US20070086029A1 - Method for estimating white point shift for auto white balance - Google Patents
Method for estimating white point shift for auto white balance Download PDFInfo
- Publication number
- US20070086029A1 US20070086029A1 US11/562,994 US56299406A US2007086029A1 US 20070086029 A1 US20070086029 A1 US 20070086029A1 US 56299406 A US56299406 A US 56299406A US 2007086029 A1 US2007086029 A1 US 2007086029A1
- Authority
- US
- United States
- Prior art keywords
- white
- color coordinate
- point
- color
- axis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Abandoned
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
- H04N23/84—Camera processing pipelines; Components thereof for processing colour signals
- H04N23/88—Camera processing pipelines; Components thereof for processing colour signals for colour balance, e.g. white-balance circuits or colour temperature control
Definitions
- the invention relates in general to an image color correction technique, and more particularly, to an estimation method of white-point shift for white balance suitable for use to correct white balance of a color image sensing apparatus.
- the image photographic device has become a necessary information appliance of our daily lives. For example, the actual image can be photographed and replayed later.
- the highly developed digital image photographic technique that records the image in pixel units, and stores and displays the image in digital form has advanced the popularity of digital cameras or digital camcorders as the color image is the information that we cannot avoid for our lives.
- a color image is basically mixed by three primary colors, that is, red (R), green (G) and blue (B). Regardless the blackbody radiation factor, a white point is formed of a mixture of isometric R, G and B lights as shown in FIG. 1 . When R, G and B lights are weighted differently, different color is produced. In FIG. 1 , the line between the white point ( 1 , 1 , 1 ) and the black point ( 0 , 0 , 0 ) indicates the gray level.
- the white point is deviated from that as illustrated in FIG. 1 .
- the hardware design of the sensing device has various color responses. Further, under different light sources such as natural light source or artificial light source, different color responses are resulted under the influence of color temperature. For example, when the color temperature of the light source is as low as about 2850° K, the color approaches to red and white. In contrast, when the color temperature of the light source is increased up to about 8500° K, the color tends to be blue. The color differences affect the image color fidelity.
- colors have been illustrated according to the characteristics of a variety of color coordinates.
- transformation matrix is readily applied between other commonly seen color coordinates such as CYM, CYMK, HIS, HSV, YCbCr, YUV and YIQ coordinates.
- the YCbCr coordinate (Y is the brightness component, Cb is the blue component, Cr is the red component) is typically used to perform color analysis because a better compression effect can be obtained for typical JPEG pictures by the YCbCr illustration.
- the white point is deflected into a curved surface in accordance with the variation of color temperature. For a particular brightness, the white point varies along a curve. For different brightness, the curvature of the curve is different. Therefore, the white point thus constructs an irregular curved surface in the YCbCr coordinated.
- the present invention further provides a method for estimating white balance point shift.
- the estimation of white balance point is simplified by using the physical characteristic of the YIQ coordinate. Further, the present invention is not affected by the ambient color temperature.
- the estimation method provided by the present invention includes the following steps.
- a white-point plane of a YIQ color coordinate is determined.
- the normal direction of the white-point plane is inclined from one of a YQ plane and the YI plane of the YIQ color coordinate by a first angle.
- the normal direction is inclined from an original Y-axis of the YIQ coordinate by a second angle.
- a coordinate rotation computation is performed to obtain a rotated YIQ color coordinate.
- the normal direction overlaps with the Y-axis of the rotated YIQ color coordinate; and consequently, the white point plane is perpendicular to the Y-axis of the rotated YIQ color coordinate.
- a Y-axis white-point value is obtained, and a YIQ image data illustrated by the YIQ color coordinate is thus obtained.
- the same rotation computation is performed to rotate the YIQ image data to the rotated YIQ color coordinate, such that a Y-axis image value is obtained.
- a difference between the Y-axis value and the Y-axis white-point value is calculated to estimate the white point shift.
- the coordination rotation computation includes perform a offset of an IQ plane of the YIQ color coordinate.
- the above estimation method further comprises a step of transform an image of an RGB color coordinate or a CYM color coordinate into the YIQ image data illustrated by the YIQ color coordinate.
- the method for estimating white balance point shift includes determining a white-point plane of a YIQ color coordinate.
- a rotation operation is performed on the YIQ color coordinate, such that a normal direction of the white-point plane is pointing an axis of the rotated YIQ color coordinate.
- the white-point plane is perpendicular to the axis with a one-dimensional white-point value of the axis.
- a YIQ image data illustrated by the YIQ color coordinate can be obtained.
- the same rotation operation is performed to rotate the YIQ image data to obtain a one-dimensional image value.
- the difference between the one-dimensional white-point value and the one-dimensional image value are then obtained to estimate the white point shift.
- the axis includes the Y-axis of the rotated YIQ color coordinate.
- FIG. 1 shows an RGB color coordinate
- FIG. 2 shows a YCbCr color coordinate and the white-point variation in response to color temperature on the YCbCr plane
- FIG. 3A shows a white-point plane in a YIQ color coordinate
- FIG. 3B shows the offset of the white-point plane along Y-axis as shown in FIG. 3A ;
- FIGS. 4A and 4B shows the mechanism for performing two coordinate rotation according to the normal direction of the white-point plane
- FIG. 5 shows the position of the white-point plane in the rotated coordinate
- FIG. 6 shows the position of white-point plane relative the Y-axis of the rotated coordinate.
- FIG. 7 shows a flow chart regarding with an estimation method of an embodiment of the present invention.
- the present invention provides a simplified method for estimating the white balance point according to the physical characteristics of the white balance point in the YIQ color coordinate. Thereby, the error caused by the color temperature is minimized.
- the estimation method provided by the present invention includes the following steps, as shown in FIG. 7 .
- a white-point plane of a YIQ color coordinate is determined.
- the normal direction of the white-point plane is inclined from one of a YQ plane and the YI plane of the YIQ color coordinate by a first angle.
- the normal direction is inclined from an original Y-axis of the YIQ coordinate by a second angle.
- a coordinate rotation computation is performed to obtain a rotated YIQ color coordinate.
- the normal direction overlaps with the Y-axis of the rotated YIQ color coordinate; and consequently, the white point plane is perpendicular to the Y-axis of the rotated YIQ color coordinate.
- a Y-axis white-point value is obtained, and a YIQ image data illustrated by the YIQ color coordinate is thus obtained.
- the same rotation computation is performed to rotate the YIQ image data to the rotated YIQ color coordinate, such that a Y-axis image value is obtained.
- a difference between the Y-axis value and the Y-axis white-point value is calculated to estimate the white point shift.
- the YIQ color coordinate has three axes representing three components of colors, where Y indicates the brightness component, I indicates the in-phase component, and Q indicates the quadrature component.
- the YUV color coordinate is similar to the YIQ color coordinate.
- the characteristic illustrated as follows can also be applied to the YUV color coordinate. It is known that a variety of color coordinates are available to illustrate an image, while transformations between the color coordinates are readily applicable.
- the present invention exemplarily provides the estimation methods for YIQ/YUV color coordinates, while the present invention can also be applied to other color coordinates by transforming image illustrated in other color coordinates into to the YIQ/YUV color coordinates.
- a white-point plane 100 can be constructed in the YIQ color coordinate for various color temperatures.
- the color temperature varies along a straight line of color temperature axis.
- a white-point gray line 102 is constructed by the white points.
- These white-point gray lines thus construct the white-point plane 100 .
- the white-point plane 100 has a normal direction 104 perpendicular to the white-point plane 100 .
- Either the white-point plane 100 or the normal direction 104 is inclined in the YIQ color coordinate.
- the inclination is determined by the characteristic parameters and operation conditions of the image sensing device, and can be determined by conventional measurements.
- the inclination of the white-point plane 100 is required, while the way to determine the inclination is not limited to any specific method.
- the normal direction 104 of the white-point plane 100 is inclined from the three axes Y, I, and Q of the YIQ color coordinate by respective angles. It is found in the present invention that the white-point plane 100 as a function of color temperature can be represented by a mathematic formula in the YIQ color coordinate. Therefore, when a color image point such as a pixel or an area of the image is to be calibrated, and the color image point deviates from the white-point plane 100 by a distance, the calibration level and calibration mode can be estimated according to the magnitude and direction of the distance. For example, a weight can be determined in response to the distance to perform statistics or analysis.
- the present invention uses the YIQ color coordinate to illustrate color, so as to obtain a white-point plane 100 .
- the estimation of the white-point shift for the color is simplified into calculation of distance between the color point to be analyzed and the white-point plane 100 .
- an appropriate weight is determined to perform analysis.
- the distance to the white-point plane 100 is calculated in three dimensions. Therefore, simplification of the computation method is further provided by the present invention.
- the white-point plane 100 has a normal direction 104 , which is inclined by an angle ⁇ and an angle ⁇ in a polar coordinate in which the Y-axis can be represented by the Z-axis.
- any of the Y, I, and Q axes can be used as the Z-axis in the polar coordinate.
- the presentation in FIG. 3A is only an exemplary option.
- the white-point plane 100 can be displaced towards the negative direction by a distance d to avoid numerical overflow.
- FIGS. 4A and 4B shows the method to obtain the distance between the color point to be analyzed and the white-point plane 100 .
- a rotation operation is performed to adjust the normal direction 104 and the Y-axis into the same direction.
- rotation operation can be achieved by Euler angle rotation.
- the Y-axis is rotated by an angle ⁇ , and the rotated coordinated is denoted as Y′I′Q′ color coordinate.
- the I′-axis of the Y′I′Q′ color coordinate is further rotated by an angle ⁇ as the Y′′I′′Q′′ color coordinate.
- the Y′′-axis can be further rotated.
- the rotation is performed for obtaining the distance between the color point and the white-point plane, such that the third rotation is not required.
- the normal direction 104 is aligned with the Y′′-axis, such that the white-point plane 100 is perpendicular to the Y′′-axis.
- the white-point plane 100 ′ is parallel to the I′′Q′′ plane, where the white-point plane 100 is located at the d′ position 500 of the Y′′ axis.
- FIG. 3B as the white-plane plane 100 ′ has been displaced along the Y-axis by a distance d, which is then converted into distance d′′ in FIG. 5 .
- the distance between a color point and the white-point plane 100 ′ in the Y′′I′′Q′′ color coordinate can be calculated only by comparing the Y′′-axis value and d′.
- the distance between the color point to be analyzed and calibrated and the white-point plane 100 ′ is either ⁇ d 1 or ⁇ d 2 .
- the magnitude of offset and the shift, a weighting value can be determined. When there is no offset, d′ is equal to zero.
- the offset d is set as the origin 128 .
- the shift of the white point can be easily computed. Therefore, only a comparator is required in hardware design, and the circuit design for automatic color calibration can be greatly simplified.
- the present invention uses the special physical characteristic of white balance point in the YIQ/YUV color coordinate to construct a white point plane as a function of color temperature.
- the white point shift can thus be simplified to the computation of distance between the color point and the white point.
- the rotation operation converts the three-dimensional distance computation into a one-dimensional computation along the Y′′-axis. Therefore, the hardware design is simplified, while the accurate automation calibration is achieved.
- the present invention may be performed by rotating the I- or Q-axis into the Z-axis with the same rotation operation. Further, the sequence of rotation is also adjustable. The purpose is to rotate the normal direction of the white-point plane aligned with the Z-axis in the polar coordinate.
Abstract
A method of estimating shift of white balance point. According to characteristics of a photograph sensing apparatus, a white-point plane of a YIQ color coordinate is determined. The white-point plane has a normal direction pointing one of the axis of the YIQ color coordinate. The white-point plane is thus perpendicular to such axis with a one-dimensional one-spot value. Consequently, a YIQ image data described by the YIQ color coordinate is obtained. The same rotation operation is performed to rotate the YIQ image data to obtain a one-dimensional image value. The one-dimensional image value is compared to the one-dimensional white-point value to obtain a differential value, so as to estimate the white-point shift. In the above method, the axis of the YIQ color coordinate includes the Y-axis after rotation.
Description
- This is a continuation application of patent application Ser. No. 10/248,373, filed on Jan., 15, 2003, which claims the priority benefit of Taiwan patent application serial no. 91133111, filed on Nov. 12, 2002 and is now allowed. The entirety of each of the above-mentioned patent applications is hereby incorporated by reference herein and made a part of this specification.
- 1. Field of the Invention
- The invention relates in general to an image color correction technique, and more particularly, to an estimation method of white-point shift for white balance suitable for use to correct white balance of a color image sensing apparatus.
- 2. Related Art of the Invention
- Along with the development of image processing technique, sensing photographic technique and the variation of information communication, the image photographic device has become a necessary information appliance of our daily lives. For example, the actual image can be photographed and replayed later. Lately, the highly developed digital image photographic technique that records the image in pixel units, and stores and displays the image in digital form has advanced the popularity of digital cameras or digital camcorders as the color image is the information that we cannot avoid for our lives.
- A color image is basically mixed by three primary colors, that is, red (R), green (G) and blue (B). Regardless the blackbody radiation factor, a white point is formed of a mixture of isometric R, G and B lights as shown in
FIG. 1 . When R, G and B lights are weighted differently, different color is produced. InFIG. 1 , the line between the white point (1,1,1) and the black point (0,0,0) indicates the gray level. - When the characteristic of actual image sensing device is taken into account, the white point is deviated from that as illustrated in
FIG. 1 . The hardware design of the sensing device has various color responses. Further, under different light sources such as natural light source or artificial light source, different color responses are resulted under the influence of color temperature. For example, when the color temperature of the light source is as low as about 2850° K, the color approaches to red and white. In contrast, when the color temperature of the light source is increased up to about 8500° K, the color tends to be blue. The color differences affect the image color fidelity. - Therefore, how to locate the white point under various color temperatures and how to estimate tone and white balance point shifts are essential for color calibration.
- Based on the color principles, colors have been illustrated according to the characteristics of a variety of color coordinates. Currently, in addition to the RGB coordinate, transformation matrix is readily applied between other commonly seen color coordinates such as CYM, CYMK, HIS, HSV, YCbCr, YUV and YIQ coordinates.
- To resolve the problem of white balance, the YCbCr coordinate (Y is the brightness component, Cb is the blue component, Cr is the red component) is typically used to perform color analysis because a better compression effect can be obtained for typical JPEG pictures by the YCbCr illustration. As shown in
FIG. 2 , the white point is deflected into a curved surface in accordance with the variation of color temperature. For a particular brightness, the white point varies along a curve. For different brightness, the curvature of the curve is different. Therefore, the white point thus constructs an irregular curved surface in the YCbCr coordinated. However, it is difficult to transform the curved surface of white point in the YCbCr coordinate into a mathematic formula. - The present invention further provides a method for estimating white balance point shift. The estimation of white balance point is simplified by using the physical characteristic of the YIQ coordinate. Further, the present invention is not affected by the ambient color temperature.
- The estimation method provided by the present invention includes the following steps. According to the design condition of a photographic device, a white-point plane of a YIQ color coordinate is determined. The normal direction of the white-point plane is inclined from one of a YQ plane and the YI plane of the YIQ color coordinate by a first angle. The normal direction is inclined from an original Y-axis of the YIQ coordinate by a second angle. According to the first and second angles, a coordinate rotation computation is performed to obtain a rotated YIQ color coordinate. Thereby, the normal direction overlaps with the Y-axis of the rotated YIQ color coordinate; and consequently, the white point plane is perpendicular to the Y-axis of the rotated YIQ color coordinate. A Y-axis white-point value is obtained, and a YIQ image data illustrated by the YIQ color coordinate is thus obtained. According to the first and second angles, the same rotation computation is performed to rotate the YIQ image data to the rotated YIQ color coordinate, such that a Y-axis image value is obtained. A difference between the Y-axis value and the Y-axis white-point value is calculated to estimate the white point shift.
- In the above estimation method, the coordination rotation computation includes perform a offset of an IQ plane of the YIQ color coordinate.
- The above estimation method further comprises a step of transform an image of an RGB color coordinate or a CYM color coordinate into the YIQ image data illustrated by the YIQ color coordinate.
- In one embodiment of the present invention, the method for estimating white balance point shift includes determining a white-point plane of a YIQ color coordinate. A rotation operation is performed on the YIQ color coordinate, such that a normal direction of the white-point plane is pointing an axis of the rotated YIQ color coordinate. Meanwhile, the white-point plane is perpendicular to the axis with a one-dimensional white-point value of the axis. Meanwhile, a YIQ image data illustrated by the YIQ color coordinate can be obtained. The same rotation operation is performed to rotate the YIQ image data to obtain a one-dimensional image value. The difference between the one-dimensional white-point value and the one-dimensional image value are then obtained to estimate the white point shift. In the above estimation method, the axis includes the Y-axis of the rotated YIQ color coordinate.
- These, as well as other features of the present invention, will become more apparent upon reference to the drawings wherein:
-
FIG. 1 shows an RGB color coordinate; -
FIG. 2 shows a YCbCr color coordinate and the white-point variation in response to color temperature on the YCbCr plane; -
FIG. 3A shows a white-point plane in a YIQ color coordinate; -
FIG. 3B shows the offset of the white-point plane along Y-axis as shown inFIG. 3A ; -
FIGS. 4A and 4B shows the mechanism for performing two coordinate rotation according to the normal direction of the white-point plane; -
FIG. 5 shows the position of the white-point plane in the rotated coordinate; and -
FIG. 6 shows the position of white-point plane relative the Y-axis of the rotated coordinate. -
FIG. 7 shows a flow chart regarding with an estimation method of an embodiment of the present invention. - The present invention provides a simplified method for estimating the white balance point according to the physical characteristics of the white balance point in the YIQ color coordinate. Thereby, the error caused by the color temperature is minimized.
- The estimation method provided by the present invention includes the following steps, as shown in
FIG. 7 . According to the design condition of a photographic device, a white-point plane of a YIQ color coordinate is determined. The normal direction of the white-point plane is inclined from one of a YQ plane and the YI plane of the YIQ color coordinate by a first angle. The normal direction is inclined from an original Y-axis of the YIQ coordinate by a second angle. According to the first and second angles, a coordinate rotation computation is performed to obtain a rotated YIQ color coordinate. Thereby, the normal direction overlaps with the Y-axis of the rotated YIQ color coordinate; and consequently, the white point plane is perpendicular to the Y-axis of the rotated YIQ color coordinate. A Y-axis white-point value is obtained, and a YIQ image data illustrated by the YIQ color coordinate is thus obtained. According to the first and second angles, the same rotation computation is performed to rotate the YIQ image data to the rotated YIQ color coordinate, such that a Y-axis image value is obtained. A difference between the Y-axis value and the Y-axis white-point value is calculated to estimate the white point shift. - Currently, an effective way to estimate the shift between the color to be calibrated and the white balance point has not been developed. Therefore, in the conventional color calibration method, errors caused by the variation of color temperature often occur to the calibration result.
- In the present invention, various color spaces or color coordinates have been investigated to find out that a specific physical characteristic of the YIQ color coordinate is particular useful for estimating the white balance point shift. The YIQ color coordinate has three axes representing three components of colors, where Y indicates the brightness component, I indicates the in-phase component, and Q indicates the quadrature component. The transformation relationship between YIQ color coordinate and the RGB color coordinate is expressed as:
- Among other color coordinates, the YUV color coordinate is similar to the YIQ color coordinate. The difference between the YUV color coordinate and the YIQ color coordinated can be expressed as:
I=0.74V−0.27U
Q=0.48V+0.41U - Therefore, the characteristic illustrated as follows can also be applied to the YUV color coordinate. It is known that a variety of color coordinates are available to illustrate an image, while transformations between the color coordinates are readily applicable. The present invention exemplarily provides the estimation methods for YIQ/YUV color coordinates, while the present invention can also be applied to other color coordinates by transforming image illustrated in other color coordinates into to the YIQ/YUV color coordinates.
- Based on the research result, it is found that white point varies as color temperature and brightness change. As shown in
FIG. 3 , a white-point plane 100 can be constructed in the YIQ color coordinate for various color temperatures. In the YIQ color coordinate, the color temperature varies along a straight line of color temperature axis. For a particular color temperature such as T2, a white-point gray line 102 is constructed by the white points. These white-point gray lines thus construct the white-point plane 100. The white-point plane 100 has anormal direction 104 perpendicular to the white-point plane 100. - Either the white-
point plane 100 or thenormal direction 104 is inclined in the YIQ color coordinate. The inclination is determined by the characteristic parameters and operation conditions of the image sensing device, and can be determined by conventional measurements. In this present invention, the inclination of the white-point plane 100 is required, while the way to determine the inclination is not limited to any specific method. - The
normal direction 104 of the white-point plane 100 is inclined from the three axes Y, I, and Q of the YIQ color coordinate by respective angles. It is found in the present invention that the white-point plane 100 as a function of color temperature can be represented by a mathematic formula in the YIQ color coordinate. Therefore, when a color image point such as a pixel or an area of the image is to be calibrated, and the color image point deviates from the white-point plane 100 by a distance, the calibration level and calibration mode can be estimated according to the magnitude and direction of the distance. For example, a weight can be determined in response to the distance to perform statistics or analysis. - Therefore, the present invention uses the YIQ color coordinate to illustrate color, so as to obtain a white-
point plane 100. Thereby, the estimation of the white-point shift for the color is simplified into calculation of distance between the color point to be analyzed and the white-point plane 100. Thereafter, according to the color calibration mode, an appropriate weight is determined to perform analysis. However, the distance to the white-point plane 100 is calculated in three dimensions. Therefore, simplification of the computation method is further provided by the present invention. - As shown in
FIG. 3A , the white-point plane 100 has anormal direction 104, which is inclined by an angle α and an angle β in a polar coordinate in which the Y-axis can be represented by the Z-axis. In fact, any of the Y, I, and Q axes can be used as the Z-axis in the polar coordinate. The presentation inFIG. 3A is only an exemplary option. In addition, as shown inFIG. 3B , as the actual physical characteristic indicated by the Y-axis is positive, the white-point plane 100 can be displaced towards the negative direction by a distance d to avoid numerical overflow. Theses are all design choices without affecting the estimation of relative deviation of the white-point plane. -
FIGS. 4A and 4B shows the method to obtain the distance between the color point to be analyzed and the white-point plane 100. As shown inFIGS. 4A and 4B , a rotation operation is performed to adjust thenormal direction 104 and the Y-axis into the same direction. In mathematics, such rotation operation can be achieved by Euler angle rotation. InFIG. 4A , the Y-axis is rotated by an angle α, and the rotated coordinated is denoted as Y′I′Q′ color coordinate. InFIG. 4B , the I′-axis of the Y′I′Q′ color coordinate is further rotated by an angle β as the Y″I″Q″ color coordinate. For certain specific requirement or condition such as to perform transformation between the YIQ and YUV color coordinates, the Y″-axis can be further rotated. However, in the present invention, the rotation is performed for obtaining the distance between the color point and the white-point plane, such that the third rotation is not required. - By performing the rotations as shown in
FIGS. 4A and 4B , thenormal direction 104 is aligned with the Y″-axis, such that the white-point plane 100 is perpendicular to the Y″-axis. Referring toFIG. 5 , the white-point plane 100′ is parallel to the I″Q″ plane, where the white-point plane 100 is located at the d′position 500 of the Y″ axis. Referring toFIG. 3B , as the white-plane plane 100′ has been displaced along the Y-axis by a distance d, which is then converted into distance d″ inFIG. 5 . If d is zero, d′ is zero, and d′ is the position of the white-point plane 100 on the Y″-axis. Therefore, the distance between a color point and the white-point plane 100′ in the Y″I″Q″ color coordinate can be calculated only by comparing the Y″-axis value and d′. The distance between the color point to be analyzed and calibrated and the white-point plane 100′ is either Δd1 or Δd2. According to the color calibration mode, the magnitude of offset and the shift, a weighting value can be determined. When there is no offset, d′ is equal to zero. - In the actual operation, the offset d is set as the origin 128. The actual YIQ numeral range is Y=[0,255], I=[−128,127] and Q=[−128,127]. According to equation (1), one can obtain
Y=(77R+150G+29B)/256
I=(128R−59G−69B)/256
Q=(52R−128G+76B)/256 (3),
Once the RGB color coordinate is transformed into the YIQ color coordinate by the transformation in equation (3), a first rotation can be performed on the Y-axis as:
Y′=Y−128
I′=I cosθ+Q sinθ
Q′=−I sinθ+Q cosθ (4)
The Q-axis can be rotated further to compensate the offset d=128 as:
Y″=Y′ cosθ+I′ sinθ+128
I″=−Y′ sinθ+I′ cosθ - After the rotations, the shift of the white point can be easily computed. Therefore, only a comparator is required in hardware design, and the circuit design for automatic color calibration can be greatly simplified.
- The present invention uses the special physical characteristic of white balance point in the YIQ/YUV color coordinate to construct a white point plane as a function of color temperature. The white point shift can thus be simplified to the computation of distance between the color point and the white point. Further, the rotation operation converts the three-dimensional distance computation into a one-dimensional computation along the Y″-axis. Therefore, the hardware design is simplified, while the accurate automation calibration is achieved. It is appreciated that the present invention may be performed by rotating the I- or Q-axis into the Z-axis with the same rotation operation. Further, the sequence of rotation is also adjustable. The purpose is to rotate the normal direction of the white-point plane aligned with the Z-axis in the polar coordinate.
- Other embodiments of the invention will appear to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. It is intended that the specification and examples to be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
Claims (11)
1. A method of estimating white point shift, comprising:
determining a white-point plane in a color coordinate according to an image sensing device, wherein the white-point plane has a normal direction inclined from a plane of the color coordinate by a first angle, and the normal direction is inclined from an Y-axis of the color coordinate by a second angle;
performing a rotation operation on the color coordinate according to the first and second angles to align the normal direction with the Y-axis, so as to obtain a Y-axis white point value;
obtaining an image data in the rotated color coordinate;
performing the rotation operation on the image data according to the first and second angles to transform the image data into the rotated color coordinate to obtain a Y-axis image value; and
obtaining a difference between the Y-axis white point value and the Y-axis image value to estimate the white point shift.
2. The method according to claim 1 , wherein the step of performing a rotation operation on the color coordinate further comprises displace the Y-axis thereof.
3. The method according to claim 1 , where the step of obtaining the difference further comprises determining a weighting value according to the difference and a calibration mode.
4. The method according to claim 1 , further comprising transforming an image in the RGB color coordinate into the image data in the color coordinate.
5. The method according to claim 1 , further comprising transforming an image in the YCM color coordinate into the image data in the color coordinate.
6. A method of estimating white point shift, comprising:
selecting a first color coordinate in which a plurality of white points are located on a single plane;
rotating the first color coordinate until the plane of white points is perpendicular to one selected axis of the first color coordinate, such that the white points have a substantially identical coordinate value of the selected axis;
obtaining a coordinate value of the selected axis of a color point to be calibrated; and
obtaining the white point shift by a difference between the coordinate values of the white points and the color point of the selected axis.
7. The method according to claim 6 , wherein the first color coordinate includes a YIQ color coordinate or a YUV color coordinate.
8. The method according to claim 6 , wherein the step of obtaining a coordinate value of the selected axis of a color point to be calibrated further comprising:
obtaining an image including a plurality of color points in the first color coordinate; and
rotation the first color coordinate until a plane of the color points is perpendicular to the selected axis.
9. The method according to claim 8 , wherein the step of obtaining an image further comprises:
obtaining an image in a second color coordinate different from the first color coordinate; and
transforming the image from the second color coordinate to the first color coordinate.
10. A method for estimating white point shift, comprising:
determining a white-point plane in a color coordinate according to characteristic of an image sensing device, the white-point plane having a normal direction;
performing a rotation operation to make the normal direction of the white-point plane perpendicular to one axis axis of the color coordinate, such that the white-point plane is perpendicular to the coordinate with a one-dimensional white-point value;
obtaining an image data in the color coordinate;
performing the rotation operation to rotate the image data to obtain a one-dimensional image value; and
calculating a difference between the one-dimensional white-point value and the one-dimensional image value to estimate the white point shift.
11. The method according to claim 10 , further comprising transforming an image of an RGB color coordinate into a YIQ/YUV image data.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/562,994 US20070086029A1 (en) | 2002-11-12 | 2006-11-23 | Method for estimating white point shift for auto white balance |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW091133111A TW548509B (en) | 2002-11-12 | 2002-11-12 | Estimation method for white-point offset of the white balance point |
TW91133111 | 2002-11-12 | ||
US10/248,373 US7177467B2 (en) | 2002-11-12 | 2003-01-15 | Method for estimating white point shift for auto white balance |
US11/562,994 US20070086029A1 (en) | 2002-11-12 | 2006-11-23 | Method for estimating white point shift for auto white balance |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/248,373 Continuation US7177467B2 (en) | 2002-11-12 | 2003-01-15 | Method for estimating white point shift for auto white balance |
Publications (1)
Publication Number | Publication Date |
---|---|
US20070086029A1 true US20070086029A1 (en) | 2007-04-19 |
Family
ID=29998119
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/248,373 Expired - Fee Related US7177467B2 (en) | 2002-11-12 | 2003-01-15 | Method for estimating white point shift for auto white balance |
US11/562,994 Abandoned US20070086029A1 (en) | 2002-11-12 | 2006-11-23 | Method for estimating white point shift for auto white balance |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/248,373 Expired - Fee Related US7177467B2 (en) | 2002-11-12 | 2003-01-15 | Method for estimating white point shift for auto white balance |
Country Status (3)
Country | Link |
---|---|
US (2) | US7177467B2 (en) |
JP (1) | JP3748862B2 (en) |
TW (1) | TW548509B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090021526A1 (en) * | 2007-07-20 | 2009-01-22 | Novatek Microelectronics Corp. | Determination method for white-point and correction method of the white balance |
US20110268356A1 (en) * | 2009-01-23 | 2011-11-03 | Masato Tsukada | Image processing method, image processing apparatus, and recording medium |
US20130050236A1 (en) * | 2011-08-23 | 2013-02-28 | Novatek Microelectronics Corp. | White balance method for display image |
WO2014204865A1 (en) * | 2013-06-17 | 2014-12-24 | Dolby Laboratories Licensing Corporation | Adaptive reshaping for layered coding of enhanced dynamic range signals |
US20170308996A1 (en) * | 2016-02-02 | 2017-10-26 | Dolby Laboratories Licensing Corporation | Single-pass and multi-pass-based polynomial approximations for reshaping functions |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3607654B2 (en) * | 2001-09-14 | 2005-01-05 | 株式会社東芝 | Illumination light color estimation method and imaging device for obtaining white balance |
TW548509B (en) * | 2002-11-12 | 2003-08-21 | Novatek Microelectronics Corp | Estimation method for white-point offset of the white balance point |
US7301543B2 (en) | 2004-04-09 | 2007-11-27 | Clairvoyante, Inc. | Systems and methods for selecting a white point for image displays |
US20050280881A1 (en) * | 2004-06-18 | 2005-12-22 | Microsoft Corporation | System and method for determination of a white point for calibration of an image capturing device |
US7804525B2 (en) * | 2006-08-24 | 2010-09-28 | Micron Technology, Inc. | Method, apparatus, and system for selecting pixels for automatic white balance processing |
US7974487B2 (en) * | 2007-04-17 | 2011-07-05 | Kabushiki Kaisha Toshiba | System and method for image white balance adjustment |
US20090046171A1 (en) * | 2007-08-16 | 2009-02-19 | C2Cure, Inc. | Non-linear color correction |
CN105026903A (en) * | 2012-11-02 | 2015-11-04 | 变量公司 | Computer-implemented system and method for color sensing, storage and comparison |
US10742850B2 (en) * | 2017-04-17 | 2020-08-11 | Canon Kabushiki Kaisha | Image processing apparatus and control method thereof for white balance adjustment |
CN108282647B (en) * | 2018-01-31 | 2019-11-05 | 上海小蚁科技有限公司 | Color correcting method and device, computer readable storage medium, terminal |
US10746599B2 (en) | 2018-10-30 | 2020-08-18 | Variable, Inc. | System and method for spectral interpolation using multiple illumination sources |
CN111565306B (en) * | 2020-03-27 | 2022-05-24 | 瑞芯微电子股份有限公司 | Automatic white balance method and storage device for three-dimensional space white point detection and brightness weighting white point |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6417892B1 (en) * | 1995-05-23 | 2002-07-09 | Colorlink, Inc. | Color filters, sequencers and displays using color selective light modulators |
US6882384B1 (en) * | 1995-05-23 | 2005-04-19 | Colorlink, Inc. | Color filters and sequencers using color selective light modulators |
US7177467B2 (en) * | 2002-11-12 | 2007-02-13 | Novatek Microelectronics Corp. | Method for estimating white point shift for auto white balance |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE3371693D1 (en) * | 1983-12-14 | 1987-06-25 | Hell Rudolf Dr Ing Gmbh | Method of and circuit arrangement for the recognition of chrominances and colours |
US7009733B2 (en) * | 2001-07-02 | 2006-03-07 | Coral Corporation | Manual correction of an image color |
US6868179B2 (en) * | 2001-07-06 | 2005-03-15 | Jasc Software, Inc. | Automatic saturation adjustment |
JP3674568B2 (en) * | 2001-10-02 | 2005-07-20 | ソニー株式会社 | Intensity modulation method and system, and light quantity modulation device |
-
2002
- 2002-11-12 TW TW091133111A patent/TW548509B/en not_active IP Right Cessation
-
2003
- 2003-01-15 US US10/248,373 patent/US7177467B2/en not_active Expired - Fee Related
- 2003-02-20 JP JP2003041953A patent/JP3748862B2/en not_active Expired - Fee Related
-
2006
- 2006-11-23 US US11/562,994 patent/US20070086029A1/en not_active Abandoned
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6417892B1 (en) * | 1995-05-23 | 2002-07-09 | Colorlink, Inc. | Color filters, sequencers and displays using color selective light modulators |
US6667784B2 (en) * | 1995-05-23 | 2003-12-23 | Colorlink, Inc. | Color filters, sequencers and displays using color selective light modulators |
US6882384B1 (en) * | 1995-05-23 | 2005-04-19 | Colorlink, Inc. | Color filters and sequencers using color selective light modulators |
US7177467B2 (en) * | 2002-11-12 | 2007-02-13 | Novatek Microelectronics Corp. | Method for estimating white point shift for auto white balance |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090021526A1 (en) * | 2007-07-20 | 2009-01-22 | Novatek Microelectronics Corp. | Determination method for white-point and correction method of the white balance |
US20110268356A1 (en) * | 2009-01-23 | 2011-11-03 | Masato Tsukada | Image processing method, image processing apparatus, and recording medium |
US8731288B2 (en) * | 2009-01-23 | 2014-05-20 | Nec Corporation | Image processing techniques for color correction |
US20130050236A1 (en) * | 2011-08-23 | 2013-02-28 | Novatek Microelectronics Corp. | White balance method for display image |
US8723880B2 (en) * | 2011-08-23 | 2014-05-13 | Novatek Microelectronics Corp. | White balance method for display image |
WO2014204865A1 (en) * | 2013-06-17 | 2014-12-24 | Dolby Laboratories Licensing Corporation | Adaptive reshaping for layered coding of enhanced dynamic range signals |
US9584811B2 (en) | 2013-06-17 | 2017-02-28 | Dolby Laboratories Licensing Corporation | Adaptive reshaping for layered coding of enhanced dynamic range signals |
US9716887B2 (en) | 2013-06-17 | 2017-07-25 | Dolby Laboratories Licensing Corporation | Chromaticity translations in coding of enhanced dynamic range signals |
US20170308996A1 (en) * | 2016-02-02 | 2017-10-26 | Dolby Laboratories Licensing Corporation | Single-pass and multi-pass-based polynomial approximations for reshaping functions |
US10223774B2 (en) * | 2016-02-02 | 2019-03-05 | Dolby Laboratories Licensing Corporation | Single-pass and multi-pass-based polynomial approximations for reshaping functions |
Also Published As
Publication number | Publication date |
---|---|
TW548509B (en) | 2003-08-21 |
JP3748862B2 (en) | 2006-02-22 |
US7177467B2 (en) | 2007-02-13 |
TW200407660A (en) | 2004-05-16 |
US20040090536A1 (en) | 2004-05-13 |
JP2004166182A (en) | 2004-06-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20070086029A1 (en) | Method for estimating white point shift for auto white balance | |
Ng et al. | Using geometry invariants for camera response function estimation | |
US7791649B2 (en) | Apparatus, medium, and method with automatic white balance control | |
US8249390B2 (en) | Method for taking panorama mosaic photograph with a portable terminal | |
Chakrabarti et al. | An Empirical Camera Model for Internet Color Vision. | |
US6594384B1 (en) | Apparatus and method for estimating and converting illuminant chromaticity using perceived illumination and highlight | |
US7949197B2 (en) | Method of and system for image processing and computer program | |
US20090021526A1 (en) | Determination method for white-point and correction method of the white balance | |
US7362357B2 (en) | Calibration of digital color imagery | |
US20120308126A1 (en) | Corresponding image processing method for compensating colour | |
US8619153B2 (en) | Radiometric calibration using temporal irradiance mixtures | |
US7889275B2 (en) | System and method for continuous flash | |
US7356180B2 (en) | Systems and methods for correcting image color balance | |
CN110213556B (en) | Automatic white balance method and system in monochrome scene, storage medium and terminal | |
CN112218065B (en) | Image white balance method, system, terminal device and storage medium | |
WO2020049779A1 (en) | Color evaluation device, color evaluation method, and indication object used in color evaluation method | |
CN114299156A (en) | Method for calibrating and unifying coordinates of multiple cameras in non-overlapping area | |
US7064769B2 (en) | Systems and methods for computing the presence of self-luminous elements in an image | |
US9030575B2 (en) | Transformations and white point constraint solutions for a novel chromaticity space | |
US8385637B2 (en) | Illuminant estimation | |
US8200007B2 (en) | Illuminant estimation | |
US20230328396A1 (en) | White balance correction method and apparatus, device, and storage medium | |
US20180176528A1 (en) | Light locus generation for automatic white balance | |
CN115426487B (en) | Color correction matrix adjustment method, device, electronic equipment and readable storage medium | |
US7019789B2 (en) | Apparatus and method for calculating color temperature |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STCB | Information on status: application discontinuation |
Free format text: ABANDONED -- FAILURE TO RESPOND TO AN OFFICE ACTION |