Sparse phase unwrapping
Sparse phase unwrapping is disclosed. A first image and a second image are received. The first image and the second image are coregistered. The first image and the second image comprise respective phase data. An unwrapped interferogram is generated, including by solving an optimization problem using a nonconvex penalty function, where minimizing the penalty function produces sparse minimizers.
1. A system, comprising:
a processor configured to:
receive a pair of images comprising a first image and a second image, wherein the first image and the second image comprise synthetic aperture radar imagery, wherein the first image and the second image are coregistered and wherein the first image and the second image comprise respective phase data;
generate an unwrapped interferogram for the pair of images including by solving an optimization problem using a nonconvex penalty function, wherein the nonconvex penalty function has an explicit proximal mapping; and
use the unwrapped interferogram, at least in part, to detect a deformation; and
a memory coupled to the processor and configured to provide the processor with instructions.
2. The system of claim 1 wherein application of the penalty function results in a sparse difference between an estimated gradient and a gradient of an estimated unwrapped phase.
3. The system of claim 1 wherein the processor is further configured to generate a wrapped differential interferogram that includes per pixel phase differences between pixels appearing in both the first image and the second image.
4. The system of claim 1 wherein the processor is further configured to detect a subsidence at least in part by using the unwrapped interferogram.
5. The system of claim 1 wherein the processor is further configured to detect mining activity at least in part by using the unwrapped interferogram.
6. The system of claim 1 wherein the processor is further configured to detect plowing activity at least in part by using the unwrapped interferogram.
7. The system of claim 1 wherein the processor is further configured to coregister the first image and the second image.
8. A method, comprising:
receiving a pair of images comprising a first image and a second image, wherein the first image and the second image comprise synthetic aperture radar imagery, wherein the first image and the second image are coregistered and wherein the first image and the second image comprise respective phase data;
generating an unwrapped interferogram for the pair of images including by solving an optimization problem using a nonconvex penalty function, wherein the nonconvex penalty function has an explicit proximal mapping; and
using the unwrapped interferogram, at least in part, to detect a deformation.
9. The method of claim 8 wherein application of the penalty function results in a sparse difference between an estimated gradient and a gradient of an estimated unwrapped phase.
10. The method of claim 8 further comprising generating a wrapped differential interferogram that includes per pixel phase differences between pixels appearing in both the first image and the second image.
11. The method of claim 8 , wherein detecting the deformation includes detecting a subsidence.
12. The method of claim 8 , wherein detecting the deformation includes detecting mining activity.
13. The method of claim 8 , wherein detecting the deformation includes detecting plowing activity.
14. The method of claim 8 further comprising coregistering the first and the second image.
15. A computer program product embodied in a non-transitory computer readable storage medium and comprising instructions which, when executed, cause a computer to:
receive a pair of images comprising a first image and a second image, wherein the first image and the second image comprise synthetic aperture radar imagery, wherein the first image and the second image are coregistered and wherein the first image and the second image comprise respective phase data;
generate an unwrapped interferogram for the pair of images including by solving an optimization problem using a nonconvex penalty function, wherein the nonconvex penalty function has an explicit proximal mapping; and
use the unwrapped interferogram, at least in part, to detect a deformation.
16. The system of claim 1 , wherein the proximal mapping is defined as follows:
proximal λ G ( x )=arg min y G ( y )+∥ y−x∥ 2 2/( 2λ)=max{ x−λ 2−p x p−1 ,0} x/|x|,
wherein the value of p is less than one, G is the nonconvex penalty function, x and y are real number value variables, lambda is a positive number value parameter, G(x) is the value of the function G applied to the variable x, and G(y) is the value of the function G applied to the variable y.
17. The method of claim wherein the proximal mapping is defined as follows:
proximal λ G ( x )=arg min y G ( y )+∥ y−x∥ 2 2/( 2λ)=max{ x−λ 2−p x p−1 ,0} x/|x|,
wherein the value of p is less than one, G is the nonconvex penalty function, x and y are real number value variables, lambda is a positive number value parameter, G(x) is the value of the function G applied to the variable x, and G(y) is the value of the function G applied to the variable y.
18. The computer program product of claim 15 , wherein the proximal mapping is defined as follows:
proximal λ G ( x )=arg min y G ( y )+∥ y−x∥ 2 2/( 2λ)=max{ x−λ 2−p x p−1 ,0} x/|x|,
wherein the value of p is less than one, G is the nonconvex penalty function, x and y are real number value variables, lambda is a positive number value parameter, G(x) is the value of the function G applied to the variable x, and G(y) is the value of the function G applied to the variable y.