpoints if said observer is in a second predetermined presence zone, a step of displaying a two-dimensional image formed by displaying a single image on each of the viewpoints of the screen, if said observer is in a third predetermined presence zone.

[0044] A control method according to the invention makes it possible to adapt the display mode (i.e., the content displayed and the manner in which this content is displayed) on an autostereoscopic display screen according to the presence zone in which the privileged observer is located. The method according to the invention provides at least three distinct display modes for at least three predetermined presence zones (a first presence zone, a second presence zone, and a third presence zone). In particular, the control method makes it possible to locate the predetermined observer in one of at least three predetermined zones (which are advantageously defined by distance intervals from the screen) and to define a display type (both with regard to the content displayed and the manner in which this content is displayed on the screen) for each zone.A control method according to the invention therefore makes it possible to adapt the display to the movements of the observer by providing a way of displaying the content of the image for each area, including in the event of loss of tracking or movements of the observer very far from or very close to the screen compared to its nominal operating distance.

[0045] For example, if the observer moves away from the screen such that the tracking device is no longer operational (advantageously, the second presence zone corresponds to a zone for which the detection device cannot provide a position for the observer), the method commands an autostereoscopic display with N viewpoints. In other words, if the detection device cannot detect the observer's presence in the first or third presence zone, then the observer is in the predetermined second presence zone. In this case, the method commands the display of an autostereoscopic image with N viewpoints.

[0046] If the presence of the observer is detected in said first zone (i.e. at a good distance from the screen to use the tracking device and detect the position of the two eyes of the observer), the method implements the method of displaying an autostereoscopic image at 2 viewpoints on the autostereoscopic display screen at N viewpoints according to the invention.

[0047] If the observer is not present in either of the two previously mentioned zones (forming the third zone), a safety display is activated, consisting of a flat, two-dimensional image. This image is formed by displaying the same image to both eyes at each of the N viewpoints of the screen.

[0048] It should be noted that in the event that the presence of the observer is detected in this third zone, the tracking device can detect the observer's head, in which case the process can adapt the display of the flat image according to the movements of the observer, which allows an interaction between the observer and the image, including in this third zone.

[0049] If the observer is in the first zone where the tracking device is fully operational, the process displays a 2-viewpoint image on the N-viewpoint screen. The image display and content can be modified based on the observer's movements. If the content is real-time 3D, the displayed content can be modified based on the observer's movements. If the content is pre-recorded, the display of that content can be modified based on the observer's movements.

[0050] Furthermore, the invention allows the use of a multi-viewpoint screen without a tracking device, initially designed to have a flat tint at an observation distance D, under standard observation conditions for viewpoint pairs 1-2; 2-3; 3-4; 4-5 etc., and the use of this same screen with a tracking device at a simulated flat tint distance corresponding, for example, to D / 3, under observation conditions for viewpoint pairs 1-4; 2-5; 3-6 etc. Under these conditions, the same screen can serve both purposes: when one is far enough from the screen, it displays multi-viewpoint images, and when one is closer, at a tracking-compatible viewing distance (i.e., if the detected distance falls within a predetermined range of observation distances), and for only one person at a time, it creates a simulated flat tint at D / 3, and the entire system remains coherent and controlled for this new use.

[0051] In other words, the display control method according to the invention allows switching from an N-viewpoint display on a screen configured to display N views to a two-viewpoint display on the same screen configured to display N views as soon as the observer enters or leaves, through their movements relative to the screen, a predetermined area enabling their detection by the detection device. Finally, the method allows the display of a 2D image when no other display is possible.

[0052] The invention also relates to a system for displaying an autostereoscopic image from 2 viewpoints, called image viewpoints, each formed of a plurality of sub-pixel bands, called image bands, on an autostereoscopic display screen configured to display N viewpoints, called screen viewpoints, N being greater than or equal to 3, said system comprising: an autostereoscopic display screen with 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, said display screen being surmounted by an array of cylindrical lenticles configured to allow the projection of two predetermined viewpoints of the screen respectively to the right eye and the left eye of a privileged observer located at a nominal distance from the screen, called the flat tint distance, a detection device, at each instant, of the position of said privileged observer with respect to a predetermined point of said display screen, allowing the X, Y, Z coordinates of the observer to be defined with a predetermined precision, in an orthonormal frame of which said predetermined point of the screen is the origin of said frame, said screen defining the X and Y axes of this frame and the direction perpendicular to the screen defining the Z axis of this frame,a determination module, for each eye of the observer, at said detected position, of the subpixels of the screen forming bands visible by that eye and forming 1 / Nth of the total screen, a determination module, for each eye of the observer, at said detected position, of the subpixels of the screen forming bands of the screen not visible by either of the two eyes of the observer and corresponding to 1-2 / Nth of the total screen, a module for assigning to said subpixels of each visible band of the screen, for each eye of the observer at said detected position, the subpixels of the corresponding band of each of the two image viewpoints, a module for assigning, to said subpixels of the non-visible bands of the screen, subpixels corresponding to viewpoints, called safety viewpoints, constructed from the subpixels of the image viewpoints and intended to compensate for said detection accuracy of said position of the privileged observer. ,

[0053] A system according to the invention advantageously implements a method according to the invention, and a method according to the invention is advantageously implemented by a device according to the invention. Therefore, the advantages and effects of the method according to the invention also apply to the display system according to the invention.

[0054] The device for detecting the position of the privileged observer can be of any type. For example, it could be a device comprising a radio frequency signal transmitter worn by the observer, working in conjunction with localization receivers attached to the screen. It could also be a sensor for detecting and recognizing the observer's eye movements, mounted on or near the display screen. Generally speaking, it can be any type of device configured to detect the position of the observer's eyes or head relative to the display screen.

[0055] The system according to the invention also includes a module for determining the bands of the screen visible to the observer and a module for determining the bands not visible to the observer. It further includes a module for assigning subpixels of the corresponding image band to the subpixels of the visible bands of the screen, and a module for assigning to said subpixels of the non-visible bands of the screen subpixels corresponding to safety points intended to compensate for the low accuracy of detecting the position of the privileged observer.

[0056] Throughout the text, the term module means a software component, a subset of a software program, which can be compiled separately, either for independent use or to be assembled with other modules of a program, or a hardware component, or a combination of a hardware component and a software subprogram.

[0057] Preferably, the modules of the system according to the invention are subsets of a software program intended to be executed by a computer or by a microprocessor connected to the display screen.

[0058] The applications of a system and a method according to the invention are numerous and varied.

[0059] In the automotive field, gauges but also head-up displays, better known by the English acronym HUD for Head-Up DisplayThey can implement the invention, which allows for the display, directly in the driving environment, in a credible and realistic manner, of information useful for driving and safety. The large depth of field and precise collimation control make it possible to superimpose, for a privileged observer—the driver, whose eye position is known at every moment—information in the correct location in front of the vehicle, on the road and slightly above, at the true distances and from the correct perspective.

[0060] In the medical field, the invention provides an image from a stereoscopic endoscope. By using only two viewpoints obtained from the stereoscopic endoscope, it avoids the need for lengthy and costly system certification procedures that would be required to create additional viewpoints derived from the two initial views. The primary observer, the surgeon, can benefit from a glasses-free, high-quality stereoscopic image.

[0061] Other applications are conceivable, such as individual projection screens on airplanes. The ability to be very close to the screen also allows for the integration of a touchscreen or a contactless interaction system for rapid interaction with the content. This latter feature could even serve as the primary tracking system.

[0062] In the field of video games, the invention makes it possible to show 3D in real time and applies equally to arcade games, slot machines, etc.

[0063] It is also possible to design touch tablets or mobile phones, for individual use by nature, to display all types of video content.

[0064] The invention also relates to a display method and a display system characterized in combination by all or part of the characteristics mentioned above or below. 5. List of figures

[0065] Other objects, features and advantages of the invention will become apparent from the following description, given by way of non-limiting example only, and which refers to the accompanying figures in which: there figure 1is a schematic view illustrating the path of light rays emitted by an autostereoscopic display screen towards an observer's eye, the figure 2 is a schematic view illustrating the change in perception for an observer's eye at the flat tint distance, the figure 3 is a schematic view illustrating the change in perception for an observer's eye as it approaches the display screen, the figure 4 is a schematic view of a display method according to one embodiment of the invention, the figure 5 is a schematic view of the steps involved in assigning a method according to an embodiment of the invention, the figure 6 is a schematic view illustrating the phase shift between the eyes of a privileged observer during a movement towards the display screen, the figure 7 is a schematic view illustrating the principle of calculating moiré bands implemented by the process according to the invention, the figure 8 is a schematic view of an autostereoscopic image display system according to an embodiment of the invention. 6. Detailed description of an embodiment of the invention

[0066] In the figures, scale and proportion are not strictly to scale for illustrative and clarity purposes. In all figures, the information contained in a screen viewpoint is represented by hatching. Different hatching patterns illustrate different screen viewpoints.

[0067] There figure 1 This schematically illustrates the path of light rays from a pixel array D of an autostereoscopic display screen E to the pupil 10 of an observer. These light rays are deflected by an optical device, such as a lenticular grating RL, to converge at the observer's pupil 10. The autostereoscopic display screen E of the figure 1It comprises three screen viewpoints schematically represented by three different types of hatching (horizontal hatching, slanted hatching, and vertical hatching). Each lens 12 of the lenticular array covers a set of pixels from the three viewpoints. On the figure 1 The observer is positioned at the flat tint distance of the screen, and the viewpoint with angled hatching is projected towards the observer's eye. The same phenomenon occurs for the observer's other eye, so that at the flat tint distance, each eye of the observer perceives a different viewpoint of the same scene, generating an impression of depth in the observer's brain.

[0068] There figure 2 schematically illustrates what happens to an observer's eye when it moves a predetermined distance, called the stereoscopic baseline and represented by the letter b, in parallel with the autostereoscopic display screen. This stereoscopic basis is an intrinsic characteristic of the display screen that depends on the pixel panel D used, the lenticular array RL used, and in particular its pitch p In this figure, the lenticular array is represented solely by the optical centers 13 of the lenses. When the observer's eye is located at the position shown in the lower left of the figure, it perceives the pixels from the viewpoint with the slanted hatching. In other words, all light rays emanating from this viewpoint with the slanted hatching are projected towards the observer's eye. If the observer moves parallel to the screen of the stereoscopic base b to reach the position shown at the top left of the figure, then his eye now perceives all the pixels from the viewpoint adjacent to the horizontal hatching.

[0069] There figure 3schematically illustrates what happens to an observer's eye when it moves along an axis connecting points with flat tints separated by a distance that is a multiple of the stereoscopic basis b and the center of the screen. When the observer's eye moves perpendicularly to the screen, the screen subpixels forming visible bands for that eye change relative to the screen subpixels visible at the distance of the flat tint. Indeed, the eye perceives not only subpixels from the viewpoint with horizontal hatching, but also subpixels from the viewpoint adjacent to the angled hatching.

[0070] The closer the observer gets to the screen, the more their eye will perceive bands of portions of viewpoints of the screen.

[0071] Thus, the invention provides for the use of an autostereoscopic display screen configured to display N viewpoints and to use it only to display two image viewpoints, intended for a privileged observer.

[0072] This process comprises a plurality of steps organized according to the schematic diagram of the figure 4 and whose operating principle is illustrated on the Figures 5 , 6 And 7 .

[0073] The method includes a first detection step E1, at each instant, of the observer's position relative to a predetermined point on the display screen, for example, the center of the screen. This detection provides the X, Y, Z coordinates of the observer in a predetermined coordinate system linked to the display screen E.

[0074] The process then includes a step E2 for determining, for each eye of the observer at that detected position, the subpixels of the screen that form bands visible to that eye. As discussed later, this step E2 for determining the bands visible to that eye can implement a step E2a for calculating the moiré bands at that position and a step E2b for calculating a phase shift between the two eyes of the observer.

[0075] The process also includes an E3 determination step of the screen subpixels that form bands of the screen not visible to either of the observer's eyes at that detected position.

[0076] Once the visible and invisible bands are determined at the observer's detected position, the process can assign to each subpixel of each visible and invisible band the information necessary to address the image bands dedicated to each eye of the observer and construct safety viewpoints intended to occupy the subpixels of the screen not visible to the observer.

[0077] To do this, the process includes an allocation step E4 to the subpixels of each visible band of the screen, for each eye of the observer at the detected position, of the subpixels of the corresponding image band for that eye and an allocation step E5 to the subpixels of the bands not visible to either of the observer's eyes, of subpixels corresponding to the safety viewpoints, constructed from the subpixels of the image bands and intended to adapt the display to the X, Y, Z movements of the observer and to compensate for the accuracy of detection of the position of the privileged observer.

[0078] At the end of step E5 or as soon as the detection device detects a movement of the observer, the process resumes at step E2 of determining the bands visible at the new detected position.

[0079] The operating principle of the different stages E2, E3, E4 and E5 is schematically represented on the Figures 5 , 6 And 7 .

[0080] In the figures, the Z-axis represents the axis perpendicular to the display screen E, which allows us to measure the distance from the observer to the display screen. The Zp distance represents the flat color distance for the display screen in question, that is, the distance at which two predetermined viewpoints of the screen are projected towards the observer's two eyes. The Zobs distance represents the distance detected by a dedicated tracking device. On the figure 5 The references P1, P2, P3, P4, and P5 represent the screen viewpoints, and the image displayed in the frame to the right of each viewpoint represents the image shown from that viewpoint. Thus, at the flat tint distance Zp, the right eye OD is considered to observe the image v4 represented by hatching inclined at +45°, and the left eye OG observes the image v2 represented by horizontal hatching.

[0081] When the observer moves perpendicular to the screen and reaches a distance Zobs, they leave the normal hue distance and thus perceive a plurality of bands from multiple viewpoints of the screen. In the figure, it is assumed that at the detected distance Zobs, the left eye (OG) should perceive bands v1, v2, v3, and the right eye (OD) bands v2, v3, v4, v5. This is represented in frame A on the figure 5 .

[0082] The process therefore allows us to assign to each visible band the sub-pixels of the corresponding image band for that eye. This is represented in the frame referenced B on the figure 5For example, if we consider the left eye (OG), the band visible with the spaced slanted hatching in image v1 from viewpoint P1 is replaced by the corresponding band in the image dedicated to that eye, that is, image v2 with horizontal hatching. The central band in image v2 with horizontal hatching from viewpoint v2 retains the corresponding band in the image dedicated to that eye. Finally, the band visible in image v3 with hatching slanted at -45° from viewpoint P3 is replaced by the corresponding band in image v2 with horizontal hatching dedicated to that eye. This is what appears in image 3 shown in frame B. The corresponding bands must also be assigned for the right eye (OD). Thus, image 2 shown in frame B includes a left band dedicated to the right eye (OD) and represented by hatching slanted at +45°. The process allows, by repeating this process, to assign to each band visible to each eye, the corresponding bands.

[0083] The images shown in frame C illustrate the image displayed by each screen viewpoint once the assignment steps are complete. The bands of each viewpoint not used by either eye allow for the display of safety information.

[0084] The discrepancy observed in the right eye (OD) compared to the left eye (OG) is explained by the phase difference between the observer's two eyes, which is schematically represented on the figure 6 .

[0085] By observing the figure 6If we consider that the right eye (OD) moves towards the screen (E) while remaining centered on the midpoint of viewpoint P1, we observe that the left eye (OG), depending on the user's distance from the screen (E), successively passes through viewpoints P2, P3, P4, and P5, in phase and opposite phase positions, continuously. We also observe that we move from a configuration in which the observer's two eyes receive information from two contiguous viewpoints (represented by the pair 1-2; 2-3), when we are far from the screen, to configurations corresponding to increasingly larger distances. From pairs 1-2; 2-3; etc., we move to pairs 1-3; 2-4; etc., then to pairs 1-4; 2-5, etc., then to pairs 1-2; 2-6, etc.

[0086] The invention makes it possible to adapt the display to these changes in stereoscopic pairs by systematically placing the transitions where the eyes cannot see them.

[0087] The invention preferably includes a calculation step E2a, which, for each detected position and for at least one eye of the observer (preferably for each of the two eyes), calculates moiré bands formed at that detected position in such a way that each sub-pixel contributing to the formation of each moiré band at that position can be substituted by the corresponding sub-pixel of the image band dedicated to that eye. This makes it possible to form a simulated flat tint at that detected position.

[0088] The calculation of moiré bands makes it possible to find the view numbers displayed on the screen according to the X, Y, Z position of the observer in the coordinate system linked to the center of the screen, for example.

[0089] For example, if we consider the point at the top of the screen, centered on the horizontal. We determine the view number to display on this pixel according to the position (X, Y, Z) in the coordinate system linked to this point.

[0090] To do this, we note ParVueBaseXthe stereoscopic basis of the system in X, that is to say the distance to travel in X to change viewpoint if one is at the flat tint distance; ParVueBaseY the stereoscopic basis of the system in Y, that is to say the distance to travel in Y to change viewpoint if one is at the flat tint distance; Zp, the flat tint distance; and X, Y and Z the coordinates of the observer (his right eye for example) in the frame linked to the screen as defined previously.

[0091] First, the simulated stereoscopic bases at distance Z are calculated using the following formulas: ParVueBaseXrect = ParVueBaseX * Z / Zp ParVueBaseYrect = ParVueBaseY * Z / Zp

[0092] These values ​​correspond to the normalization according to the distance of the position detected in X and Y, in order to find the angles from which the light rays having passed through the lenticular grating arrive at the observer.

[0093] We then find the values ​​of the view observed on the point of the screen by finding the phase shift in the number of views traversed according to the positions in X and Y.

[0094] The overall gap DG the number of views is equal to Dg = DecalColorOri + X / ParVueBaseXrect + Y / ParVueBaseYrect

[0095] In this formula, DecalColorOri translates to a possible misalignment of the optical system with the panel during the screen manufacturing process.

[0096] The set of values ​​of the different pixels can be found relative to this reference pixel by taking into account the X position of the pixel whose value is being sought relative to the lens (with its inclination), which corresponds to the repetition of the phenomena along this same axis; and the Z position of the observer and the X position of the pixel which defines the shift in moiré bands traversed.

[0097] According to the rules of Thales' geometry, schematically illustrated on the figure 7We can then find the width of the moiré bands according to the observation distance Z. Indeed, this width corresponds to the number of pixels NbPixDec (more or less than the original number of pixels) covered by the lenticular network from the distance of the observer.

[0098] If we note NbPix The number of pixels on half a screen is then given by the following formula: NbPixDec = NbPix * Z + Focal / Z * Zp / Zp + Focal − 1

[0099] To find the phase shift in number of views relative to the central lens, we then add to the previously found phase shift the signed value of the phase shift, knowing that NbPixDec can be negative if the network covers fewer pixels (if we are further away than the flat tint).

[0100] There figure 8 is a schematic view of a display system according to the invention comprising an autostereoscopic display screen E comprising a slab Dof pixels and a cylindrical lenticule array RL. The display system also includes a device 15 for detecting the position of a privileged observer 22 relative to a predetermined point on the display screen.

[0101] The system also includes a module 16 for determining, for each eye of the observer, at said detected position, the subpixels of the screen forming bands visible to that eye and forming 1 / Nth of the total screen, a module 17 for determining, for each eye of the observer, at said detected position, the subpixels of the screen forming bands of the screen not visible to either of the two eyes of the observer and corresponding to 1-2 / Nth of the total screen, a module 18 for assigning to said subpixels of each visible band of the screen, for each eye of the observer at said detected position, the subpixels of the corresponding band from each of the two image viewpoints, and a module 19 for assigning, to said subpixels of the non-visible bands of the screen, subpixels corresponding to viewpoints, called safety viewpoints,constructed from the sub-pixels of the image viewpoints and intended to compensate for said detection accuracy of said position of the privileged observer.

[0102] Modules 16, 17, 18, and 19 of the system are, for example, subsets of a software program intended to be executed by a computer or microprocessor connected to the display screen. This configuration is schematically represented on the figure 8 The software implemented by this computer also preferably implements the calculation module, preferably for each eye of the observer 22 for each detected position of the moiré bands formed at that detected position so as to be able to substitute each sub-pixel contributing to form each moiré band at that position, by the corresponding sub-pixel of the image band dedicated to each eye, so as to form a false flat tint at the detected position.

[0103] The invention can be implemented on all types of autostereoscopic display screens whose parameters are known and which allow in particular the calculation of moiré bands visible at each position detected by the tracking device.

Claims

1. A method of displaying an autostereoscopic image of two viewpoints, called image viewpoints, each formed of a plurality of strips of subpixels, called image strips, on an autostereoscopic display screen (E) configured to display N viewpoints, called screen viewpoints, N being greater than or equal to 3, and comprising a matrix (D) of pixels arranged by lines and columns, each pixel being composed of a plurality of subpixels of different colours, the said display screen (E) comprising an optical component, such as an array (RL) of cylindrical lenticules or parallax barrier, configured to allow the projection of two predetermined screen viewpoints respectively to the right eye (OD) and to the left eye (OG) of a privileged observer (22) located at a nominal distance from the screen, called solid colour distance (Zp), the said method comprising the following steps: - a step (E1) of detection, at each instant, of the position of the said observer relative to a predetermined point of the display screen (E), enabling the coordinates X, Y, Z of the observer to be defined with a predetermined accuracy, in an orthonormal reference system in which the said predetermined point of the screen is the origin of the said reference system, the said screen defining the X and Y axes of the said reference system and the direction perpendicular to the screen defining the Z axis of the said reference system, - a step (E2) of determining, for each eye of the observer at the said detected position, subpixels of the screen forming strips visible by that eye and corresponding to 1 / Nth of the total screen, - a step (E3) of determining subpixels of the screen forming strips of the screen that are not visible by either of the observer's two eyes at the said detected position and corresponding to (N-2) / Nth of the total screen, - a step (E4) of assigning to the said subpixels of each visible strip of the screen, for each eye of the observer at the said detected position, subpixels of the respective image strip for that eye, - a step (E5) of assigning to the said subpixels of the strips of the screen not visible by either of the two eyes of the observer, subpixels corresponding to viewpoints, called safety viewpoints, constructed from subpixels of image strips and intended to adapt the display to the X, Y, Z movements of the observer and to compensate for the said accuracy of detection of the said position of the privileged observer.

2. The method according to claim 1, characterized in that it comprises a step (E2a) of calculating, for each detected position and for at least one eye of the observer, called director eye, moiré strips formed at that detected position, so as to be able to substitute each subpixel contributing to form each moiré strip at that position, for the corresponding subpixel of the image strip dedicated to that eye, in such a way as to form a simulated solid colour, called false solid colour, at the said detected position.

3. The method according to claim 2, characterized in that it comprises, for each position, a step (E2b) of calculating a phase shift between the two eyes of the observer so as to be able to determine the said moiré strips visible by the complementary eye of the director eye, from the calculation of the moiré strips of the director eye and of the calculated phase shift.

4. The method according to any one of claims 1 to 3, characterized in that at least one safety viewpoint is constructed from a mixture of image strips that is formed by the addition to the first viewpoint of the image of a negative percentage of the viewpoint of the complementary image.

5. The method according to any one of claims 1 to 4, characterized in that the said display screen (E) is a seven-viewpoint screen configured so that the viewpoints visible by the eyes of the privileged observer (22) at the said solid colour distance (Zp) are screen viewpoints separated from each other by two viewpoints, so that the visible pairs are the views 1-4, 2-5, 3-7 for viewpoints numbered successively from 1 to 7.

6. A method of controlling the display on an autostereoscopic display screen (E) of N viewpoints, N being greater than or equal to 3, associated with a device for detecting the position of a privileged observer in front of the said screen, comprising: - a step of detecting the position of the observer by the said detection device allowing a presence of the observer to be defined in one presence zone from among at least three predetermined presence zones delimiting the space in front of the screen, - a step of displaying an autostereoscopic image of two viewpoints by the implementation of the display method according to any one of claims 1 to 5, if the said observer is in the said first predetermined presence zone, - a step of displaying an autostereoscopic image of N viewpoints if the said observer is in the said second predetermined presence zone, - a step of displaying a two-dimensional image formed by the display of a single image on each of the N screen viewpoints, if the observer is in a third predetermined presence zone.

7. The method according to claim 6, characterized in that the said predetermined presence zones are defined by distance intervals relative to the screen.

8. The method according to either one of claims 6 or 7, characterized in that the said second presence zone corresponds to a zone for which the said detection device does not enable a position of the observer to be supplied.

9. A display system of an autostereoscopic image of two viewpoints, called image viewpoints, each formed of a plurality of strips of subpixels, called image strips, on an autostereoscopic display screen (E) configured to display N viewpoints, called screen viewpoints, N being greater than or equal to 3, the said system comprising: - an autostereoscopic display screen (E) of N viewpoints comprising a matrix (D) of pixels arranged in lines and columns, each pixel being composed of a plurality of subpixels of different colours, the said display screen comprising an array (RL) of cylindrical lenticules configured to allow the projection of two predetermined screen viewpoints respectively to the right eye (OD) and the left eye (OG) of a privileged observer (22), located at a nominal distance from the screen, called solid colour distance (Zp). - a device (15) adapted for detecting, at each instant, the position of the said privileged observer (22), relative to a predetermined point of the display screen, allowing the coordinates X, Y, Z of the observer to be defined with a predetermined accuracy, in an orthonormal reference system in which the said predetermined point of the screen is the origin of the said reference system, the said screen defining the X and Y axes of the said reference system and the direction perpendicular to the screen defining the Z axis of the said reference system, - a first module (16) adapted for determining, for each eye of the observer at the said detected position, subpixels of the screen forming strips visible by that eye and forming 1 / Nth of the total screen, - a second module (17) adapted for determining, for each eye of the observer at the said detected position, subpixels of the screen forming strips of the screen that are not visible by either of the observer's two eyes and corresponding to (N-2) / Nth of the total screen, - a first module (18) adapted for assigning to the said subpixels of each visible strip of the screen, for each eye of the observer at the said detected position, subpixels of the respective strip of each of the two image viewpoints, - a second module (19) adapted for assigning to the said subpixels of the strips not visible of the screen, subpixels corresponding to viewpoints, called safety viewpoints, constructed from subpixels of image viewpoints and intended to compensate for the said accuracy of detection of the said position of the privileged observer.

10. The system according to claim 9, characterized in that it comprises a first module adapted for calculating, for each detected position and for at least one eye of the observer, called director eye, moiré strips formed at that detected position so as to be able to substitute each subpixel contributing to form each moiré strip at that position, for the corresponding subpixel of the image strip dedicated to that eye, in such a way as to form a false solid colour at the said detected position.

11. The system according to claim 10, characterized in that it comprises a second module adapted for calculating a phase shift between the two eyes of the observer so as to be able to determine the said moiré strips visible by the complementary eye of the director eye, from the calculation of the moiré strips of the director eye and of the calculated phase shift.

12. The system according to any one of claims 9 to 11, characterized in that the said display screen is a seven-viewpoint screen configured so that the viewpoints visible by the eyes of the privileged observer at the said solid colour distance are screen viewpoints separated from each other by two viewpoints, so that the visible pairs are the views 1-4, 2-5, 3-7 for viewpoints numbered successively from 1 to 7.