Back to results
US 7,609,228 B1Granted
Automatic keystone correction system and method
AssigneePixelworks, Inc.
InventorTianbing Brian Teng
Life of this patent
Estimating the term…
Abstract
The invention relates to an automatic keystone correction system and method. The system includes a projection means to project an image on a projection surface and a means for selecting a maximally distorted corner on the image. And the system includes means for moving the maximally distorted corner from a first to a second position and means for distorting image data responsive to the moving.
Claims30
A keystone correction system, comprising:
- an interface to select a maximally distorted corner on a projected image;
- and
- a keystone controller to distort image data responsive to moving the maximally distorted corner and responsive to an identification of a tilt and a rotation of the system, the keystone controller comprising:
- a plurality of vertical offset registers comprising a column even register (YE), a column odd register (YO), and a column increment (1st order term) register (YI);
- a plurality of vertical registers comprising a line position register (VDN), an increment per line (1st order term) register (VDNI), and an increment per column register (VDNPI);
- a vertical enable register;
- a vertical scalar to vertically scale image data responsive to at least the vertical offset registers, the vertical registers, and the vertical enable register;
- a plurality of horizontal offset registers comprising a row even register (XE), a row odd register (XO), and a row increment (1st order term) register (XI);
- a plurality of horizontal registers comprising a row position register (HDN), an increment per line (1st order term) register (HDNI), an increment per line (2nd order term) register (HDNII), an increment per column (1st order term) register (HDNPI), and an increment per column (2nd order term) register (HDNPII);
- a horizontal enable register;
- and
- a horizontal scalar to horizontally scale the vertically scaled image data responsive to at least the horizontal offset registers, the horizontal registers, and the horizontal enable register.
9 more elements
A method for keystone correction, comprising:
- projecting an image on a projection surface;
- selecting a distorted corner of the projected image;
- and
- a keystone controller predistorting image data responsive to the selected distorted corner, wherein the predistorting comprises:
- a vertical scalar vertically scaling the image data responsive to at least a set of vertical offset registers and a set of vertical registers, wherein the set of vertical offset registers comprises a column even register (YE), a column odd register (YO), and a column increment (1st order term) register (YI), and wherein the set of vertical registers comprises a line position register (VDN), an increment per line (1st order term) register (VDNI), and an increment per column register (VDNPI);
- and
- a horizontal scalar horizontally scaling the vertically scaled image data responsive to at least a set of horizontal offset registers and a set of horizontal registers, wherein the set of horizontal offset registers comprises a row even register (XE), a row odd register (XO), and a row increment (1st order term) register (XI), and wherein the set of horizontal registers comprises a row position register (HDN), an increment per line (1st order term) register (HDNI), an increment per line (2nd order term) register (HDNII), an increment per column (1st order term) register (HDNPI), and an increment per column (2nd order term) register (HDNPII).
4 more elements
A keystone apparatus, comprising:
- means for projecting an image on a projection surface;
- means for selecting a maximally distorted corner on the image;
- means for moving the maximally distorted corner from a first to a second position;
- means for identifying a position of the keystone apparatus in two dimensions;
- means for generating vertically scaled data by vertically scaling image data responsive to at least a set of vertical offset registers and a set of vertical registers, wherein the set of vertical offset registers comprises a column odd register (YO) and a column increment (1st order term) register (YI), and wherein the set of vertical registers comprises a line position register (VDN), an increment per line (1st order term) register (VDNI), and an increment per column register (VDNPI), wherein the means for generating vertically scaled data comprises a first polynomial function f2, wherein the image data corresponds to the image, the first polynomial function f2 comprising the following function:
- f
-
-
- 2
- =
- eqnV
- =
- ∑
- j
- =
- 0
- b
-
-
- (
- VDN
- 2
- 16
- +
- a
- ⨯
- VDNPI
- 2
- 16
- +
- j
- ⨯
- VDNI
- 2
- 16
- )
- +
- (
- YO
- +
- a
- ⨯
- YI
- 2
- 16
- )
- ;
- means for generating horizontally scaled data by horizontally scaling the generated vertically scaled data responsive to at least a set of horizontal offset registers and a set of horizontal registers, wherein the set of horizontal offset registers comprises a row odd register (XO), a row increment (1st order term) register (XI), and a row increment (2nd order term) register (XII), and wherein the set of horizontal registers comprises a row position register (HDN), an increment per line (1st order term) register (HDNI), an increment per line (2nd order term) register (HDNII), an increment per column (1st order term) register (HDNPI), and an increment per column (2nd order term) register (HDNPII), wherein the means for generating horizontally scaled data comprises a second polynomial function f3 comprising the following function:
- f
-
-
- 3
- =
- eqnH
- =
- ∑
- i
- =
- 0
- e
-
-
- (
- HDN
- 2
- 16
- +
- HDNPII
- 2
- ⨯
- 2
- 26
- ⨯
- i
- 2
- +
- (
- HDNPI
- 2
- 16
- +
- HDNPII
- 2
- ⨯
- 2
- 26
- )
- ⨯
- i
- +
- HDNII
- 2
- ⨯
- 2
- 26
- ⨯
- b
- 2
- +
- (
- HDNI
- 2
- 16
- +
- HDNII
- 2
- ⨯
- 2
- 26
- )
- ⨯
- b
- )
- +
- (
- XO
- +
- b
- ⨯
- XI
- 2
- 16
- +
- b
- 2
- ⨯
- XII
- 2
- 19
- )
- ;
- and
- means for distorting the image data responsive to the moving, the identifying, and the generated horizontally scaled data.
130 more elements
Dependencies are inferred from the claim text.