DE102007043574A1 - Autostereoscopic flat display for e.g. computer tomography, has rear lens scanning disk with horizontal cylinder lenses that are focused on surface of subpixels, where barrier mask is permitted to be used in place of scanning disk - Google Patents

Abstract

The display has a liquid crystal with red, green and blue subpixels (1.4) controlled by a graphics computer, in which perception-oriented highly resolving (HR) coding is installed. A rear lens scanning disk (1.5) is mounted with horizontal cylinder lenses (1.13) adjacent to a front lens scanning disk (1.6) with perpendicular or transverse cylinder lenses before a flat screen with red, green and blue subpixels (1.7), where the horizontal cylinder lenses are focused on a surface of the subpixels (1.7). A barrier mask is permitted to be used in the place of the rear lens scanning disk.

Description

Subject of the invention

The Autostereoscopes Multi User Window 3D technique concerns a multi View 3D presentation and coding facility, in the 3D window can be opened in a 2D display while the environment continues to operate in the usual 2D mode, without to lose resolution. Of course, too the entire screen can be operated in 3D mode, though desired. This is achieved by small structured Lens raster discs, the pitch distances on the order have the desired 2D / 3D pixel size, the z. T. focus directly on the subpixel surface with possibly additional small prism grid or barrier masks. For a Multi View 3D representation, the resolution of the individual perspectives, called views for short, to increase the factor 3, a so-called perception-oriented subpixel HR coding is used, which each subpixel for increased brightness resolution Uses and the correct color components on the environment of a distributed every pixel. By focusing different cylindrical lenses directly on the pixel surface is achieved that the optical transition areas from one perspective to the next be kept as small as possible - in Compared to the angular ranges in front of the display where exactly a perspective is clearly visible. When using a base display with a QWUXGA resolution (3840 × 2400 pixels) through a vertical-to-horizontal conversion of the resolution achieved that with a full HDTV resolution of 1920 × 1200 a total of 12 separate perspectives (views) can be displayed. When using a new technology, such as an OLED Technology in which 5 μm narrow subpixels are realized It is possible to have over 30 perspectives at HDTV resolution for every perspective achieve. This leads to a new quality of Depth representation in a picture: to a stress-free 3D TV presentation.

State of the art

Multi View 3D representations are known [1], [2], [3], [5], [6], and developed systems have been built for several years.

What but still unsatisfactory, are clear depth position perceptions in front of and behind the display level. This is due to low resolution the individual views and too bad separation of the individual Views. Also 2D displays, which are switched to a 3D mode, are known in different versions [10], and for example presented by PHILIPS and SeaReal at conferences and fairs Service. An elaborate version is that in front of the lenticular screen incorporated a refractive index switchable liquid crystal became [PHILIPS].

Also autostereoscopic illustrations, with only two perspectives, with the right and the left work are from numerous publications [4], [7], [9], [10]. Stereoscopes Film and projection techniques have been in use for years. Mostly polarized light (horizontal / vertical, or right / left circular) used to the right and left image to separate. With the advancement of LCD technology, it has become possible the translucency of crystals electronically too Taxes. This made the development of the shutter technology possible in synchronism with the field frequency alternately the right one and the left lens becomes opaque and synchronous right and left images appear sequentially on the screen. This method also use autostereoscopic shutter monitors [10].

since A few years ago, autostereoscopic TFT displays are already in use [1], [2], [8], [9], [10], the two views, the right and left Picture, display horizontally multiplexed on the display and the spatially separated radiation directions through barrier masks or cylindrical lens grids produce. Furthermore Such monitors have also been constructed positionally adaptive Use of head trackers [4], [7], [9], [9], [10], [11], [12], [13].

applications

In Many applications already have 3-dimensional data before, can be generated in standard 3D graphics cards in real time are or will be recorded in stereo video cameras. In medical technology is it computed tomography or stereo endoscopy, in architecture Building and landscape animations, in computer graphic It is virtual reality staging, in the home computer area For example, they are 3D games.

Problem:

Of the Desire to define 3-dimensional windows on serial monitors and It is available on all 3D designers. The previous approaches worsen the 2D presentation too strong and are not priced in series displays used. Therefore, a flexible solution is sought, in the case of an installable software the switching on of 3D windows allows, if prepared for it Monitor is present or can be operated as a secondary monitor. It should also be possible multi-view representations which do not require a fixed position of the viewer and also several Bring viewers a 3D visual profit at the same time.

Principle of the invention:

In order to enable a standard 2D representation, and yet be able to separate several perspectives at the same time, vertical, or rather obliquely different from the vertical direction cylindrical lenses ( 2.6 ) 5.2 ) as well as horizontal cylindrical lenses ( 2.5 ) 5.1 ), with the largest possible aperture so that they focus exactly on the surface of the pixels and thus hit very specific subpixel rows from certain viewing directions. In this case, the color sequences of the otherwise adjacent subpixels appear alternately in a vertical or slightly oblique direction. However, the resolution in the vertical direction must not be reduced. This is achieved by the perception-optimized brightness-color subpixel coding already known from [2] and [3]. (Called HRC). In order to keep the transition areas from one perspective to the neighboring small, or to separate the perspectives clearly, ie not to mix, additionally grid ( 1.5 ) with horizontal cylindrical lenses - with a cylindrical lens for each pixel row. This can possibly be replaced by a barrier mask ( 6.4 ) - directly on the pixel surface. In order to avoid color moire, the visible position of the subpixels is shifted line by line differently to the right, constant, or to the left by a subpixel position optically shifted (offset) by prism grid ( 4.4 ) 4.5 ) 4.6 ). In slightly oblique front cylindrical lenses this abwechselden superimposed color sequences takes over the slope of the cylindrical lens, see. 5 ,

For a very high resolution basic display (eg 3840x2400), one can group a 2D / 3D pixel in an approximately quadratic area of 4 pixels, ie 12 subpixels and code them so that the 4 pixels receive the same information of a 2D pixel or divided into 12 perspectives with 12 perspectives. As 5 and 6 show that a vertical-to-horizontal conversion of the resolution can be achieved such that the odd-numbered perspectives, for example, are copied into odd lines and straight into. The logical separation is then achieved by means of the inclination of the front cylindrical lenses ( 1.6 ) and the additional barrier mask ( 6.4 ) or rear screen ( 5.1 ).

Detailed description different embodiments

2 View Window Technology: The Muli-User 2D / 3D Window Technology Auto-stereoscopes are described below with the help of 7 drawings (2, 3, 4, 12 and a 32 View Version). 1 to 7 ) described in detail. 1 shows a display arrangement with the beam paths to the eye positions ( 1.8 . 1.9 ) with two perspectives. Image part A is a horizontal section - in each case by the pixel line i, i + 1 and i + 2, while B represents a section with the beam path in a vertical plane. The basic TFT display includes: backlight ( 1.1 ), Active matrix for controlling the color subpixels, rear and, to the vertical, front polarizing filter ( 1.3 ) and the liquid crystal layer of the analogically driven subpixels ( 1.4 ). In front of it, there are μm-accurate: the rear lenticular screen ( 1.5 ) with horizontal cylindrical lenses and the front lens grid disc ( 1.6 ) with vertical cylindrical lenses. The horizontal cylindrical lenses have a pitch, ie a distance from the beginning of a cylindrical lens to the beginning of the next, which corresponds to a pixel height, while the front vertical cylindrical lenses have a pitch which corresponds to the width of two subpixels. Both cylindrical lenses ( 1.11 . 1.13 ) focus on the surface of the pixels. On the back of the front lenticular screen ( 1.6 ) prism grid are mounted, the height of which corresponds to about one pixel and the width of about two subpixels, that is, the width and position of a vertical cylindrical lens have. In adjacent pixel lines i, i + 1 and i + 2, three different prism structures occur: in line i are prisms (to the left) 1.12 ), which cause an optical left shift by one sub-pixel. In line i + 1 there are no or evenly deep rectangular recesses, and in line i + 2 are the prisms opened to the right, which cause a right shift by one subpixel. Sizing examples for prism slopes and depths, lens grid disk thicknesses and lens radii are given in tables later for various embodiments.

The beam paths ( 1.10 ) from the left eye ( 1.8 ) or from the right eye ( 1.9 ) show that they lead to the center of respectively adjacent subpixels, which are then driven by HR coded left and right images, respectively, for a 3D representation in the marked window. Since the vertical lenses are narrower than a pixel, even a 2D view is hardly changed. By a precise focusing of the cylindrical lenses on the pixel surface is achieved that a small freedom of movement of the viewer within the window ( 1.16 ) is possible - with correct 3D view. In addition to the central position, there is also a right and left side position with little freedom of movement for correct 3D perception.

4 View window technology, cf. 2 : The same principle as explained for the 2 view representation can be transferred to a 4 view representation: instead of 2 subpixels, 4 subpixels are under a vertical cylindrical lens ( 2.11 ) arranged. The prism grid ( 2.12 ) are also about 4 subpixels.

As 2 shows are now 3 central optimal viewer zones with freedom of movement ( 2.14 ): the middle optimal position and the left optimal head position ( 2.8 ) and the right optimal head position ( 2.9 ). The transition positions, in which two views merge and certain blurs on edges, which are displayed outside the display level, are very small and tolerable, so that also freedom of movement in the distance to the display arise. In addition to the marked central opening angle (aperture) then also two valid lateral areas occur. The transitions from the central aperture area to the lateral areas should be avoided for 3D perception since depth positions are falsified there.

Naturally, the horizontal resolution per View decreases with their number. When using perceptual HR coding, the resolution per view can be preserved in the 3D representation (cf. 3 ), while in 4 views it goes back to ¾. For example, with a WUXGA basic display with 1920 × 1200 pixels, a 3D resolution of 4 × 1440 × 1200 (number of views, horizontal pixels, vertical pixels) is achieved.

3 View window technology, cf. 3 : A natural 2D / 3D technique is achieved on the basis of TFT displays with 3 views, as both cylindrical lenses have the width of a pixel and pixel sizes are thus preserved. Also the regular visible color pixel arrangement (cf. 3B ) shows that the orders R, G, B (red, green, blue) occur in both the horizontal and vertical directions.

The exact arrangement of the grid discs in front of the display is in 4 represented by the three adjacent pixel rows by means of the horizontal sections: line i in section C-D, line 1 + 1 in section E-F and line i + 2 in section G-H. Furthermore, the vertical section I-J illustrates the vertical assignment of the prism positions. Further down, a sizing example is also given in a table.

When 3D capable viewer positions also come closer to the center Distances from the display in question, where the eyes the Views 1 and 3 see, or more distant, where the eyes are the views 1 and 2 or 2 and 3 see. Next to them are also the two side areas available.

12 View window technology, cf. 5 , The large number of 12 views causes a significant reduction in resolution. In order to still achieve a high resolution per view, it is advisable to use a basic display with a QWUXGA resolution (3840 × 2400 pixels). In order to obtain also square 3D pixels, a resolution conversion from the vertical direction to the horizontal direction is performed. The reduction of the resolution in the vertical direction from 2400 to 1200 causes horizontally 2 × 3480 pixels or 12 × 1920 sub-pixels are available. When using the HR coding one achieves thus 12 views, each with a resolution of 1920 × 1200 pixels. That leads to a full 3D HDTV quality. For real-time representations and transmissions this special hardware is required, but this is quite feasible today.

For the vertical / horizontal conversion, it is advisable to make a slight inclination of the front lenses in the ratio 1: 6 (h: v) instead of prism grids. A 2D / 3D pixel then takes the form of a rhombus ( 5.3 ). For the distribution of the views, the odd-numbered views (1, 3, 5, 7, 9, 11) can be written to the odd lines (1, 3, 5, ..., 2399) and the even views on the base display (2, 4, 6, 8, 10, 12) in the even lines (2, 4, 6, ..., 2400). There is a separating black space between each subpixel on the display. By focusing the two cylindrical lenses on one point on the pixel surface, this space can be used for vertical / horizontal conversion. From the odd directions ( 5.11 ) are then only the subpixels of the odd lines (1, 3, 5, 2399) visible and on the intermediate straight lines of sight (2, 4, 6, ..., 12) are then only the subpixels of the even lines (2, 4, 6, ..., 2400) visible.

The 12 subpixels on the adjoining 2D / 3D pixels, of which There are 1920 × 1200, then are HR coded by software Positively interlaced with the pixels of the 12 Views, where the inclination of the front lenses compensated accordingly becomes.

A comparable 12 View quality can be achieved by replacing the rear lenticular screen with the horizontal cylindrical lenses line-by-line staggered barrier strips of the mask ( 6.4 ) are used directly on the pixel surface oriented in the same oblique direction 1: 6, cf. 6 , It covers the part of a subpixel that should not be visible from the even or odd directions. However, one must accept that some brightness is lost through the partial concealment.

The black shown in the magnification of the subpixels Interspaces are not for the human eye visible if the correct viewing distance is maintained. In a first approximation one can reach the resolution limit the eyes with 0.02 ° × 0.02 °. smaller Structures are no longer recognizable. That's about a relationship from 1: 3000 in terms of pixel size to viewing distance. In a 24 inch monitor with 1920 horizontal pixels leads this to a minimum viewing distance of about 80 cm.

32 View window technology, cf. 7 : If you do not want to do without good image resolution, the TFT technology is overwhelmed for a 32 View window technique. But if one uses the possibilities offered by a new OLED technology, one can go from the outset new ways in the design of pixels and choose such structures, as they are ideal for a 3D representation. As shown in test structures, the pixel width can be realized down to 5 μm. Switching speeds in the μs range can also be achieved. This leads to a realistic 3D pixel size of 0.16 mm × 0.16 mm In 7 is such a design for a 2D / 3D pixel ( 7.1 ) for 32 views, how it is suitable for a stress-free 3D presentation. In order to achieve a perceptual image quality, in addition to the colors red (R, 7.2 ), Green (G, 7.4 ) and blue (B, 7.5 ) also white (W, 7.3 ), which also contributes to a better light output. For every perspective ( 7.6 For example, a 5 μm narrow and 160 μm high pixel with the color order R, W, G, W, B, W is available. The image B shows in horizontal section, as the beam path from view 5 ( 7.9 ) to the subpixel column 5 ( 7.10 ) - via the vertical cylindrical lens ( 7.8 ).

Some 100 view test images with this subpixel size and a resolution of 1600 × 1200, instead of OLED displays a film is used on exhibitions and measuring has been shown.

dimensioning examples

• [1] S. Hentschke: Autostereoskope Multi-View Subpixel Render-Methode für Echtzeit-3D. Anm. Nr. 10 2005 036 744.5, (2005), bzw. PCT/DE 2006/001365 .
• [2] S. Hentschke: Autostereoskopisches Wiedergabesystem für 3D Displays (PARSC II), Anm. Nr. WO2004/081 863 , PCT (2004).
• [3] S. Hentschke: Positionsadaptiver, autostereoskoper hochauflösender Subpixel-Coder (PARSC I), Anm. Nr. DE 103 11 389.4 (2003).
• [4] S. Hentschke: Position-adaptive autostereoscopic monitor (PAM) vom 23.10.2001, US 6, 307,585 B1 .
• [5] S. Hentschke: 3D-Display including cylindrical lenses and binary ED Micro fields, vom 03.04.2001, US 6,212.007 B1 .
• [6] 4 Division Patent: Wellenlängenselektives Filter Array für Multi User 3D, Patentanmeldung DE 100 03 326.1 , (1999).
• [7] Großman: Patentanmeldung DE 19827590 A1 vom 23.12.1999.
• [8] Van Berkel: Image Preparation for 3D-LCD, Philips Research Laborstories, UK, SPIE Vol. 3639 (1999), S. 84–91 .
• [9] Schwerdtner. Anordnung zur dreidimensionalen Darstellung von Informationen, Offenlegungsschrift DE 198 22 342 A1 , (1998).
• [10] S. Hentschke: Observer-adaptive autostereoscopic shutter Monitor, US 5,771,121 B1 .
• [11] Jong-man Kim: Liquid crystal display device where in each scanning electrode includes three gate line corresponding separate Pixels for displaying three dimensional image, Samsung Electronics Co., Ltd., Korea, Patent Nr. 5.850.69 (1998).
• [12] Andiel, Hentschke u. a.: Eye-Tracking for Autostereoscopic Displays using Web Cams; SPIE, Bd. 4660 (2002), S. 200–206 .
• [13] Andiel, Hentschke: Position and Velocity Degending Subpixel Correction for Spatial-Multiplexed Autostereoscopic Displays; veröffentlicht in SPIE, Bd. 5006, (2003) .
• [14] S. Hentschke: Stereoskoper Bildschirm. Patentanmeldung P 41 14 023.0 (1991).

In the following two tables (Table 1) and (Table 2) sizing examples are given, which contain the most important parameters for the construction of the 3D window technique, such as lens lenticular, lens pitch and radius, prism pitch and depth, pixel and screen size, viewer angle and optimal viewer distance. # Designation / Parameter 1 2 Bern 1 Surname 12 View TV 3 View TV 2 Screen diagonal 22 '' 40 '' 3 Umbrella size (W × H in mm) 474 × 296 885 × 494 4 Resolution Base TFT Display (W × H) 3840 × 2400 1366 × 768 5 Resolution 2 / 3D Display (V × W × H) 12 × 1920 × 1200 3 × 1369 × 768 HRC 6 Pixel size 2 / 3D (hxv in mm) 0.2468 x 0.2468 0.647 × 0.647 7 Observer angle (horizontal aperture) 28 ° 14 ° Discret. 8th Optimal viewing distance [mm] 1000 1200 9 Lenticular pitch (verti, hor. LRS) 0.2468; 0.1234 0.647; 0.647 10 LSR thickness (v, h lenticular) [mm] 0.49; 0.26 2.78; 1.38 11 Lens radius (v, h cylindrical lenses) 0.25; 0,095 1.39; 0.50 12 Shift prism height - 0.32 13 Lens slope 1: 6 0 14 drawing sheet Fig. 5, (or Fig. 6) Fig. 4 Tab. 1: Dimensioning for the construction of a 3D window technique on available TFT base displays for a 12 View TV with highest HDTV quality and a 3 View TV with 16: 9 HQ. # Designation / Parameter 3 4 Bern 1 Surname 3 View Monitor 4 View Laptop 2 Screen diagonal 24 '' 17 '' 3 Umbrella size (W × H in mm) 517 × 323 366 × 228 4 Resolution Base TFT Display (W × H) 1920 × 1200 1920 × 1200 5 Resolution 2 / 3D Display (V × W × H) 3 × 1920 × 1200 4 × 1440 × 1200 HRC 6 Pixel size 2 / 3D (hxv in mm) 0.258 × 0.258 0.254 × 0.191 7 Observer angle (horizontal aperture) 23 ° 28 ° Discret. 8th Optimal viewing distance [mm] 700 700 9 Lenticular pitch (vertical LRS) 0.269; 0,269 0.254; 0.191 10 LSR thickness (v, h lenticular) [mm] 0.432; 0,575 0.353; 0,407 11 Lens radius (v, h cylindrical lenses) 0.335; 0,258 0.253; 0,147 12 Shift prism height 0.135 0,127 13 Lens slope 0 0 14 drawing sheet 4, Fig. 2 Tab. 2: Dimensioning for the construction of a 3D window technique on available 24 '' TFT monitors with WUXGA resolution for a 3 View 3D monitor and a 4 View 3D laptop.

• [1] S. Hentschke: Autostereoscopes Multi-View Subpixel Rendering method for real-time 3D. Note 10 2005 036 744.5, (2005), or PCT / DE 2006/001365 ,
• [2] S. Hentschke: Autostereoscopic Display System for 3D Displays (PARSC II), Note no. WO2004 / 081,863 , PCT (2004).
• [3] S. Hentschke: Position Adaptive, Autostereoscopic High Resolution Subpixel Coder (PARSC I), note no. DE 103 11 389.4 (2003).
• [4] S. Hentschke: Position-adaptive autostereoscopic monitor (PAM) of 23.10.2001, US Pat. No. 6,307,585 B1 ,
• [5] S. Hentschke: 3D-Display with cylindrical lenses and binary ED Micro fields, dated 03.04.2001, US 6,212,007 B1 ,
• [6] 4 Division Patent: Wavelength-selective filter array for multi-user 3D, patent application DE 100 03 326.1 , (1999).
• [7] Grossman: Patent Application DE 19827590 A1 from 23.12.1999.
• [8th] Van Berkel: Image Preparation for 3D-LCD, Philips Research Laboratories, UK, SPIE Vol. 3639 (1999), pp. 84-91 ,
• [9] Schwerdtner. Arrangement for the three-dimensional representation of information, published patent application DE 198 22 342 A1 , (1998).
• [10] S. Hentschke: Observer Adaptive Autostereoscopic Shutter Monitor, US 5,771,121 B1 ,
• [11] Jong-man Kim: "" In each scanning electrode, there are separate pixels for displaying three dimensional image, "" Samsung Electronics Co., Ltd., Korea. Patent No. 5,850,669 (1998).
• [12] Andiel, Hentschke et al.: Eye Tracking for Autostereoscopic Displays using Web Cams; SPIE, Vol. 4660 (2002), pp. 200-206 ,
• [13] Andiel, Hentschke: Position and Velocity Degending Subpixel Correction for Spatial-Multiplexed Autostereoscopic Displays; published in SPIE, Vol. 5006, (2003) ,
• [14] S. Hentschke: Stereoscopic Screen. Patent Application P 41 14 023.0 (1991).

Captions and signatures

1 : Display arrangement with beam path to the observer position; 3D window technique with two perspectives. A) Section in horizontal plane - through line i, i + 1 and i + 2; B) Section in vertical plane.

1.1

Backlight

1.2

control active matrix

1.3

rear or front polarizer

1.4

liquid crystal and color filters

1.5

Rear Lenticular disk with horizontal cylindrical lenses

1.6

Front Lens raster disc with vertical or oblique cylindrical lenses

1.7

R G, B subpixels

1.8

left eye

1.9

right eye

1.10

horizontal Beam path in horizontal direction from left or right eye on the subpixel surface

1.11

Front vertical cylindrical lenses

1.12

prism grid

1.13

Rear horizontal cylindrical lenses Beam path to the display surface in a vertical plane

1.15

central Movement Space Of Observer Eyes For Optimal 3D view

1.16

central 3D viewing window

1.17

right 3D side viewing window

1.18

left Side 3D viewing window

2 : Display arrangement with beam path to the observer position; 3D window technique with four perspectives. A) Section in horizontal plane with eye positions; B) Section in vertical plane.

2.1

Backlight

2.2

control active matrix

2.3

rear or front polarizer

2.4

liquid crystal and color filters

2.5

Rear Lenticular disk with horizontal cylindrical lenses

2.6

Front Lens raster disc with vertical or oblique cylindrical lenses

2.7

protector

2.8

left optimal head position

2.9

right optimal head position

2.10

beam path from the right viewing position to the surface of the subpixels

2.11

Front vertical cylindrical lenses

2.12

prism grid

2.13

Rear horizontal cylindrical lenses

2.14

horizontal Motion scope of viewer eyes for correct 3D view

2.15

vertical Freedom of movement

3 : Visible subpixel arrangement in front of the cylindrical lenses and prism grid from marked "vertical" line of sight A) Horizontal section through display in pixel line i) B) Arrangement of the subpixels enlarged by cylindrical lenses.

3.1

Backlight

3.2

control active matrix

3.3

rear or front polarizer

3.4

liquid crystal and color filters

3.5

Rear Lenticular disk with horizontal cylindrical lenses

3.6

Front Lenticular disk with vertical cylindrical lenses

3.7

line of sight on the display with grid discs

3.9

visible Pixie line i

3.8

visible Pixeize i + 1

3.10

visible Pixeize i + 2

4 : Arrangement of the lenticular disks in front of the RGB subpixels - for 3 views. The lenticular lenses have the width of a pixel, so that with a standard 2D control the usual image quality is almost maintained. To simplify the illustration, the individual layers are divided and shown in vertical sections.

4.1

RGB Pixel on TFT screen

4.2

horizontal Lens Grid Disc (LRS)

4.3

vertical The lenticular lens sheet

4.4

prism grid for legal shift

4.5

Shift Free recess grid

4.6

prism grid for left shift

4.a

front view the display area

4.b

cut through vertical plane I-J

4.c

cut through horizontal plane C-D

4.d

cut through horizontal plane C-D in pixel line i

4.e

cut by horizontal plane E-F in pixel line i + 1

4.f

cut by horizontal plane G-H in pixel line i + 2

5 : 2D / 3D pixels (3) for a 12 view display in a full HDTV resolution of 1920 × 1200 pixels per view - with a QWUXGA TFT display as a base, a 1: 6 oblique lenticular screen (1), a horizontal lenticular screen (2) and with HR subpixel encoding. 5.1

5.1

Rear Lenticular disk with horizontal cylindrical lenses.

5.2

Front Lenticular disk with slightly inclined from the vertical direction, 1: 6 oblique cylindrical lenses.

5.3

2 / 3D Pixel with up to 12 views

5.4

TFT Basic pixels (RGB)

5.5

Cylindrical lens focal line for view # 1 on the left edge of the 2 / 3D pixel

5.6

Cylindrical lens focal line for view # 12 on the right edge of the 2 / 3D pixel

5.7

Thickness of the front lenticular disc D _vL with focal line on RGP pixel surface

5.8

Thickness of the rear lenticular _disk D _hL with focal line on RGP pixel surface

5.9

Radius of the front cylindrical _lens R _vL

5.10

Radius of the rear cylinder line R _nL

5.11

viewpoints

6 : 2D / 3D pixels for a 12 view display in a full HDTV resolution of 1920 × 1200 pixels per view - with a QWUXGA TFT display as a base, a 1: 6 oblique lenticular screen and a v / h conversion barrier mask at subpixel HR coding.

6.1

2 / 3D Pixels with up to 12 views over a 1: 6 oblique Cylindrical lens and two base pixel rows.

6.2

focal line on TFT pixel surface for View # 1 on the left Edge of the 2 / 3D pixel

6.3

focal line on TFT pixel surface for View # 12 on the right Edge of the 2 / 3D pixel

6.4

Barrier mask structure for vertical-to-horizontal conversion

6.5

barriers mask in horizontal section

6.6

The lenticular lens sheet 1: 6 with focal lines on pixel surface

7 : 2D / 3D pixels for 32 views with 6 × 32 (OLED) subpixels in the colors red, white, green and blue with 4 × 32 separate matrix controls. A) Front view of a 3D OLED pixel. B) Horizontal section through pixel with cylindrical lens.

7.1

3D (OLED) pixels with 32 views

7.2

Red Subpixel row

7.3

white Subpixel row

7.4

green Subpixel row

7.5

Blueness Subpixel row

7.6

numbers Views 1-32

7.7

cut through an emitting subpixel

7.8

vertical cylindrical lens

7.9

line of sight 5

7.10

focus for viewing direction

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - DE 2006/001365 [0025]
• - WO 2004/081863 [0025]
• - DE 10311389 [0025]
• - US 6307585 B1 [0025]
• - US 6212007 B1 [0025]
• - DE 10003326 [0025]
• - DE 19827590 A1 [0025]
• - DE 19822342 A1 [0025]
• US 5771121 B1 [0025]
• - KR 585069 [0025]

Cited non-patent literature

• Van Berkel: Image Preparation for 3D-LCD, Philips Research Laboratories, UK, SPIE Vol. 3639 (1999), pp. 84-91 [0025]
• - Andiel, Hentschke et al.: Eye Tracking for Autostereoscopic Displays using Web Cams; SPIE, Vol. 4660 (2002), pp. 200-206 [0025]
• - Andiel, Hentschke: Position and Velocity Degending Subpixel Correction for Spatial-Multiplexed Autostereoscopic Displays; published in SPIE, Vol. 5006, (2003) [0025]

Claims (10)

Autostereoscope flat panel display with R, G, B subpixels ( 1.4 ), driven by a graphics computer, on which the perception-oriented high-resolution (HR) coding [2] can be installed, characterized in that in addition to a front lenticular screen ( 1.6 ) with vertical or oblique cylindrical lenses in front of a flat display with RGB subpixels ( 1.7 ) another rear lenticular screen ( 1.5 ) is mounted with horizontal cylindrical lenses ( 1.13 ) pointing to the surface of the RGB subpixels ( 1 . 7 ) or that instead of the rear lenticular lens ( 1.5 ) a barrier mask, (e.g. 6.4 ) is available. Autostereoscopic flat display according to claim (1), characterized in that behind the front lenticular screen ( 1.6 ) another, line by line different prism grid ( 1.12 ) located in each of two of three adjacent rows of pixels ( 4.4 ) 4.5 ) the optical beam path ( 1.10 ) from the eye positions ( 1.8 ) 1.9 ) shifts to the sub-pixel surface by one sub-pixel to the right or left and in a row of pixels ( 4.5 ) leaves the beam path unchanged. Autostereoscopic flat display according to claim (1), characterized in that suitable subpixel groups (eg. 5.3 . 6.1 ) are combined as 2D / 3D pixels and can be controlled as 3D pixels in a marked window on the screen. Autostereoscopic flat display according to one of the preceding claims, characterized in that a horizontal / vertical conversion is achieved in that in alternating pixel lines adjacent different perspectives or views from the associated viewing directions ( 5.11 ) are visible separately. Autostereoscopic flat display according to claim (1), in which both lens raster discs focus on the pixel surface, characterized in that the cylindrical lenses ( 2.11 ) of the front lenticular screen ( 5.2 ) in the horizontal direction have a pitch of about 2, 3, 4 or more subpixels and have a skew of 1: 6 or 1: 3 Autostereoscopic flat display according to one of the preceding claims, characterized in that all subpixels in 2D or 3D mode with the perceptually oriented HR subpixel coding. Autostereoscopic flat display according to claim (1) or (2), characterized in that the cylindrical lenses of the front lens grid disc ( 1.6 ) focus not exactly on the surface of the subpixel, but that the beam path ( 1.10 ) goes from the eye position over about ½ to 1 subpixel and that in each case two horizontally adjacent subpixels with the Pixelinformarionen from the same perspectives are controlled color. Autostereoscopic flat display according to claim (7), characterized in that a front lens raster disc ( 1.6 ) has an arbitrary skew and this is taken into account and compensated in the HR subpixel coding. Autostereoscopic flat display according to claim (7) and (8), characterized in that one to the lens pitch with respect to a given optimal viewer spacing belonging to subpixelpitch, the slightly different, per HR subpixel encoding again is corrected. Autostereoscope flat display with a head tracker on the computer connected according to claim (7), (8) or (9), characterized in that the optimal viewer position for a Users using adaptive subpixel encoders, as in simpler earlier ones Systems (see [3]) also tracked in the multi-user 3D window technology becomes.