System and method for digital image intensity correction
The present invention provides a method, system and image to enhance the image contrast of a digital image device while simultaneously compensating for image intensity inhomogeneity, regardless of the source. The present invention corrects intensity inhomogeneities producing a more uniform image appearance. Also, the image is enhanced through increased contrast, e.g., tissue contrast in a medical image. The method makes no assumptions as to the source of the inhomogeneities, e.g., physical device characteristics or positioning of the object being imaged. In the method, the error between the histogram of the spatially-weighted original image and a specified histogram is minimized. The specified histogram may be selected to increase contrast generally or particularly for accentuation, e.g., on localized regions of interest. The weighting is preferably achieved by two-dimensional interpolation of a sparse grid of control points overlaying the image. A sparse grid is used rather than a dense one to compensate for slowly-varying image non-uniformity. Also, sparseness reduces the computational complexity, as the final weight set involves the solution of simultaneous linear equations whose number is the size of the chosen grid.
1. A method for modifying a digital image, said method comprising:
transforming, using a processor, at least one pixel in pixel data for a digital image of an object,
wherein said transforming minimizes error between a histogram of said pixel information and a specified histogram.
2. The method according to claim 1 , wherein said step of transforming comprises:
modifying said pixel data;
computing a weighting vector by minimizing the error between said histogram and said specified histogram; and
correcting said pixel data using said weighting vector.
3. The method according to claim 2 , wherein said weighting vector (W) is described by the formula:
W=B −1 Z.
4. The method according to claim 3 , wherein matrix B and vector Z are described by the formulas:
B
=
[
∑
i
=
1
N
P
i
2
b
1
i
2
∑
i
=
1
N
P
i
2
b
1
i
b
2
i
...
∑
i
=
1
N
P
i
2
b
1
i
b
M
i
∑
i
=
1
N
P
i
2
b
1
i
b
2
i
∑
i
=
1
N
P
i
2
b
2
i
2
...
∑
i
=
1
N
P
i
2
b
2
i
b
M
i
⋮
⋮
⋮
∑
i
=
1
N
P
i
2
b
M
i
b
1
i
∑
i
=
1
N
P
i
2
b
M
i
b
2
i
...
∑
i
=
1
N
P
i
2
b
M
i
2
]
and
Z
=
[
∑
i
=
1
N
s
i
P
i
b
1
i
∑
i
=
1
N
s
i
P
i
b
2
i
⋮
∑
i
=
1
N
s
i
P
i
b
Mi
]
where s i is the histogram specified transformed gray level of the i th pixel, N is the number of pixels in the image, P i is the original gray level of the i th pixel, b ji is the interpolation coefficient of the j th grid point acting on the i th pixel, W j is the weight value of the j th interpolating function, and M is the number of points in the grid.
5. The method according to claim 1 , wherein a plurality of grid points overlay said image of said object, each said grid point having associated therewith at least one interpolating function.
6. The method according to claim 5 , wherein said at least one interpolating function is selected from the group consisting of:
two-dimensional Gaussian functions, three-dimensional Gaussian functions, multiquadratic basis functions, and combinations thereof.
7. The method according to claim 1 , wherein said specified histogram is selected from the group consisting of:
uniform histograms, linearly-rising histograms, tissue contrast enhancement histograms, tissue class accentuation histograms and combinations thereof.
8. The method according to claim 1 , wherein said histogram is generated from said digital image.
9. An imaging system comprising:
a camera, said camera comprising a processor,
wherein said processor processes image data of an object, a histogram being generated from a portion of said image data,
wherein said processor transforms at least one pixel in said image data by minimizing error between said histogram and a specified histogram.
10. The imaging system according to claim 9 , wherein said processor transforms said at least one pixel by weighting each pixel according to a weighting function.
11. The imaging system according to claim 9 , wherein said processor in transforming:
modifies said at least one pixel;
computes a weighting vector by minimizing the error between said histogram and said specified histogram; and
corrects said at least one pixel using said weighting vector.
12. The imaging system according to claim 11 , wherein said weighting vector (W) is described by the formula:
W=B −1 Z.
13. The imaging system according to claim 12 , wherein matrix B and vector Z are described by the formulas:
B
=
[
∑
i
=
1
N
P
i
2
b
1
i
2
∑
i
=
1
N
P
i
2
b
1
i
b
2
i
...
∑
i
=
1
N
P
i
2
b
1
i
b
M
i
∑
i
=
1
N
P
i
2
b
1
i
b
2
i
∑
i
=
1
N
P
i
2
b
2
i
2
...
∑
i
=
1
N
P
i
2
b
2
i
b
M
i
⋮
⋮
⋮
∑
i
=
1
N
P
i
2
b
M
i
b
1
i
∑
i
=
1
N
P
i
2
b
M
i
b
2
i
...
∑
i
=
1
N
P
i
2
b
M
i
2
]
and
Z
=
[
∑
i
=
1
N
s
i
P
i
b
1
i
∑
i
=
1
N
s
i
P
i
b
2
i
⋮
∑
i
=
1
N
s
i
P
i
b
Mi
]
where s i is the histogram-specified transformed gray level of the i th pixel, N is the number of pixels in the image, P i is the original gray level of the i th pixel, b ji is the interpolation coefficient of the j th grid point acting on the i th pixel, W j is the weight value of the j th interpolating function, and M is the number of points in the grid.
14. The imaging system according to claim 9 , wherein a plurality of grid points overlay an image portion of said object, each said grid point having associated therewith at least one interpolating function.
15. The imaging system according to claim 9 , wherein said at least one interpolating function is selected from the group consisting of:
two-dimensional Gaussian functions, three-dimensional Gaussian functions, multiquadratic basis functions, and combinations thereof.
16. The imaging system according to claim 9 , wherein said specified histogram is selected from the group consisting of:
uniform histograms, linearly-rising histograms, tissue contrast enhancement histograms, tissue accentuation histograms and combinations thereof.
17. The imaging system according to of claim 9 , wherein said histogram is generated from a digital image of said object.
18. A computer-generated image, said image on a computer display comprising:
a plurality of pixels in an image portion of said image,
wherein said plurality of pixels have image intensity inhomogeneity therein; and
wherein at least one pixel in said plurality of pixels is transformed by minimizing error between a histogram generated from said plurality of pixels and a specified histogram.
19. The computer-generated image according to claim 18 , wherein a plurality of grid points overlay said image portion, each said grid point having associated therewith at least one interpolating function.
20. The computer-generated image according to claim 18 , wherein said specified histogram is selected from the group consisting of:
uniform histograms, linearly-rising histograms, tissue contrast enhancement histograms, tissue accentuation histograms and combinations thereof.