USRE47341EExpiredUtilityPatentIndex 73
Filtering control method for improving image quality of bi-linear interpolated image
Est. expiryOct 21, 2019(expired)· nominal 20-yr term from priority
G06T 3/4007G06T 5/003G06T 5/20H04N 19/59H04N 19/80G06T 5/73
73
PatentIndex Score
1
Cited by
34
References
24
Claims
Abstract
The present invention relates to an interpolation method for enlarging a digital image or predicting a moving vector of a compressed image system as a sub-pixel unit when the image digitized through a CCD (Charge Coupled Device) camera ect. has a low resolution in a video phone or video conference or general digital video system, particularly the present invention can be adapted to a post processor of a compressed digital image in order to improve the image quality, and can be used for finding a moving vector of a moving picture compressed type, accordingly the present invention is capable of improving the image quality.
Claims
exact text as granted — not AI-modifiedWhat is claimed is:
1. A filtering control method for improving the image quality of a bi-linear interpolated image when recovering a high resolution image from a low resolution image, the method comprising performing operations using at least one processor, the operations comprising:
restoring a requested high resolution image f by finding an added filter coefficient Q of a PSF(P) and a bi-linear interpolation filter B from an equation f=Pg=PBz=Qz, wherein f is the high resolution image as requested, P is the PSF (Point Spread Function), g is the high resolution image found by the bi-linear interpolation method, and z is the low resolution image; wherein the high resolution image f can be restored by performing an added function M(f) definition process for finding the PSF(H) from an equation g=Bz=Hf+n, wherein B, H are bi-linear interpolation filters, and n is a noise component generated by the assumed H; and wherein the added function M(f) is defined as M(f)=∥g−Hf∥ 2 +α∥Cf∥ 2 , wherein α is a regularization parameter, and C is a two-dimensional high frequency filter for finding mitigation of the original image.
2. The filtering control method for improving the image quality of the bi-linear interpolated image according to claim 1 , wherein the regularization parameter α is fixed as ‘1’ in order to reduce a computational complexity.
3. The filtering control method for improving image quality of the b-linear interpolated image according to claim 1 , wherein a two-dimensional gaussian filter is used as the two-dimensional high frequency filer C in order to determine the mitigation of the original image.
4. A filtering control method for improving the image quality of a bi-linear interpolated image when recovering a high resolution image from a low resolution image, the method comprising performing operations using at least one processor, the operations comprising: restoring a requested high resolution image f by finding an added filter coefficient Q of a PSF(P) and a bi-linear interpolation filter B from an equation f=Pg=PBz=Qz, wherein f is the high resolution image as requested, P is the PSF (Point Spread Function), g is the high resolution image found by the bi-linear interpolation method, and z is the low resolution image; wherein the high resolution image f can be restored by performing an added function M(f) definition process for finding the PSF(H) from an equation g=Bz=Hf+n, wherein B, H are bi-linear interpolation filters, and n is a noise component generated by the assumed H; wherein the high resolution image f is restored by finding a PSF(P) of a f=Pg function after finding the PSF(H) from the added function M(f); and wherein the PSF(H) is found by using an equation
H
(
k
,
l
)
=
G
(
k
,
l
)
F
(
k
,
l
)
.
G(k,l) is the component in the k,l frequency region of the bi-linear interpolated image, and F(k,l) is the component in the k,l frequency region of the high resolution image.
5. A filtering control method for improving the image quality of a bi-linear interpolated image when recovering a high resolution image from a low resolution image, the method comprising performing operations using at least one processor, the operations comprising:
restoring a requested high resolution image f by finding an added filter coefficient Q of a PSF(P) and a bi-linear interpolation filter B from an equation f=Pg=PBz=Qz, wherein f is the high resolution image as requested, P is the PSF (Point Spread Function), g is the high resolution image found by the bi-linear interpolation method, and z is the low resolution image; wherein the PSF(P) can be found by getting an IFT (Inverse Fourier Transform) by an equation
P
(
k
,
l
)
=
H
*
(
k
,
l
)
H
*
(
k
,
l
)
H
(
k
,
l
)
+
C
*
(
k
,
l
)
C
(
k
,
l
)
H(k,l) is a component in the k,l frequency region of the PSF(H), and C is a two-dimensional high frequency filter.
6. The filtering control method for improving the image quality of the bi-linear interpolated image according to claim 5 , wherein the number of a kernal of the PSF(P) is set in accordance with an up-sampling value of the image.
7. A filtering control method for improving the image quality of a bi-linear interpolated image when recovering a high resolution image from a low resolution image, the method comprising performing operations using at least one processor the operations comprising:
defining an added function M(f) for finding a PSF(H) from an equation g=Bz=Hf+n (wherein B, H are bi-linear filters, n is a noise component generated by an assumed H when the H is a PSF (Point Spread Function), f is a requested high resolution image, z is a low resolution image, and g is a high resolution image gotten by the bi-linear interpolation method); finding a PSF(P) of a f=Pg function after finding the PSF (H) from the defined added function M(f); and restoring the requested high resolution image f by finding an added filter coefficient Q of the PSF(P) and interpolation filter B from the equation f=Pg=PBZ=Qz; wherein the added function M(f) is defined as M(f)=∥g−Hf∥ 2 +α∥Cf∥ 2 , wherein α is a regularization parameter, and C is a two-dimensional high frequency filter for finding the mitigation of the original image.
8. The filtering control method for improving the image quality a of the bi-linear interpolated image according to claim 7 , wherein the regularization parameter α is fixed as ‘1’ in order to reduce a computational complexity.
9. The filtering control method for improving image quality of the bi-linear interpolated image according to claim 7 , wherein a two-dimensional gaussian filter is used as the two-dimensional high frequency filter C in order to determine the mitigation of the original image.
10. A filtering control method for improving the image quality of a bi-linear interpolated image when recovering a high resolution image from a low resolution image, the method comprising performing operations using at least one processor the operations comprising:
defining an added function M(f) for finding a PSF(H) from an equation g=Bz=Hf+n (wherein B, H are bi-linear filters, n is a noise component generated by an assumed H when the H is a PSF (Point Spread Function), f is a requested high resolution image, z is a low resolution image, and g is a high resolution image gotten by the bi-linear interpolation method); finding a PSF(P) of a f=Pg function after finding the PSF (H) from the defined added function M(f); and restoring the requested high resolution image f by finding an added filter coefficient Q of the PSF(P) and interpolation filter B from the equation f=Pg=PBZ=Qz; wherein the PSF(H) is found by an equation
H(k,l)=(G(k,l)/F(−k,l),
wherein G(k,l) is the component in the k,l frequency region of the bi-linear interpolated image, and F(k,l) is the component in the k,l frequency region of the high resolution image.
11. A filtering control method for improving the image quality of a bi-linear interpolated image when recovering a high resolution image from a low resolution image, the method comprising performing operations using at least one processor, the operations comprising:
defining an added function M(f) for finding a PSF(H) from an equation g=Bz=Hf+n (wherein B, H are bi-linear filters, n is a noise component generated by an assumed H when the H is a PSF (Point Spread Function), f is a requested high resolution image, z is a low resolution image, and g is a high resolution image gotten by the bi-linear interpolation method); finding a PSF(P) of a f=Pg function after finding the PSF(H) from the defined added function M(f); and restoring the requested high resolution image f by finding an added filter coefficient Q of the PSF(P) and interpolation filter B from the equation f=Pg=PBZ=Qz; wherein the PSF(P) is found by using an IFT (Inverse Fourier Transform) by an equation
P
(
k
,
l
)
=
H
*
(
k
,
l
)
H
*
(
k
,
l
)
H
(
k
,
l
)
+
C
*
(
k
,
l
)
C
(
k
,
l
)
H(k,l) is a component in the k,l frequency region of the PSF(H), and C is a two-dimensional high frequency filter.
12. The filtering control method for improving the image quality of the bi-linear interpolated image according to claim 11 , wherein the number of a kernal of the PSF(P) is differently set in accordance with an up-sampling value of the image.
13. A method for generating an interpolated pixel data, the method comprising generating a set of the interpolated pixel data from a set of original pixel data from an original image, wherein interpolated pixel data for a particular pixel is generated by performing operations using at least one processor, the operations comprising:
selecting original pixel data for more than three pixels of the original image; obtaining at least a first filter coefficient and a second filter coefficient, the first filter coefficient and the second filter coefficient being configured to interpolate the original pixel data; applying the first filter coefficient to the selected original pixel data to produce first interpolated pixel data, the first filter coefficient including weighting factors having at least three different numerical values, wherein applying the first filter coefficient to the selected original pixel data comprises:
multiplying each of the weighting factors and the selected original pixel data to produce weighted pixel data, and
summing the weighted pixel data to produce the first interpolated pixel data;
multiplying the second filter coefficient and the first interpolated pixel data to produce second interpolated pixel data; and identifying the interpolated pixel data as the second interpolated pixel data.
14. The method of claim 13, wherein the second filter coefficient is a matrix that includes one or more individual numeric values.
15. The method of claim 13, wherein the first filter coefficient and the second filter coefficient each comprise at least one integer value.
16. The method of claim 13, wherein a value of the first filter coefficient and a value of the second filter coefficient are one.
17. The method of claim 13, wherein a value of the second filter coefficient is one.
18. The method of claim 13, wherein the original image is obtained from a low-resolution imaging system.
19. The method of claim 13, wherein the second filter coefficient is a point spread function (P) and the first filter coefficient is a bi-linear interpolation filter (B).
20. A method for a moving picture compression with a video system by generating an interpolated pixel data, the method comprising:
generating a set of interpolated pixel data from a set of original pixel data from an original image; and finding a sub-pixel motion vector using the set of interpolated pixel data, wherein interpolated pixel data for a particular pixel is generated by performing operations using the video system, the operations comprising: selecting, by the video system, original pixel data including more than three pixels of the original image; obtaining, by the video system, at least a first filter coefficient and a second filter coefficient, the first filter coefficient and the second filter coefficient being configured to interpolate the original pixel data; applying, by the video system, the first filter coefficient to the selected original pixel data to produce first interpolated pixel data, the first filter coefficient including weighting factors having at least three different numerical values, the at least three different numerical values including integer values, wherein said applying the first filter coefficient to the selected original pixel data comprises: multiplying each of the weighting factors and the selected original pixel data to produce weighted pixel data, and summing the weighted pixel data to produce the first interpolated pixel data; applying, by the video system, the second filter coefficient to the first interpolated pixel data to produce second interpolated pixel data, wherein said applying the second filter coefficient comprises multiplying the second filter coefficient and the first interpolated pixel data and summing results of multiplying the second filter coefficient and the first interpolated pixel data to produce the second interpolated pixel data; and identifying, by the video system, the interpolated pixel data for the particular pixel as the second interpolated pixel data.
21. The method of claim 20, wherein the second filter coefficient is a matrix that includes one or more individual numeric values.
22. The method of claim 20, wherein the second filter coefficient includes at least one integer value.
23. The method of claim 20, wherein the original image is obtained from a low-resolution imaging system.
24. The method of claim 20, wherein the at least three different numerical values include at least one positive integer value and at least one negative integer value.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.