WO2003083559A2

WO2003083559A2 - Color non-uniformity alignment for light engines

Info

Publication number: WO2003083559A2
Application number: PCT/US2003/009453
Authority: WO
Inventors: Frank Albert Glad; Peter Scott Hillenberg
Original assignee: Thomson Licensing S.A.
Priority date: 2002-03-27
Filing date: 2003-03-26
Publication date: 2003-10-09
Also published as: AU2003228384A8; AU2003228384A1; WO2003083559A3

Abstract

A system of color non-uniformity alignment for a liquid crystal display (290), includes light measuring equipment (292 and 294) for measuring correction information at a plurality of plane points on the liquid crystal display and memory (250) for storing the correction information, a plurality of matrix memories (205, 275 and 245) utilizing the correction information, a plurality of horizontal and vertical interpolators (210, 260, and 220) coupled to the plurality of matrix memories, a pixel brightness detector (225), a soft switch (230) receiving control information from the pixel brightness detector, a dither circuit (235) to provide effective resolution to the correction information, and a processor (240) to add the correction information to an input video signal.

Description

COLOR NON-UNIFORMITY ALIGNMENT FOR LIGHT ENGINES

Cross Reference to Related Applications

This invention claims priority under 35 U.S.C. section 119(e) to U.S. Provisional Application Serial No. 60/367,974 filed March 27, 2002, the teachings of which are incorporated by reference herein.

Background of the Invention Technical Field This invention relates generally to the field of liquid crystal projection displays and micro-mirror image displays, and in particular, to method for color uniformity alignment in imager based display systems having electronic compensation for the color non-uniformity of light engines having liquid crystal imagers, for example liquid crystal on silicon (LCOS) imagers, and micro-mirror imagers.

Description of the Related Art

LCOS display systems incorporate a high pressure lamp and a light engine for generating a video display in lieu of a cathode ray tube found in traditional video displays. The light engine receives ultra bright light from the high pressure lamp and processes the light through display optics contained within the light engine. Display optics are typically provided for each of the base colors, namely red, green and blue. Variations between the display optics tend to cause color non-uniformity in the LCOS displays. For example, the green optics of a particular LCOS light-engine may be slightly more transmissive in the top left corner. This would produce a green zone in the top left corner of the displayed image. Other LCOS light-engines can have non-uniform zones in other areas of the display. The degree and nature of color non-uniformity varies tremendously from light-engine to light-engine. Notwithstanding, color non- uniformity also changes as a function of pixel brightness. Hence the color non- uniformity can vary as the brightness level of video changes. Color non-uniformity correction has been implemented to address color non- uniformity in LCOS displays, however current implementation requires substantial processing resources to implement color non-uniformity correction. Color non- uniformity correction is typically performed after frame rate doubling and gamma correction have been applied to a video signal. To perform color non-uniformity correction after frame rate doubling results in the color non-uniformity correction being applied to twice as much data as contained in an original video signal. Further, gamma correction increases the size of the video data. Hence, performing color non-uniformity correction after gamma correction further increases the amount of video data that must be processed.

On the other hand, applying color non-uniformity correction prior to frame rate doubling and gamma correction creates an even greater obstacle. Color non-uniformity correction typically increases the amount of video data contained in a video signal as well. However, the frame rate doubler is limited with respect the amount of video data that can be processed. Thus a "bit bottleneck" is created and video data incorporating color non-uniformity correction cannot be adequately processed by the frame rate doubler. Even if the color non-uniformity corrected video could be adequately doubled, the amount of video data handled by the rest of the processing components would be increased. New color non-uniformity correction systems need new alignment systems. Thus, a need exists for a method and system for color non-uniformity alignment of light engines prior to increasing the frame rate that overcomes the problems described above.

Summary of the Invention

A machine vision measurement method and algorithm for aligning color temperature of an array of points and/or areas on a liquid crystal display, such as an LCOS or micro-mirror video display, in accordance with the inventive arrangements, is a particularly useful with certain electronic systems for correcting color non-uniformity correction.

An example of such a system is shown in FIGs 1 -4. There are several advantages to having such a system, wherein color non-uniformity correction is performed at the front end, that is, prior to frame rate doubling and gamma correction. A first advantage is that a fewer number of correction planes are needed, as compared to the prior art topology, because the correction occurs prior to the non-linearities introduced by the gamma correction. The compensation process is therefore linear and thus less complex. A second advantage is that the color non-uniformity correction can operate at 1x rather than 2x, that is, not at frame-doubled rate. No frame-alternate inversion is necessary, thus further reducing the complexity. A third advantage relates to yield performance in the manufacture of the light engines. Color non-uniformity can be a frequent cause of quality control failure in manufacturing light engines. The enhanced ability to correct color non-uniformity electronically, with signal processing, has the added benefit of increasing manufacturing yields for the light engines. In accordance with one aspect of the present invention, a method for correcting color non-uniformity in a liquid crystal display system that receives an incoming video signal comprises the steps of determining coordinates of a grid including an overall center coordinate and a center coordinate for each rectangle in the grid and measuring an RGB ratio of the overall center coordinate and using this measure as a target RGB ratio for each center coordinate in the grid for a selected gain level range. The method further comprises the step of determining a dominant color at each rectangle in the grid, assigning a zero correction value to the dominant color and increasing correction values for a remaining set of weaker colors to match the target RGB ratio for each coordinate in the grid and changing the correction values for the remaining set of weaker colors for all coordinates in the grid using until color non-uniformity is corrected in the liquid crystal display system.

In another aspect of the present invention, a system of color non-uniformity alignment for a liquid crystal display comprises light measuring equipment for measuring correction information at a plurality of plane points on the liquid crystal display, memory for storing the correction information, a plurality of matrix memories of RGB correction planes utilizing the correction information, a plurality of horizontal and vertical interpolators coupled to respective ones of the plurality of matrix memories, and a pixel brightness detector for determining whether pixels being processed have a brightness in the ranges corresponding to the RGB correction planes. The system further comprises a soft switch receiving control information from the pixel brightness detector to control a blend of brightness planes, a dither circuit to provide effective resolution to correction information from the RGB correction planes, and a processor to add the correction information to an input video signal.

Brief Description of the Drawings

FIG. 1 is a block diagram useful for explaining the topology of the color non- uniformity correction system.

FIG. 2 shows a block diagram of an embodiment of a system for correcting color non-uniformity in accordance with the inventive arrangements.

FIG. 3 illustrates a diagram representing a correction plane or projection screen and useful for explaining the operation of the system shown in FIG. 2 and the method shown in FIG. 5.

FIG. 4 illustrates a diagram representing a correction plane or projection screen and useful for explaining the operation of the system shown in FIG. 2 and the method shown in FIG. 5.

Figure 5 is a flow chart useful for explaining the process of the alignment system.

Detailed Description of the Preferred Embodiments

Overview of System Architecture

Regardless of the location of the color non-uniformity correction block, frame rate doubling and gamma correction must be performed upstream of the light engine. These blocks are needed to satisfy the requirements of the light engine, which can use a liquid crystal imager or a micro-mirror imager or the like.

FIG. 1 shows a block diagram of a color non-uniformity correction system 100 incorporating color non-uniformity correction. The system 100 can receive video data from a video source. The video source can be a multimedia device, for example a television tuner, a digital video disk (DVD) player, a video cassette recorder (VCR), a personal video recorder (PVR), a digital broadcast receiver, etc. Typically, the video source can incorporate a video processor for processing multimedia data and outputting video data in a format compatible with the system 100.

The system 100 can comprise a video input, color non-uniformity corrector 110, a frame rate doubler 120, a gamma corrector 125 and associated lookup tables in a memory and a light engine 140. The video input can receive an input video signal from a video source and forward the input video signal to the color non-uniformity corrector 110.

The color non-uniformity corrector 110 can incorporate color non-uniformity correction into the input video signal to generate a color corrected video signal. A memory (not shown) can store a plurality of red, green, and blue (RGB) correction matrices for correcting the RGB brightness levels of a LCOS screen (not shown). The RGB correction matrices can be used for generating non-uniformity color correction and can be created from test measurements made on the LCOS display using the systeml 00 during or after manufacture of the LCOS display 100.

Liquid crystals of the light engine 140 can be damaged by DC voltage and may only be driven by AC voltage. Since liquid crystals are responsive to the absolute value of a voltage, pixels can be driven by a first polarity of signal, then an opposite polarity of signal. More precisely, liquid crystal imagers are driven by electric potentials of alternating positive and negative polarity. Many alternating schemes can be implemented provided that the average DC level of the electric fields driving the imager, for example an LCOS imager, is zero. Pixels in the imagers can be driven by opposite polarity signals because liquid crystals are responsive to the absolute value of the applied voltage. Accordingly, frame rate doubler 120 can generate two frames of video in opposite polarities from each frame of color corrected video received. Hence, each frame of the color corrected video signal may be displayed twice, once for each polarity. The frame rate doubler stores each frame of video so that each frame can be displayed twice; once for each drive polarity. This means the video after frame rate doubling is at a 2x rate. If the input video is at 1fH (where fH is a horizontal scanning frequency) the clock rate is doubled to 2fH, if the input video is at 2fH the clock rate is doubled to 4fH, and so on. All processing subsequent to the frame doubling must run at the higher 2x clock rate or be implemented with multiple parallel processors. Either approach adds considerable complexity and cost. Nevertheless, it should be noted that the LCOS display 100 is not limited to only doubling the frame rate. For example, in another arrangement the frame rate of the color corrected video signal can be increased by any multiple, for example 3X, 4X, 5X, 6X, etc.

Gamma corrector 125 can apply gamma correction to a color corrected video signal after the frame rate of the signal has been doubled to generate an output video signal. Gamma correction is a nonlinear transfer characteristic that corrects for the nonlinear relationship between a liquid crystal imager reflectivity and an applied voltage. Gamma correction is normally done after frame rate doubling to conserve memory. Since gamma correction increases the number of bits, from 8-bits to 10-bits in most systems, it is advantageous to store a frame of video before increasing the number of bits of resolution. In other words, since gamma correction increases the number of bits in a video signal and a frame rate doubler 120 is typically limited with respect to the amount of video data that can be processed, having the gamma corrector 125 after the frame rate doubler 120 is advantageous. The memory in the gamma corrector 125 can store a plurality of gamma correction data in the form of lookup tables associated with the light engine 140.

Light engine 140 can receive light from high pressure lamp (not shown). The light can be directed into a prism in the light engine where it is divided into the three RGB primary color streams. These three individual streams of light are directed to corresponding light valves, which are highly reflective mirrors contained in the light engine known as imagers, where they can pick up the output video signal as they are reflected off the surface of the imager. The three streams of light, each with their specific piece of the same video signal attached, then can be recombined within the prism into a single synchronized video image that is forwarded to the LCOS screen for display. It should be noted that although the present embodiment is an LCOS display, the present invention is not limited to such and can be used with other liquid crystal displays.

The color non-uniformity correction system 110 of FIG. 1 is shown in block diagram form in FIG. 2. This system provides a programmable RGB correction to video, dependent on spatial position and brightness of the pixels.

Interposing this system 110 ahead of gamma correction and prior to frame doubling in the video signal flow as shown in FIG. 1 makes it possible to utilize a smaller number of correction planes than in the prior art, for example only two or three correction planes. In a two-correction plane system, one correction plane is used for , lower brightness correction and the other correction plane is used for higher brightness correction. In a three-correction plane system, the additional correction plane is used for a middle brightness range.

The color non-uniformity alignment system 110 shown in FIG. 2 comprises three memories or grids of 16 x 9 or 16 x 12 point RGB correction planes 205, 275, and 245, respectively, for low, middle and high ranges of brightness levels and respective horizontal and vertical interpolators 210, 260 and 220. The system 110 further comprises a pixel brightness detector 225 and a soft switch 230 to control blend of low and high brightness planes, a dither circuit 235 to provide effective sub-LSB resolution to the correction information, and, a processor 240 such as an adder circuit to add the correction information to the video. The high brightness 16 x 9 or 16 x 12 grid block 205 has a matrix of color non-uniformity correction values stored therein for a higher range of pixel brightness values, one for each grid point of the 16-wide by 9-high grid or matrix (as shown in FIG. 3) or 16-wide by 12-high grid or matrix (as shown in FIG. 4). The low brightness 16 x 9 or 16 x 12 grid block 245 has a matrix of color non-uniformity correction values stored therein for a low range of pixel brightness values, one for each grid point of the 16-wide by 9-high grid or matrix or 16-wide by 12-high grid or matrix. The middle brightness 16 x 9 or 16 x 12 grid block 275 has a matrix of color non- uniformity correction values stored therein for a middle range of pixel brightness values, one for each grid point of the 16-wide by 9-high grid or matrix or 16-wide by 12-high grid or matrix. The respective horizontal and vertical interpolation blocks (210, 260 or 220) compute color non-uniformity correction values that lie in between the stored values. The brightness detector 225 determines whether the pixels being processed have a brightness in the high, middle or low brightness range. The brightness detector 225 then provides one or more control signals, for example a control signal with two or more bits, to the soft switch 230 indicative of two threshold levels, one between the low and middle range brightness values and another between middle and high range brightness values. The soft switch 235 supplies high brightness, middle brightness or low brightness correction values responsive to the brightness detector 225. The selected correction values are supplied to a dither circuit 235. The dithered correction signal is added back to the video input signal and the result is supplied to the frame doubler. 16 x 9 or 16 x 12 Correction Planes

Color non-uniformity has a fairly broad and smooth characteristic. It has been determined that it is sufficient to represent the non-uniformity with a 16 x 9 plane as shown in FIG. 3 or a 16 x 12 (4 x 3) plane as shown in FIG. 4, depending upon the aspect ratio of the imager (not shown), or the aspect ratio of a portion of an imager if only that portion is utilized. A portion of an image is used, for example, when a 4 x 3 imager is used to display a 16 x 9 image. The information for each correction plane must be produced with light measuring equipment (camera 292 and external alignment computer 294) at each of the plane points on the television display, and then stored in an EEPROM 250 in the television receiver, as explained below.

It is useful to note the difference between the number of bits in a signal and the effective or apparent quantization resolution of a signal. The number of bits in a signal is evident from the bit designation, for example, 6-bit or 8-bit. Effective or apparent quantization resolution is determined by the incremental value of the least significant bit (LSB). Accordingly, a 6-bit signal said to have 8-bit quantization has no fractional bits (i.e., no bits that represent fractional values), and in the context of this invention, the LSB represents a normalized brightness level of 100/256 IRE. An 8-bit signal said to have 10-bit effective or apparent quantization has 2 fractional bits (i.e., bits that represent fractional values) and the LSB represents 100/1024 IRE.

The red, green and blue coefficients are 6-bit values with 8-bit effective quantization. The coefficients are stored for each plane point, for both low brightness and high brightness conditions in a two-plane system, and also for middle brightness conditions in the three-plane system. All of the bits including the least significant bit (LSB) are integers.

Horizontal and Vertical Interpolators

Even though 16 x 9 or 16 x 12 plane points are sufficient to represent the correction plane, the correction plane must be smoothed both horizontally and vertically or the correction information will take on a blocky appearance. Linear interpolation is used to calculate the correction for any spatial location, using the four surrounding plane points. The smoothed correction is calculated in real time based on the stored 16 x 9 or 16 x 12 plane and the row and column of the pixel being displayed. The interpolators add two fractional bits of resolution to the six integer bits of the stored coefficients, resulting in an 8-bit signal having two least significant fractional bits, that is, having an effective quantization of 10-bits. Brightness Detector and Soft Switch

The brightness detector 225 and soft switch 230 work together to determine the brightness of a pixel, and then apply the appropriate blend of correction planes. The soft switch provides a soft (in other words, gradual) switching characteristic in the vicinity of the brightness-switching threshold, to reduce switching artifacts. In a two- plane system, for pixels having a brightness level near the threshold between low and high brightness conditions, a blend of low and high correction planes is used. The threshold of the soft switch is programmable to allow optimum performance over a wide range of light engine characteristics. A three-plane system utilizes two such thresholds. The correction system is additive. This processing has the further advantage of reducing the gain of the correction signal at very high brightness levels to prevent limiting of the combined video plus correction signal. Dithering The disadvantage of having color non-uniformity correction before the frame rate doubler and gamma correction is that processing must be performed with reduced quantization accuracy. This is due at least in part to the design-imposed "bit bottleneck" (8-bits, to conserve memory) through the frame rate doubler, a not uncommon engineering compromise. To prevent visible contouring of the correction planes, color non-uniformity correction normally requires 10-bit quantization. The invention effectively solves this problem through the use of dithering (235).

Although the stored correction plane coefficients have 8-bit effective quantization, the horizontal and vertical interpolators and soft switch generate two additional fractional bits of resolution, providing 10-bit effective quantization. This resolution is needed to produce invisible correction contours.

Because the video through the color non-uniformity corrector is 8-bits, a dither 235 is required to make use of the fractional resolution correction signal. The dither 235 adds the line rate sequence 0, 2, 1 , 3 . . . to the 10-bit effective quantized correction signal from the soft switch, and the result is truncated to a 6-bit signal having 8-bit quantization before being summed with the video signal. The information content of the least significant bits is thus migrated upwardly into the more significant bits so that the information content of the bits to be truncated remains in the signal after truncation. This technique fools the eye into seeing the fractional information. On alternate frames the line rate sequence can be advantageously changed to the complementary phase 3, 1 , 2, 0 . . to reduce the visibility of the dither temporally.

It should be noted that the dither used for this system is advantageously a vertical dither, designed not to introduce horizontal transitions into the signal. The reason is that the imagers used in these kinds of systems are susceptible to declination errors causing an artifact often referred to as "green sparkle". Declination errors are aggravated by horizontal transitions, particularly at lower brightness levels where the gamma table is most non-linear. Alignment System and Method

The color non-uniformity correction system 110 utilizes the internal microprocessor (μP) 252 of the television receiver and a programmable EEPROM 250 coupled thereto, as shown in FIG 2. The microprocessor 252 controls numerous functions in the television receiver utilizing signals transmitted on an i²c bus. The television receiver has a plug or jack for connecting the i²c bus to an external device. In the alignment system, the external device is an alignment computer 294. The computer 294 is coupled to a camera 292 that is used as the measurement device. The camera is, in the presently preferred embodiment, an RGB CCD camera with computer controlled gain and with a compatible computer controlled frame storage device.

In accordance with the process a screen 290 is illuminated by a bright, over- scanned, flat white field and an image of the screen is acquired by the camera 292. The screen is divided into images of rectangles that enclose each of the required measurement points on the screen, as shown in FIG. 4. Measuring is defined as performing a neighborhood average of the gray scale level in each of the bounding rectangles (32 pixels x 32 pixels) for red, green, and blue independently, at multiple brightness levels or correction planes, and calculating ratios for each level for use in color matching, that is, correcting color non-uniformity. The rectangles need not be square, for example as might be the case on a 16 x 9 screen.

The correction values are stored in registers (not shown) that make up the high, middle and low grids 205, 275, 245 shown in FIG. 2. The registers hold 6-bit correction values. There is one register, and corresponding correction value, for each primary color (RGB) at each of the 192 alignment points illustrated in FIG. 4. There are 3 complete sets of the 576 (192 multiplied by 3) registers, one set for each of 3 luminance levels represented by the high, middle and low brightness grids. The number of alignment points can be different for a 16 x 9 screen, such as that shown in FIG. 3, but the alignment process is the same. In the presently preferred embodiment, the three luminance or brightness levels are chosen to be in the range of: 7-10 IRE for low; 25-30 IRE for middle; and, 50-70 IRE for high. Camera gain or aperture is preferably adjusted or preset to optimize the gray scale range at each of the three luminance levels. A stepwise alignment procedure 500 is explained in connection with the flow chart in FIG. 5.

The first step 501 , undertaken with a lens cap on the camera, is measuring and recording noise floors of the camera for red, green, and blue each of the three different gain levels or ranges used for the alignment process. The noise values are subtracted from all subsequent RGB measurements.

The second step 502 is displaying a bright, over-scanned, flat white field on the television screen, preferably in a darkened space, acquiring an image with the camera such as shown in FIG. 2, locating the edges of the screen, calculating the x, y coordinates of the center of the screen and calculating the x, y coordinates of the center of each rectangle in the matrix, such as shown in FIG. 4.

The third step 503 is selecting a video brightness level or range, in this example either high middle or low, setting the camera gain to match the selected video level, with respect to optimizing the gray scale range as explained above, and measuring and recording the center color reference. The RGB ratio measurement at the center of the screen is used as the target RGB ratio for all locations in the matrix, 192 points in the example of FIG. 4. As noted, 1024 pixels in the center of the screen (32 multiplied by 32) are averaged for the center reference measurements.

The fourth step 504 is determining the strongest or dominant, color, that is the brightest primary color, relative to the center reference, at each of the locations in the matrix. The register for the dominant primary color is assigned zero correction value at that point. The two weaker colors are increased to match the RGB ratio at the center. The fifth step 505 is simultaneously increasing the brightness of the two weaker primary colors at all points using large increments, that is multiple steps, until both of the weaker colors are above their respective target threshold, or until the registers reach a maximum value. The maximum value setting for a 6-bit register is 63. This step quickly provides rough values that are close to the targets.

The sixth step 506 is simultaneously decreasing the two weaker color register settings, in single register setting steps, until the weaker colors at all points are just below their targets, or until their register settings are zero.

The seventh step 507 is determining if more than one of the three color registers at each point is set to zero, and if so, determining which color is currently the strongest color. This step is important because interaction between points can cause the dominant primary color at a matrix point to change during alignment. The eighth step 508 is simultaneously increasing the two weaker color register settings, in single register setting steps, until both colors are above their targets, or until their register settings reach the maximum value.

The ninth step 509 is determining whether more iterations are needed to compensate for the interaction. If more iterations are needed, the process branches back to step 506. Steps 506 through 509 are repeated until further iterations are not needed, at which time the process continues with the tenth step.

The tenth step 510 is recording the current RGB levels and RGB targets for all of the colors at all of the matrix locations. These measurements represent a first result for each of the two primary colors at each point of the matrix, stored in the external alignment computer.

The eleventh step 511 is decreasing the weaker two colors at each point by one register setting step, provided that the result is not a zero setting, and recording the current RGB levels and RGB targets for all colors at all of the matrix locations. These measurements represent a second result for each of the two primary colors at each point of the matrix, stored in the external alignment computer.

The twelfth step 512 is examining the first and second sets of recorded results, and selecting the best value for each of the weaker colors at each of the matrix locations. Due to the digital nature of the system, the alignment values will be either a little to high or a little too low, and a choice must be made between the values of the first and second results on a point by point basis for each weaker color at each matrix location. The selected values are expected to be a blend of the values of the first and second results. The selected values for setting the registers for the weaker colors are transmitted over the i²c bus to the microprocessor (252) in the television receiver, for storage in the EEPROM (250). The thirteenth step 513 is determining if there are more video levels requiring alignment. If any video levels still need to be aligned, then the process branches back to step 503, so that steps 503 through 512 are repeated for each remaining video level.

If no video levels need to be aligned, the process ends at the fourteenth step 514.

The alignment values stored in the EEPROM 250 are retrieved by the microprocessor 252 and loaded into the registers in the grid blocks (205, 275, and 245) shown in FIG. 2 whenever the television receiver is turned on.

It should be understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof can be suggested by persons skilled in the art and are to be included within the spirit and purview of this application. For example, the present invention is equally applicable to micro-mirror based systems as well as LCOS systems using multiple imagers or a single imager with sequential color systems. The invention can take many other specific forms without departing from the spirit or essential attributes thereof for an indication of the scope of the invention.

Claims

1. A method for correcting color non-uniformity in a liquid crystal display system that receives an incoming video signal, CHARACTERIZED BY: determining coordinates of a grid including an overall center coordinate and a center coordinate for each rectangle in the grid; measuring an RGB ratio of the overall center coordinate and using this measure as a target RGB ratio for each center coordinate in the grid for a selected gain level range; determining a dominant color at each rectangle in the grid, assigning a zero correction value to the dominant color and increasing correction values for a remaining set of weaker colors to match the target RGB ratio for each coordinate in the grid; changing the correction values for the remaining set of weaker colors for all coordinates in the grid until color non-uniformity is corrected in the liquid crystal display system.

2. The method of claim 1 , wherein the method further comprises the step of determining a noise floor for a plurality of gain level ranges for red, green, and blue.

3. The method of claim 1 , wherein the step of changing the correction values comprises simultaneously at all coordinates on the grid increasing the correction values for the remaining set of weaker colors for all coordinates in the grid using large increments until the remaining set of weaker colors surpass the target RGB ratio and simultaneously for all coordinates on the grid decreasing the correction values for the remaining set of weaker colors for all coordinates in the grid using small increments until the remaining set of weaker colors are below their respective target RGB ratio wherein if more than one of the color correction values is zero, determining which of the colors is the strongest.

4. The method of claim 3, wherein the method further comprises the step of simultaneously increasing the correction values for the remaining set of weaker colors for all coordinates in the grid using small increments until the remaining set of weaker colors surpass their respective target RGB ratio.

5. The method of claim 1 , wherein the method further comprises the step of repeating the steps of simultaneously increasing, decreasing, and increasing the correction values until no more than one of the correction values for each coordinate in the grid is set to zero.

6. The method of claim 5, wherein the method further comprises the step of recording the current RGB levels and RGB targets for all the colors at all of the coordinates of the grid to represent a first result for each of two primary colors for each coordinate of the grid stored in an external storage.

7. The method of claim 6, wherein the method further comprises the step of decreasing the correction values for the remaining set of weaker colors for all coordinates in the grid using small increments provided that the result is above zero and recording the current RGB levels and RGB targets for all the colors at all of the coordinates of the grid to represent a second result for each of two primary colors for each coordinate of the grid stored in the external storage.

9. The method of claim 7, wherein the method further comprises the step of examining levels recording in the first and second results and selecting the best correction value for each weaker color at each location.

10. The method of claim 8, wherein the method further comprises the step of transmitting the best correction values to a microprocessor for storage in an EEPROM (250).

11. The method of claim 9, wherein the method further comprises the step of retrieving the best correction values and other alignment correction values from the EEPROM (250) and loading the best correction values and other alignment correction values into registers of grid blocks for a respective gain level range.

12. The method of claim 2, wherein the plurality of gain level ranges comprises 3 levels: a high brightness gain level range, a middle brightness gain level range, and a low brightness gain level range.

13. The method of claim 1 , wherein the step of measuring the RGB ratio of the overall center coordinate comprises the step of averaging the RGB ratio for a predetermined matrix of pixels at the center of the liquid crystal display.

14. The method of claim 3, wherein the step simultaneously decreasing comprises the step of simultaneously decreasing the correction values for the remaining set of weaker colors for all coordinates in the grid using small increments until the remaining set of weaker colors are below their respective target RGB ratio or until the remaining set of weaker colors have a correction value equals zero.

15. The method of claim 10, wherein the method further comprises the step of repeating the method for a subsequently selected gain level starting at the step of measuring the RGB ratio of an overall center coordinate and using this measure as a target RGB ratio for each center coordinate in the grid until there are no more selected gain levels.

16. The method of claim 2, wherein the step of determining the noise floor comprises the step of measuring and recording the noise floors at 3 different gain levels corresponding to the plurality of gain level ranges using a camera with a cap placed over the lens.

17. The method of claim 15, wherein the method further comprises the step of subtracting the noise floors from all subsequent RGB measurements.

18. The method of claim 1 , wherein the step of determining coordinates comprises the step of detecting the edges of a screen displaying a bright white flat field patterned in a darkened room.

19. A method of color non-uniformity alignment for a liquid crystal display,: determining a noise floor for a plurality of gain level ranges for red, green, and blue, CHARACTERIZED BY; determining coordinates of a grid including an overall center coordinate and a center coordinate for each rectangle in the grid; measuring an RGB ratio of an overall center coordinate and using this measure as a target RGB ratio for each center coordinate in the grid for a selected gain level range; determining a dominant color among red, green, and blue at each rectangle in the grid, assigning a zero correction value to the dominant color and increasing correction values for a remaining set of two weaker colors to match the target RGB ratio for each coordinate in the grid; performing a coarse upward correction for the remaining set of two weaker colors for all coordinates simultaneously in the grid until the remaining set of two weaker colors surpass their target RGB ratios; performing a fine downward correction for the remaining set of two weaker colors for all coordinates in the grid simultaneously until the remaining set of two weaker colors are below their respective target RGB ratios; if more than one of the color correction values is zero, determining which of the colors is the strongest for all coordinates in the grid simultaneously; and performing a fine upward correction for the remaining set of two weaker colors for all coordinates in the grid until the remaining set of two weaker colors surpass their respective target RGB ratios.

20. The method of claim 18, wherein the method further comprises the step of recording the current RGB levels and RGB targets for all the colors at all of the coordinates of the grid to represent a first result for each of two primary colors for each coordinate of the grid stored in an external storage.

21. The method of claim 19, wherein the method further comprises the step of decreasing the correction values for the remaining set of weaker colors for all coordinates in the grid using small increments provided that the result is above zero and recording the current RGB levels and RGB targets for all the colors at all of the coordinates of the grid to represent a second result for each of two primary colors for each coordinate of the grid stored in the external storage.

22. The method of claim 20, wherein the method further comprises the step of examining levels recorded in the first and second results and selecting a best correction value for each weaker color at each location.

23. The method of claim 21 , wherein the method further comprises the step of transmitting the best correction values to a microprocessor for storage in an EEPROM (250).

24. The method of claim 22, wherein the method further comprises the step of retrieving the best correction values and other alignment correction values from the EEPROM (250) and loading the best correction values and other alignment correction values into registers of grid blocks for a respective gain level range.

25. A system of color non-uniformity alignment for a liquid crystal display, CHARACTERIZED BY: light measuring equipment (292, 294)for measuring correction information at a plurality of plane points on the liquid crystal display; memory for storing the correction information (250); a plurality of matrix memories of RGB correction planes (205, 245, 275) utilizing the correction information; a plurality of horizontal and vertical interpolators (210, 220 AND 260) coupled to respective ones of the plurality of matrix memories; a pixel brightness detector (225) for determining whether pixels being processed have a brightness in the ranges corresponding to the RGB correction planes; a soft switch (230) receiving control information from the pixel brightness detector to control a blend of brightness planes; a dither circuit (235) to provide effective resolution to correction information from the RGB correction planes; and a processor (240) to add the correction information to an input video signal.

26. The system of claim 24, wherein the plurality of matrix memories include RGB correction planes respectively for lower, middle and higher ranges of brightness levels.

27. The system of claim 24, wherein the brightness detector further provides one or more control signals to the soft switch indicative of two threshold levels, one between lower and middle range brightness values and another between middle and higher range brightness values.

28. The system of claim 24, wherein the light measuring equipment comprises a camera and an external alignment computer.