专利摘要:

公开号:SE0901121A1
申请号:SE0901121
申请日:2009-08-26
公开日:2011-02-27
发明作者:Robert Malmstroem
申请人:Robert Malmstroem;
IPC主号:
专利说明:

type CRT, DLP, LCD, LCOS and LED all have some type of color balance problem. Theseproblems may be that one or more of the three color components are not reproduced linearly fromweakest to strongest light. The color reproduction in these projectors can also be affected by howstrong or weak the total light in a single pixel is. It is thus a prerequisite, thatcreate a perfect anti-aliasing, that you adjust the anti-aliasing mask to the color thateach individual pixel has before applying the mask. Today, there are systems that adaptafter the total brightness of the whole image but these systems do not take into account that in a lightimage there may be smaller areas that are dark and that in a dark image there may be smaller areaswhich are bright.
A common solution for creating large display systems is that each projector is connected to each othercomputer that renders the part of the image that the projector is to project. These display systemsmust also use complementary systems to display moving images. reasonthis is that you have to achieve both so-called frame lock, synchronization betweenthe computers of which image is to be displayed at a given moment, and so-called genlock,synchronization between the projectors of their image sweep, in order to create a cohesive imageappears in all projectors at any given moment. The complementary systems that createframe lock and genlock are costly and require expert knowledge to be installed andto be confirmed.
Summary of the inventionThe present invention relates to a method for creating large high resolution and borderlessprojected images with cost-effective consumer hardware. The method is used indisplay system (1) which may consist of a computer with a graphics card having three video outputs (2)which is connected to three projectors (3a, 3b and 30) which together project onehigh resolution and borderless image (4).
As can be seen in Figure 2, this method is divided into four steps. In the first step (6) rendersthe graphics card the individual large image to be projected by the projectors. Unlike the USBl 6 545 685, which renders an image for each projector and where the overlapping parts ofthe images are rendered fl times, this method only needs to render a large image (10) whichsimplifies the rendering process and significantly reduces the rendering time. In step two (7) is copiedthree overlapping sections to separate rendering buffers. Each section corresponds to that part ofthe large image that one of the projectors is to project. In step three (8), edge smoothing is addedto each section by processing each individual pixel that is overlapped by other sections.
The edge smoothing process is performed with so-called pixel shader which are small programs likeexecuted per pixel on the graph card. The pixel shader program performs the edge smoothing throughto combine a number of color curves based on the color of the individual pixel beforethe edge alignment is added. The color curves are produced manually by projecting test imagesdifferent colors and different brightness in the display system and then adjust a number of color curvesafter each test image. This way of adjusting the edge smoothing to the pixel color in real time improvesconsiderable edge compensation for images that are light, dark or have both light and dark areascompared with previously known methods for edge smoothing. In step four (9), they are copiedkantutj intended the sections to a common rendering buffer.
This invention refers to a display system where all projectors are connected to the samecomputer. The invention solves two problems in this way when moving pictures are to be shown over fl eraprojectors. The first problem is the so-called frame lock. By frame lock is meant that onesynchronizes which image, in a moving image sequence, is currently to be displayed in allproj ector. The invention solves this problem by rendering a large image in the samerendering buffer which is then divided into a section for each projector. In this way is guaranteedthat the same image is displayed by all projectors at any given moment. The second problem is socalled genlock. By genlock is meant that you synchronize all in the display systemimage projectors of input projectors. This is done to ensure that the projectors change from an imageto the next exactly at the same time. The invention creates backlinks between the projectors byuse a graphics card with multiple video outputs to connect all projectorssame Greek card. Since a projector's image sweep is controlled by the graphics card, genlock comesto occur between the projectors because they are connected to the same graphics card. The method solvesthe problem of frame lock and genlock without using any specialized hardware whichminimizes the cost of hardware to display systems that use the method. Display systemusing the method can be created with a standard computer and a simpler graphics card with twoor vide your video outputs. You thus do not need a network to connect the devices inthe system, which minimizes installation and configuration work.
Brief description of the drawingsFigure 1 shows a display system (1) consisting of a computer with a Greek code having three orfl are video outputs (2) connected to three projectors (3a, 3b and 3c) whichtogether project a high-resolution and borderless image (4).
Figure 2 shows a fate diagram (5) where the first step (6) consists of rendering an image into arendering buffer on the graphics card, step two (7) consists of dividing the rendered image intotwo or fl your overlapping sections corresponding to the part of the image that each projectorproject, step three (8) consists of creating one or fl your edge equalizations for each sectionto each adjacent section and step four (9) consists of copying all sections into onecommon rendering buffer for all projectors.
Figure 3 shows the rendered original image (10) which is then divided into overlapping sections(ll, 12 and 13) and then the edge smoothings rendered adjacent to other sections(14, 15 and 16) with pixel shader functions that adapt to each individual pixel colorand finally the resulting common rendering buffer (17) created bythe sections are copied to this buffer.
Figure 4 shows three color intensity curves (18) where the left curve (19) corresponds to the red onethe color intensity, the middle curve (20) corresponds to the green color intensity and the rightthe curve (21) corresponds to the blue color intensity. These color intensity curves are used tocreate an anti-aliasing texture (23) which in turn is used to create anti-aliasing betweenprojector images in the display system. To configure the color intensity curves on oneIn a user-friendly way, a few control points (22) are used that build up the curves with helpof bezier curves.
Detailed description of the inventionThis invention consists of a method (5) for creating large images to be projected by onemultiple projectors in one display system (1). The display system consists of a computer (2) running onesoftware using the method of the invention.
To concretize the method, we take the example that the display system consists of three projectors(3a, 3b and 3c) with an image resolution of 1920 pixels in horizontal direction and 1080 pixels invertical and that we want to use 200 pixels in the overlap between the projectors to createedge alignment. The number of pixels used for edge smoothing between the projectors determineshow sensitive the display system is because the projectors are not perfectly positioned in relation toeach other and the display surface.
What needs to be created on the graphics card is a primary rendering buffer that you use forto render the image to be displayed by the display system. A secondary also needs to be createdrendering buffer which is the buffer used by the graph card to send sections ofthe large image with edge smoothing for each projector.
The primary rendering buffer should be three times the size of the horizontalprojector resolution minus the two 200 pixel overlaps for edge smoothing ascreated at a later stage. This provides a primary rendering buffer with a resolution of 5360 pixelshorizontally and 1080 pixels vertically.
The secondary rendering buffer must have a horizontal size that is three timesprojector resolution. This provides a secondary rendering buffer with a resolution of 5760pixels in the horizontal direction and 1080 pixels in the vertical direction.
In the first step of the method (6), the image to be displayed by the display system is rendered inthe primary rendering buffer.
In step two (7), the left section (11) to be projected by the left is first copied.projector (3 a). The size of this section should be the same as the resolution of itleft projector. This gives the section the size of 1920 pixels horizontally and 1080 pixels in horizontalvertically. To describe the sections' positions and extensions, we divide itthe primary rendering buffer in 5360 pixel columns with 1080 pixels in each column.
Pixel column number 1 corresponds to the column of pixels that starts in the upper leftcorner and ends in the lower left corner of the primary rendering buffer. This meansthat, to copy the left section, copy a rectangle of pixels frompixel column 1 to pixel column 1920 from the primary rendering buffer. Then soyou copy the middle section (12). The middle section is the same size asthe left but to create the overlap of 200 pixels, which is used for edge smoothingbetween the projectors, a rectangle of pixels starting from pixel column 1720 is copied topixel column 3640. Finally, the right section (13) is copied. The right section isalso of the same size as the left one starting from pixel column 3440 to pixel column5360 which creates an overlap of 200 pixels to the middle section.
In step three (8), an edge smoothing is first created on the right side in the left section (14).
This edge smoothing is created by rendering a partially transparent rectangle that is200 pixels horizontally and 1080 pixels vertically. The rendering is performed by a pixel shaderwhich adapts to the color of each pixel and how this happens is described in detail below.
Then two edge equalizers are also created, one on the right side and one on the left side, in itmiddle section (15) and finally an edge leveling on the left side in the right section(16).
In step four (9), you copy the three sections edge to edge in the secondarythe rendering buffer. This rendering buffer is then used to create a texture likerendered across the entire desktop of the computer. The operating system, which must be able to handlemultiple screens, should be set up so that the desktop is distributed over three screens horizontally.
The operating system then ensures that the desktop is divided into three sections, each of which is sent viavideo signals to their corresponding projector. If the projectors in the display system are set upso that the edge edges overlap exactly 200 pixels between the adjacent projectors soan borderless image should be projected on the screen.
In order for a pixel shader to be able to render an edge smoothing in the three sections, you have tofirst create what this method calls edge smoothing textures (23). Theseedging textures are one pixel high and at least as many pixels wide as the numberpixels used for anti-aliasing. The edge texture textures describe the transitionfrom full light to no light in the edge equalizers. A number of anti-alignment textures mustadapted for each projector in the display system because each projector has a uniquecolor reproduction depending on projector technology, product quality, age of the projector lamp, etc.
To create anti-aliases that can adapt to images that have both light and darkbatches, you produce edge smoothing textures through two separate test procedures where youprojects test images in a set-up display system. The test images in the first test procedureconsists of light test images and the second test procedure consists of dark test images. For eachtest image, you can manually or automatically adjust three color intensity curves. Eachcolor intensity curve has as many values as there are pixels in the anti-aliasing texture andthe values range from zero to 255. The values in the first curve (19) correspond to the redthe color components in the pixels of the anti-aliasing texture, the second curve (20) corresponds to thethe green color components of the texture and the last curve (21) correspond to the blue onesthe color components of the texture. These color curves are then used to createedge smoothing textures by copying the values of the three color curves from left to rightinto the corresponding pixel color component of the texture from left to right.
The test images in the first test series are called the light series and the first image in the light series is onea single color image where the entire image is 100 percent red, followed by an image that is 100 percent green,then an image that is 100 percent blue and finally an image that is 100 percent white. For eachtest image, you configure the three color curves for each edge smoothing and these curvescreates a real-time anti-aliasing texture for each anti-aliasing texture. Sincethe edge texture textures are created in real time so you can always see what effect one haschange of a color curve gets on the overlapping areas where two projectors projectsame surface. In order not to have to surface hundreds of points in eachcolor intensity curve, you create a few control points (22) with which you create onebezier curve that fills in intermediate values between the control points.
The test images in the second test series are called the dark series and the first image inthe dark series is a solid color image where the entire image is 25 percent red, followed by an image that is 25percent green, then an image that is 25 percent blue and finally a gray image that is 25 percentred, green and blue.
The actual rendering of an individual edge smoothing is done by using the pixel shaderrenders one or fl your polygons that cover the edge smoothing area withthe edge texture texture. Because the edge smoothing texture is only one pixel high, so mustthe texture is repeated in height so that it covers the entire edge leveling area and thus formsa gradient from full light to no light over the entire area. When to render the pixel shaderthe edge compensation for an individual pixel, it first calculates how light-intensive the pixel is inpercent and we call this result the color intensity factor. The color intensity factor calculatespixel shader by adding together the value of the three firg components red, green and blue whereeach color component has a value between 0 to 255. Then the pixel shadem divides thisresult with the number 765 which is the maximum value of the color components multiplied by the numbercolor components. This creates a color intensity factor that is a fl number between 0 and 1.
The color intensity factor is then used by the pixel shadem to calculate oneedge compensation component for each color component in the pixel. Edge leveling componentsis produced by pixel shadem calculating what percentage of the anti-aliasing color inthe anti-aliasing texture from the light series and what percentage of the anti-aliasing color fromthe dark series to be added together into an anti-aliasing component. This makes pixelsthe shader by multiplying the edge smoothing color from the light series bythe color intensity factor and then multiply the edge smoothing color from the dark series by aminus the color intensity factor and finally add these two results together into oneanti-aliasing component. The pixel shader tends to the final and edge-smoothed pixelby rendering a transparent pixel where the colors of the pixel are taken fromthe edge smoothing components and the transparency, which is a number between 0 and 255, are calculatedby adding the three anti-aliasing components and dividing the result by the number three.
Pixel shadem performs this canoe smoothing procedure for each pixel included inthe edgesPatent claimsMethod for creating large, high-resolution and borderless display systems (1), characterizedby using an individual computer with one graphics card and where the graphics card has twoor fl your video outputs connected to two or fl your projectors as togethercreates borderless projected images by rendering the graphics card a large andhigh-resolution image to a primary rendering buffer (10) and then split itrendered the image in overlapping sections (11) which are then edge smoothed (14) beforemerged into a secondary rendering buffer (17).Device according to Claim 1, characterized in that the edge smoothings are renderedthe graphics card with pixel shader that adapts to the image that is currently to beprojected by calculating the color intensity of each individual pixel in real time and based onthis color intensity factor renders an edged pixel that is customized according to ifthe pixel has high or low color intensity.Device according to claim 2, characterized in that the individual computer has two orgra your graphics cards installed and where the graphics cards are connected to actas a greeting card to increase rendering performance.
权利要求:
Claims (3)
[1]
Method for creating large, high-resolution and borderless display systems (l), characterized by the use of a single computer with a graphics card and where the graphics card has two or more video outputs connected to two or more projectors that together create borderless projected images by the graphics card renders a large and high-resolution image to a primary rendering buffer (10) and then divides the rendered image into overlapping sections (1 1) which are then edge-ironed (14) before being merged into a secondary rendering buffer (17).
[2]
Device according to claim 1, characterized in that the edge services are rendered on the graphics card with pixel shader which adapts to the image to be projected at the moment by calculating the color intensity of each individual pixel in real time and based on this color intensity factor rendering an edge smoothed pixel which is adapted to whether the pixel has high or low color intensity.
[3]
Device according to claim 2, characterized in that the individual computer has two or fl your graphics cards installed and where the graphics cards are interconnected to act as a graphics card to increase the rendering performance.
类似技术:
公开号 | 公开日 | 专利标题
JP6145123B2|2017-06-07|System and method for injection of mapping function
JP2009200613A|2009-09-03|Projector and display adjusting method of projector
CN102081284A|2011-06-01|Edge blending method for splicing multiple projection images
JP4777675B2|2011-09-21|Image processing apparatus, image display apparatus, image processing method, program for causing computer to execute the method, and recording medium
CN101727880B|2012-02-15|Projection fusion method of true seamless rear projection large screen display image
CA2399678C|2010-04-27|Digital correction module for video projector
US8440955B2|2013-05-14|Method for spatial smoothing in a shader pipeline for a multi-projector display
CN111192552B|2021-03-23|Multi-channel LED spherical screen geometric correction method
CN101888511A|2010-11-17|Projection integration method for large screen display device
CN104702929A|2015-06-10|Method of correcting distortion of image overlap area, recording medium, and execution apparatus
CN105574918A|2016-05-11|Material adding method and apparatus of 3D model, and terminal
JP2006081151A|2006-03-23|Graphical user interface for keyer
CN101409790A|2009-04-15|High-efficiency multi-projector splicing and amalgamation method
JP2006033672A|2006-02-02|Curved surface multi-screen projection method, and its device
JP4355508B2|2009-11-04|Video image projector on dome screen
CN100552713C|2009-10-21|A kind of image processing method
CN103716603A|2014-04-09|Camera auxiliary correcting method, device and system based on image fusion processing
JP2009181323A|2009-08-13|Signal processor and projection video display
CN105306852A|2016-02-03|Multi-projector stitching fusion method for high-quality visual effect
SE0901121A1|2011-02-27|Method for creating high-resolution display systems with multiple projectors
JP4008333B2|2007-11-14|Multi-image projection method using a plurality of projectors, projector apparatus for using the method, program, and recording medium
Minomo et al.2005|Transforming your shadow into colorful visual media: multi-projection of complementary colors
US20090167949A1|2009-07-02|Method And Apparatus For Performing Edge Blending Using Production Switchers
JP2001306024A|2001-11-02|Device and method for generating luminance-corrected image
US20020158877A1|2002-10-31|Shadow buffer control module method and software construct for adjusting per pixel raster images attributes to screen space and projector features for digital wrap, intensity transforms, color matching, soft-edge blending and filtering for multiple projectors and laser projectors
同族专利:
公开号 | 公开日
SE534866C2|2012-01-24|
引用文献:
公开号 | 申请日 | 公开日 | 申请人 | 专利标题

法律状态:
优先权:
申请号 | 申请日 | 专利标题
SE0901121A|SE534866C2|2009-08-26|2009-08-26|Method for creating high-resolution display systems with multiple projectors|SE0901121A| SE534866C2|2009-08-26|2009-08-26|Method for creating high-resolution display systems with multiple projectors|
[返回顶部]