(19) United States
`(12) Patent Application Publication (10) Pub. No.: US 2005/0093798 A1
`Kamada et al.
`(43) Pub. Date:
`May 5, 2005
`
`US 20050093798A1
`
`(54) CORRECTION OF UNEVEN IMAGE
`APPEARANCE BY USE OF SMALL-SIZE
`DATA
`
`(75) Inventors: Tsuyoshi Kamada, Kawasaki (JP);
`Kazuhiro Nukiyama, Kawasaki (JP);
`Toshiaki Suzuki, Kawasaki (JP)
`
`Correspondence Address:
`Patrick G. Burns, Esq.
`GREER, BURNS & CRAIN, LTD.
`Suite 2500
`300 South Wacker Dr.
`Chicago, IL 60606 (US)
`
`(73) Assignee: FUJITSU DISPLAY TECHNOLO
`GIES CORPORATION
`
`(21) Appl. No.:
`
`10/843,039
`
`(22) Filed:
`
`May 11, 2004
`
`(30)
`
`Foreign Application Priority Data
`
`Oct. 29, 2003 (JP) .................................... .. 2003-369317
`
`Publication Classi?cation
`
`(51) Int. Cl? ........................... .. G09G 3/36; 6096 5/00;
`H04N 1/46
`(52) US. Cl. ............................................ .. 345/89; 345/204
`
`(57)
`
`ABSTRACT
`
`A circuit for display correction includes a memory which
`stores ?rst data indicative of siZe and position of a rectan
`gular region on a display screen and second data indicative
`of gray level changes in a surrounding region around the
`rectangular region in an isometric manner with respect to a
`horizontal direction and a vertical direction, and an image
`processing unit which adjusts gray levels of image data in
`response to the ?rst data and the second data stored in the
`memory.
`
`12
`
`MEMORY
`
`13
`
`SIGNAL
`SOURCE
`
`11 lMAGE PROCESSING APPARATUS
`
`21
`CORRECTION
`DATA
`STORAGE
`UNIT
`
`FIFO
`
`22
`
`CORRECTION
`PROCESSING
`
`23
`
`UNIT
`
`14
`
`~
`LIQUID CRYSTAL
`DISPLAY PANEL
`
`

`
`Patent Application Publication May 5, 2005 Sheet 1 0f 11
`
`US 2005/0093798 A1
`
`TOE
`
`:z: 2 .
`
`
`
`ozmmmooma . :z:
`
`zo?ommmoo M9205
`
`
`
`
`
`mm \|\ . . R 1
`
`. zopommmou
`
`<20 A
`
`\ll\
`
`

`
`Patent Application Publication May 5, 2005 Sheet 2 0f 11
`
`US 2005/0093798 A1
`
`GRAY
`
`‘ LEVEL
`
`SHIFT /
`
`FIG.3
`
`FIG.4
`
`'éei
`'
`'
`(W1)
`
`POSITION
`
`GRAY
`LEVEL
`SHIFT k
`
`GRAY LEVEL
`
`

`
`Patent Application Publication May 5, 2005 Sheet 3 0f 11
`
`US 2005/0093798 A1
`
`FIG.5A
`
`FIG.5B
`
`if x<x1-w1 then
`k=0
`elsehc x<x1 then
`k=k * (x1-x)/w1
`elseif x>x2+w1 then‘
`
`v
`k=0
`elseif x>x2 then
`k=k * (x-x2)/w1
`
`else
`
`end
`
`.
`
`k=k
`
`elseif y<y1 then
`
`FIG.5C
`
`end
`
`FIG.5D
`
`Output Gray Scale = Input Gray Scale
`
`

`
`Patent Application Publication May 5, 2005 Sheet 4 of 11
`
`US 2005/0093798 A1
`
`EEHIHHHHE
`
`o.0_..._i
`
`
`
`zm_>mz:oz.<m.4.EzoNEo:
`
`
`
`
`
`
`EEHEEEEmoz<m<m_n_n_<zm>m_z:m<.5oEo
`
`HEHHHEEHE-
`
`HEHHIEEHE
`
`
`
`
`
`moz<m<m_n_n_<zm_>m_z:x<m_Em
`
`
`
`

`
`Patent Application Publication May 5, 2005 Sheet 5 0f 11
`
`US 2005/0093798 A1
`
`r_____________________._____
`
`TUE
`
`0251552
`
`.
`
`zo?ommmoo
`
`L
`
`.523
`
`, mn_<Im
`
`
`
`zOFOmmmOO Alll
`
`
`
`ZOFOmEmOO .AITII
`
`:2:
`
`.52:
`
`
`
`.52: mO<mO._.w
`
`<P<Q
`
`ZOFOmmEOO
`
`

`
`Patent Application Publication May 5, 2005 Sheet 6 0f 11
`
`US 2005/0093798 A1
`
`FIG.8A
`
`k= Hg“
`
`k=k
`
`-
`
`end
`
`if y<y1~w1 then
`k=0
`elseif y<y1 then
`
`else'rf y>y2+w1 then
`k=0
`elseif y>y2 then
`
`end
`
`FIGBC
`
`Output Gray Scale = Input Gray Scale
`
`+
`
`k
`
`

`
`Patent Application Publication May 5, 2005 Sheet 7 0f 11
`
`US 2005/0093798 A1
`
`LUT (GRAY LEVEL)
`gs
`k0
`
`0
`
`1
`
`1
`
`2
`
`'
`gs = gs * 1/4
`
`7
`
`k0 = f( gs)
`
`_
`
`~
`
`'
`
`1
`
`'
`
`>
`
`2
`
`32
`
`12
`
`164 ' \A\ 0
`
`40
`
`_
`
`FIG.9B
`
`-
`
`FIG.9C
`
`if x<x1 then
`x = x1 - x
`
`elseif x>x2 then
`
`X = X - X2
`else
`x = 0
`end
`
`if y<y1 then
`y = v1 - y
`elseif y>y2 then
`v = v - v2
`else
`»
`
`v = 0
`end
`
`I
`
`41
`
`LUT (POSITIONNJ
`k = k0 - ?x) - ?y)
`ifk*k0 <0 then \ gs
`k
`k=0
`\ 1
`end
`2
`
`1
`1
`
`i
`
`'
`
`30
`31
`
`12
`12
`
`Output Gray Scale = Input Gray Scale
`
`+
`
`k
`
`

`
`Patent Application Publication May 5, 2005 Sheet 8 0f 11
`
`US 2005/0093798 A1
`
`FIG.1O
`
`GRAY
`LEVEL
`SHIFT
`
`k1
`
`k2
`
`/;
`
`A E _L v -
`
`A E —L v
`
`POSITION
`
`FIG.11 j
`
`GRAY
`LEVEL
`‘ SHIFT
`
`k1
`
`I
`
`-:
`
`k2 "
`(w2)i _
`
`§(wz§ GRAY LEVEL
`
`(g1)
`
`'
`
`(g2)
`
`

`
`Patent Application Publication May 5, 2005 Sheet 9 0f 11
`
`US 2005/0093798 A1
`
`FIG.12A
`
`else
`
`end if
`
`k=kl
`
`FIG.12B
`
`else
`
`end if
`
`k=k + k2
`
`FlG.12CV ‘
`
`else
`
`end if
`
`k=k + k2
`
`FIG.12D Output Gray Scale = Input Gray Scale
`
`+
`
`k
`
`

`
`Patent Application Publication May 5, 2005 Sheet 10 0f 11
`
`US 2005/0093798 A1
`
`
`
`zOFOmmmOO .EmIS V
`
`
`
`
`
`ZOFOmFEOO .tmlm I
`
`
`
`
`
`ZOFOwmEOO .EmQ. .
`
`

`
`Patent Application Publication May 5, 2005 Sheet 11 0f 11
`
`US 2005/0093798 A1
`
`5% is 255
`
`2053382.
`
`E2
`
`93:
`
`S 8
`. I I
`
`.. ><._n_m5 .
`
`.2550 . 0E
`
`wzwwwooma
`
`. :2:
`
`81
`
`9E ‘
`
`.2205
`
`mOmDOw
`
`
`
`zo?owmmoo wo<mo5
`
`4 :z:
`
`<3 1 < K 1
`
`\||\
`
`zoFomEmoo
`
`<P<Q ‘
`
`

`
`US 2005/0093798 A1
`
`May 5, 2005
`
`CORRECTION OF UNEVEN IMAGE
`APPEARANCE BY USE OF SMALL-SIZE DATA
`
`BACKGROUND OF THE INVENTION
`
`[0001] 1. Field of the Invention
`[0002] The present invention generally relates to display
`correction circuits and display apparatuses, and particularly
`relates to a display correction circuit and a display apparatus
`Which correct uneven image appearance caused by the
`characteristics of the display apparatus.
`
`[0003] 2. Description of the Related Art
`[0004] In liquid crystal display apparatuses, plasma dis
`play apparatuses, or the like, uneven image appearance may
`be observed When display brightness becomes darker or
`brighter, for example, than desired brightness at some places
`on the screen. In the liquid crystal display apparatuses, for
`example, such uneven image appearance is caused by varia
`tion in the thickness of liquid crystal display cells, the
`thickness of electrode patterns, etc.
`
`[0005] A circular uneven appearance has a circular shape
`appearing on the screen, and is caused by a locally different
`cell thickness (thinner or thicker) than surrounding areas,
`local abnormality of TFT characteristics, local abnormality
`of electrode pattern siZe, the presence of a pinhole in the
`orientation layer, the inadvertent mixing of tarnishing for
`eign material, etc. A band uneven appearance has a band
`shape appearing on the screen, and is caused by variation in
`the siZe of electrode patterns, variation in the siZe of BM
`patterns, variation in the Way the orientation layer is formed.
`A frame uneven appearance has a frame shape appearing on
`the periphery of the screen, and is caused by a different cell
`thickness around the periphery of the display area. A streak
`uneven appearance has a streak appearing on the screen, and
`is caused by abnormal characteristics of TFT that may be
`present on a bus-line-speci?c basis. A shot uneven appear
`ance has a rectangular shape appearing on the screen, and is
`caused by area variation, line Width variation, positional
`displacement, etc., that take place during stepper exposure.
`
`[0006] In addition to those uneven appearances as
`described above, there are uneven appearances having unde
`?ned shape that is dif?cult to describe in Word. In most
`cases, hoWever, uneven appearance appears as an area
`having a de?ned shape such as a circle, a band, a rectangle,
`a line, a periphery frame, etc. If the density of uneven
`appearance exceeds a spec of the manufactured liquid crys
`tal display apparatus, such apparatus is generally treated as
`a defect product.
`
`[0007] As a method of reducing uneven appearance by use
`of a circuit, information indicative of the shape and density
`of an uneven appearance are stored in memory as mapping
`information, and a liquid crystal display apparatus is con
`trolled based on the stored information to correct the uneven
`appearance (Patent Document 1). Another method includes
`specifying the coordinates of a center, specifying spread
`from the center in four directions, and performing approxi
`mation to obtain correction values (Patent Document 2)
`
`[0008] Patent Document 1: Japanese Patent Application
`Publication No. 9-318929
`
`[0009] Patent Document 2: Japanese Patent Application
`Publication No. 11-113019
`
`[0010] Patent Document 3: Japanese Patent Application
`Publication No. 02-108096
`[0011] In Patent Document 1, the siZe of data becomes
`enormous as the area of uneven appearance increases. For
`example, if uneven appearance occurs in 1/10 the entire
`display area of XGA (1024x768), a large siZe memory
`having approximately a l-Gbit capacity (1024X768X3/10X8X
`2x256) is necessary in order to store the mapping informa
`tion for :8 level correction With respect to each of 256 gray
`levels.
`
`[0012] In Patent Document 2, uneven appearance is
`removed by computing correction data based on the center
`coordinates and spread in the four directions. In reality,
`hoWever, a function for correction needs to be speci?ed With
`respect to the spread of uneven appearance from the center
`in each direction, resulting in a large number of required
`parameters. Further, although this method can remove an
`uneven appearance having a circular or ellipse shape,
`uneven appearances of non-circular shape such as a band
`uneven appearance, a frame uneven appearance, a streak
`uneven appearance, a shot uneven appearance, etc., cannot
`be removed.
`
`[0013] Moreover, 8-bit data cannot properly represent the
`periphery of uneven appearance or the like Where uneven
`ness is extremely thin. According to study conducted by the
`applicants of this application, correction by use of 8-bit data
`ends up generating step-like level changes at the periphery.
`In the conventional liquid crystal display apparatus, gray
`levels are represented by 8 bits (i.e., 256 gray levels). and,
`in some apparatuses for use in notebook computers, 6-bit
`gray levels are used. The related-art methods described
`above give no consideration to the ?neness of gray level
`representation at corrected portions.
`
`[0014] Further, addition of a correction circuit for remov
`ing uneven appearance results in a cost increase of a
`controller IC. It is thus necessary to reduce the siZe of the
`correction circuit as much as possible. The possibility of
`having uneven appearance is rather small relative to a total
`manufacturing quantity. For example, only 0.01% to 1% of
`the total manufacturing quantity are treated as defect units.
`When defect units accounting for 0.1% are to be recovered
`by a cost increase of 50 yen, the SO-yen cost increase is
`applied to all the units including nondefective units. Busi
`ness is not pro?table unless economical loss caused by the
`disposal of a defective unit exceeds 50,000 yen.
`
`[0015] Accordingly, there is a need for a display correction
`circuit and a display apparatus Which can reduce uneven
`appearance by use of small siZe correction data and a simple
`circuit construction.
`
`[0016] There is also a need for a display correction circuit
`and a display apparatus Which can properly reduce uneven
`appearance even at a portion Where the density of uneven
`appearance is loW (thin).
`
`SUMMARY OF THE INVENTION
`
`[0017] It is a general object of the present invention to
`provide a display correction circuit and a display apparatus
`that substantially obviate one or more problems caused by
`the limitations and disadvantages of the related art.
`[0018] Features and advantages of the present invention
`Will be presented in the description Which folloWs, and in
`
`

`
`US 2005/0093798 A1
`
`May 5, 2005
`
`part Will become apparent from the description and the
`accompanying drawings, or may be learned by practice of
`the invention according to the teachings provided in the
`description. Objects as Well as other features and advantages
`of the present invention Will be realized and attained by a
`display correction circuit and a display apparatus particu
`larly pointed out in the speci?cation in such full, clear,
`concise, and exact terms as to enable a person having
`ordinary skill in the art to practice the invention.
`[0019] To achieve these and other advantages in accor
`dance With the purpose of the invention, the invention
`provides a circuit for display correction, including a memory
`Which stores ?rst data indicative of siZe and position of a
`rectangular region on a display screen and second data
`indicative of gray level changes in a surrounding region
`around the rectangular region in an isometric manner With
`respect to a horiZontal direction and a vertical direction, and
`an image processing unit Which adjusts gray levels of image
`data in response to the ?rst data and the second data stored
`in the memory.
`[0020] According to one aspect of the invention, a display
`apparatus includes a memory Which stores ?rst data indica
`tive of siZe and position of a rectangular region on a display
`screen and second data indicative of gray level changes in a
`surrounding region around the rectangular region in an
`isometric manner With respect to a horiZontal direction and
`a vertical direction, an image processing unit Which adjusts
`gray levels of image data in response to the ?rst data and the
`second data stored in the memory, and a display unit Which
`displays the image data having the gray levels thereof
`adjusted that is output from the image processing unit.
`[0021] According to another aspect of the invention, the
`image processing unit as described above adjusts the gray
`levels of the image data by representing the gray levels of
`the image data by use of 9or more bits in at least a portion
`of a display area.
`[0022] In the display correction circuit and the display
`apparatus as described above, uneven appearance is cor
`rected based on the ?rst data indicative of siZe and position
`of a rectangular region and second data indicative of gray
`level changes in a surrounding region around the rectangular
`region in an isometric manner With respect to the horiZontal
`direction and the vertical direction. Accordingly, the siZe of
`the data for correction is small, and a small-siZe circuit for
`simple computation suf?ces.
`[0023] The rectangular region can approach a single point
`by reducing the siZe of the rectangular region. In the extreme
`case, the rectangular region is turned into a single point. In
`such a case, correction is such that its effect decreases
`toWard an outer perimeter Within a circle around the speci
`?ed point. This makes it possible to properly correct a
`circular uneven appearance. Alternatively, a ?nite rectangu
`lar region may be speci?ed, With the Width of the surround
`ing being set to Zero, thereby providing for the correction of
`a rectangular region. This successfully corrects a shot
`uneven appearance. The Width of the rectangular correction
`region may be set substantially equal to one line, providing
`for a streak uneven appearance to be properly corrected. An
`extension from one edge of the screen to an opposite edge
`of the screen may be speci?ed to correct a band uneven
`appearance.
`[0024] Further, the gray levels for correction may be
`provided in 512 levels (i.e., 9-bit representation) to achieve
`
`the representation of ?ne brightness. This makes it possible
`to properly reduce uneven appearance even at a portion
`Where density is loW.
`[0025] Other objects and further features of the present
`invention Will be apparent from the folloWing detailed
`description When read in conjunction With the accompany
`ing draWings.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`[0026] FIG. 1 is a block diagram shoWing an example of
`the construction of a liquid crystal display apparatus accord
`ing to the invention;
`[0027] FIG. 2 is an illustrative draWing for explaining
`correction data for correcting uneven gray levels;
`
`[0028] FIG. 3 is a diagram shoWing the changes of a
`correction value (gray level shift) according to positions;
`[0029] FIG. 4 is a diagram shoWing an example of gray
`level shifts With respect to the respective gray levels of input
`image data to be displayed;
`[0030] FIGS. 5A through 5D are diagrams shoWing an
`example of a correction algorithm according to the inven
`tion;
`[0031] FIG. 6 is a chart shoWing an example of actual
`correction data for use in an SXGA panel;
`
`[0032] FIG. 7 is a block diagram shoWing an example of
`a further detailed construction of an image processing appa
`ratus shoWn in FIG. 1;
`[0033] FIGS. 8A through 8D are diagrams shoWing
`another example of a correction algorithm according to the
`invention;
`[0034] FIGS. 9A through 9E are diagrams shoWing yet
`another example of a correction algorithm according to the
`invention;
`[0035] FIG. 10 is a diagram shoWing another example of
`changes of the correction value (gray level shift) according
`to positions;
`
`[0036] FIG. 11 is a diagram shoWing an example of gray
`level shifts With respect to respective input gray levels in the
`case of correction value settings shoWn in FIG. 10;
`
`[0037] FIGS. 12A through 12D are diagrams shoWing an
`example of a correction algorithm according to the invention
`in the case of FIG. 10 and FIG. 11;
`
`[0038] FIG. 13 is a draWing for explaining an effect of
`gray levels on ideal correction values; and
`
`[0039] FIG. 14 is a draWing shoWing a construction that
`reduces the number of bits through frame modulation after
`correction using a large number of bits.
`
`DESCRIPTION OF THE PREFERRED
`EMBODIMENTS
`
`[0040] In the folloWing, embodiments of the present
`invention Will be described With reference to the accompa
`nying draWings.
`[0041] FIG. 1 is a block diagram shoWing an example of
`the construction of a liquid crystal display apparatus accord
`ing to the invention. Although a description Will be given
`
`

`
`US 2005/0093798 A1
`
`May 5, 2005
`
`here With reference to the liquid crystal display apparatus of
`FIG. 1, it should be noted that the invention is equally
`applicable to other types of display apparatuses such as a
`plasma display apparatus.
`[0042] A liquid crystal display apparatus 10 of FIG. 1
`includes an image processing apparatus 11, a memory 12, a
`signal source 13, and a liquid crystal display panel 14. The
`memory 12 stores correction data for use in the correction of
`uneven appearance. The signal source 13 supplies image
`data signals for display on the liquid crystal display panel
`14. The image processing apparatus 11 corrects the image
`data signals supplied from the signal source 13 based on the
`correction data supplied from the memory 12, thereby
`adjusting the gray levels of the image data signals. The
`image processing apparatus 11 supplies the image data
`signals having their gray levels adjusted to the liquid crystal
`display panel 14. The gray levels of the image data signals
`are adjusted such as to reduce uneven appearance speci?c to
`the liquid crystal display panel 14. This makes it possible to
`display an image With reduced uneven appearance.
`[0043] The image processing apparatus 11 includes a
`correction data storage unit 21, a correction processing unit
`22, and a FIFO 23. The correction data storage unit 21 stores
`the correction data supplied from the correction data storage
`unit 21, and provide the stored data to the correction
`processing unit 22. The FIFO 23 receives the image data
`signals from the signal source 13, and stores a ?xed number
`of data (i.e., display data equal in amount to one frame),
`folloWed by supplying the data to the correction processing
`unit 22 in an order in Which the data is received. The
`correction processing unit 22 corrects the image data signals
`supplied from the FIFO 23 based on the correction data
`supplied from the correction data storage unit 21, thereby
`adjusting the gray levels of the image data signals.
`[0044] FIG. 2 is an illustrative draWing for explaining the
`correction data for correcting uneven gray levels.
`
`[0045] As shoWn in FIG. 2, an area to be corrected is
`speci?ed by tWo points corresponding to a top left corner
`(X1, y1) and a bottom right corner (X2, y2) of a rectangular
`region according to the invention. Within the rectangular
`region de?ned by these tWo points, a constant correction
`value k is applied, for eXample. This correction value
`corresponds to an amount of shift by Which a gray level is
`changed. A surrounding region having a Width W1 is de?ned
`around the rectangular region, and a correction value is
`gradually decreased in this surrounding region. That is, the
`correction value is k at the edge of the rectangular region,
`and decreases in the surrounding region toWard an outer
`edge of the surrounding region until it becomes Zero at
`distance W1 from the edge of the rectangular region.
`[0046] FIG. 3 is a diagram shoWing changes of the
`correction value (gray level shift) according to positions.
`[0047] In FIG. 3, a ?at portion Where the gray level shift
`is constant at k corresponds to the rectangular region shoWn
`in FIG. 2. In an eXample of FIG. 3, the gray level shift
`linearly decreases from k to Zero in the surrounding region
`speci?ed as having the Width W1. In this embodiment, only
`the Width W1 is speci?ed, thereby de?ning gray level
`changes in the surrounding region in an isotropic manner
`With respect to the X direction and the y direction. The
`invention thus has an advantage in the small siZe of correc
`tion data.
`
`[0048] FIG. 4 is a diagram shoWing an eXample of gray
`level shifts With respect to the respective gray levels of the
`input image data to be displayed.
`
`[0049] Uneven appearance becomes conspicuous When
`data to be displayed is halftone. Namely, When the display
`data is close to black (i.e., close to Zero) or close to White
`(i.e., close to 255 in the case of 256 gray levels), there is no
`need for uneven appearance correction. In the eXample of.
`FIG. 4, such characteristics of uneven appearance are taken
`into consideration, so that the correction value is set to k for
`halftones inside a range betWeen a gray level g1 and a gray
`level g2, and decreases as the gray level of interest moves
`aWay from this range. Speci?cally, regions having a Width
`W2 are provided above and beloW the above range, such that
`the correction value linearly decreases from k to Zero in
`these regions having the Width W2.
`
`[0050] FIGS. 5A through 5D are diagrams shoWing an
`eXample of a correction algorithm according to the inven
`tion.
`
`[0051] As shoWn in FIG. 5A, the gray level shift is
`adjusted according to an input gray level. Speci?cally, the
`correction value is set to Zero if an input gray level gs is
`smaller than g1—W2. OtherWise, if gs is smaller than gl, the
`correction value is set to (k) (g1—gs)/W2. This provides the
`correction value that linearly increases as the gray level
`increases. If the input gray level gs is larger than g2+W2, the
`correction value is set to Zero. OtherWise, if gs is larger than
`g2, the correction value is set to (k)(gs—g2)/W2. This pro
`vides the correction value that linearly decreases as the gray
`level increases. In other areas, the correction value is set to
`k.
`
`[0052] In FIG. 5B, the gray level shift is adjusted accord
`ing to position. Speci?cally, the correction value is set to
`Zero if a piXel position X of the input display data is smaller
`than X1—W1. OtherWise, if X is smaller than X1, the correc
`tion value is set to (k)(X1—X)/W1. Here, k is a value of the
`correction value that is adjusted according to the input gray
`level as described With reference to FIG. 5A. This provides
`the correction value that linearly increases in the surround
`ing region around the rectangular region. If the piXel posi
`tion X of the input display data is larger than X2+W1, the
`correction value is set to Zero. OtherWise, if X is larger than
`X2, the correction value is set to (k)(X—X2)/W1. This provides
`the correction value that linearly decreases in the surround
`ing region around the rectangular region. In other areas, the
`correction value is maintained at k. The same adjustment
`process is also performed in the y direction (FIG. SC).
`
`[0053] At the end, as shoWn in FIG. 5D, the correction
`value k obtained in the manner as described above is added
`to the input gray level (Input Gray Scale) to produce an
`output gray level (Output Gray Scale).
`[0054] In the embodiment described above, the rectangu
`lar region can approach a single point by reducing the siZe
`of the rectangular region, Which is situated at the center of
`a corrected region. In the eXtreme case, the top left corner
`(X1, y1) and the bottom right corner (X2, y2) coincide,
`turning the rectangular region into a single point. In such a
`case, correction is such that its effect decreases toWard the
`outer perimeter Within the radius W1. This makes it possible
`to properly correct a circular uneven appearance that Was
`described in the background of the invention.
`
`

`
`US 2005/0093798 A1
`
`May 5, 2005
`
`[0055] The top left corner (x1 y1) and the bottom right
`corner (x2, y2) may be provided as separate points to de?ne
`a rectangular region, and the Width W1 of the surrounding
`region may be set to Zero, providing for the correction of a
`rectangular region. This successfully corrects a shot uneven
`appearance that Was described in the background of the
`invention. The Width of the rectangular correction region
`may be set substantially equal to one line, providing for a
`streak uneven appearance to be properly corrected. An
`extension from one edge of the screen to an opposite edge
`of the screen may be speci?ed to correct a band uneven
`appearance.
`[0056] FIG. 6 is a chart shoWing an example of actual
`correction data for use in an SXGA panel. The SXGA panel
`is comprised of 1280-by-768 pixels. Data for panel correc
`tion in the case of 8-bit image data are shoWn in FIG. 6.
`With respect to uneven appearances having the same circular
`shape, as shoWn in FIG. 6, the correction value k is positive
`for a black uneven appearance (i.e., uneven appearance
`darker than the surrounding). and is negative for a White
`uneven appearance (i.e., uneven appearance brighter than
`the surrounding).
`[0057] FIG. 7 is a block diagram shoWing an example of
`a further detailed construction of the image processing
`apparatus 11 shoWn in FIG. 1.
`[0058] As shoWn in FIG. 7, the image processing appa
`ratus 11 includes the correction data storage unit 21, the
`FIFO 23, a shape correction unit 31, a gray level correction
`unit 32, a multiplying correction unit 33, and an adding and
`subtracting unit 34.
`[0059] In FIG. 7, the image processing apparatus 11 may
`be implemented as an ASIC, for example. The shape cor
`rection unit 31 computes correction coefficients according to
`display coordinates, and, at the same time, the gray level
`correction unit 32 computes correction coef?cients accord
`ing to input signal gray levels. The correction coefficients
`obtained by the shape correction unit 31 and the correction
`coef?cients obtained by the gray level correction unit 32 are
`multiplied by the multiplying correction unit 33, thereby
`producing correction values (i.e., gray level shifts). In the
`program shoWn in FIGS. 5A through 5D, a gray level shift
`responsive to an input gray level is obtained in FIG. 5A, and
`a gray level shift responsive to a pixel position (display
`coordinates) is obtained by multiplication (k=(k responsive
`to the input gray level)><(coef?cient responsive to the display
`position)) in FIGS. 5B and 5C. In FIG. 7, the correction
`coef?cient responsive to the input gray level and the cor
`rection coef?cient responsive to the display position are
`obtained concurrently, and are multiplied to achieve the
`same computation as in FIGS. 5A through 5D.
`[0060] The adding and subtracting unit 34 adds the cor
`rection value obtained in the manner described above to the
`image data signals retrieved from the FIFO 23. This per
`forms the correction of uneven appearance With respect to
`the input display signals. Further, the correction data stored
`in the memory 12 are temporarily stored in the correction
`data storage unit 21, and are then supplied to the shape
`correction unit 31 and the gray level correction unit 32. This
`makes it possible to cope With any types of uneven appear
`ances such as a circular shape, a band shape, a rectangular
`shape, a streak shape, a frame shape, etc.
`[0061] In FIG. 7, the shape correction unit 31 may be
`provided separately from a processing unit 36 that includes
`
`the gray level correction unit 32, the multiplying correction
`unit 33, and the adding and subtracting unit 34. With such a
`construction, the processing unit 36 may be provided as
`many as three, corresponding to respective RGB colors, and
`the single shape correction unit 31 is shared by all the three
`RGB colors. With this provision, circuit siZe can be reduced
`to a minimum siZe that is no more than necessary. In general,
`an uneven appearance. due to a single cause may have
`different densities for respective RGB colors, but is not
`likely to have different shapes for different colors. Accord
`ingly, the unit for computing a shape-related correction
`coef?cient is separately provided for shared use by all the
`colors.
`
`[0062] FIGS. 8A through 8D are diagrams shoWing
`another example of a correction algorithm according to the
`invention.
`
`[0063] In the algorithm of FIG. 5, simple linear compu
`tation is performed by a logic circuit, thereby adjusting a
`correction value according to the gray level of input data.
`Depending on the types of uneven appearances, hoWever,
`density may become higher or loWer With respect to speci?c
`gray levels, resulting in linear approximation failing to
`properly represent the density of uneven appearance. Fur
`ther, circuit-based approximation requires a large number of
`multiplications. Since multiplication computation results in
`a drastic increase in the number of data bits, circuit siZe is
`greatly affected.
`
`[0064] In the embodiment shoWn in FIGS. 8A through
`8D, the computation of a correction value responsive to the
`input gray level is not performed, but instead the correction
`value is retrieved from a lookup table 40 (f(gs) in FIG. 8A)
`stored in memory. Data of the lookup table 40 is a one
`dimensional data array corresponding to respective gray
`levels, so that its data siZe is suf?ciently small so as not to
`give rise to a problem in terms of circuit siZe. Further, there
`is an advantage in that such data can be freely adjusted
`according to the characteristics of uneven appearance. The
`algorithm shoWn in FIGS. 8B through 8D are the same as
`that of FIG. 5B through 5D.
`
`[0065] FIGS. 9A through 9E are diagrams shoWing yet
`another example of a correction algorithm according to the
`invention.
`
`[0066] As shoWn in FIG. 9D, a correction value respon
`sive to a display position is obtained by retrieving data from
`a lookup table 41. In FIG. 9D, f(x) obtains data from the
`lookup table 41 according to a position in the x direction,
`and f(y) obtains data from the lookup table 41 according to
`a position in the y direction. In this manner, this embodiment
`uses only the lookup table 41 to de?ne gray level shifts in the
`surrounding of the rectangular region in an isometric manner
`With respect to the x direction and the y direction. The siZe
`of correction data is thus small.
`
`[0067] With this provision, the siZe of a logic circuit can
`be reduced further. Moreover, since the gray level slope at
`the periphery of uneven appearance can be controlled by use
`of desired correction values, rather than by use of linearly
`approximated correction values, correction is possible even
`With respect to an uneven appearance that has an irregular
`distribution of brightness at the periphery. With the algo
`rithm shoWn in FIGS. 9A through 9E, hoWever, the sur
`rounding region near the corners of the rectangular region
`
`

`
`US 2005/0093798 A1
`
`May 5, 2005
`
`has a straight-line outer boundary as opposed to a round
`(circular) boundary of the previous embodiments. When a
`circular uneven appearance is to be removed, thus, a cor
`rected region becomes an octagonal shape rather than a
`circular shape. HoWever, a difference betWeen an octagonal
`shape and a circular shape does not present a problem in
`appearance if the density of uneven appearance is loW. If a
`smooth curve instead of a straight line is desired at the
`corners, correction data relating to the X and y coordinates
`may be stored in the lookup table exclusively for the corners.
`[0068] FIG. 10 is a diagram shoWing another eXample of
`changes of the correction value (gray level shift) according
`to positions. FIG. 11 is a diagram shoWing an eXample of
`gray level shifts With respect to respective input gray levels
`in the case of correction value settings shoWn in FIG. 10.
`
`[0069] FIG. 10 and FIG. 11 correspond to FIG. 3 and
`FIG. 4, respectively. In FIG. 3, the gray level shift is Zero
`outside the surrounding region de?ned by the Width W1. In
`FIG. 10, on the other hand, the correction value is not Zero
`outside the correction region, but is set to a correction value
`k2 that can be any