Removal of image artifacts from sensor dust
View Patent ↗Removal of the effects of dust or other impurities on image data is described. In one example, a model of artifact formation from sensor dust is determined. From the model of artifact formation, contextual information in the image and a color consistency constraint may be applied on the dust to remove the dust artifacts. Artifacts may also be removed from multiple images from the same or different cameras or camera settings.
1. A system comprising a computing device and at least one program module that are together configured for performing actions comprising:
identifying an artifact in an image, where the artifact results from an impurity, and where the identifying comprises:
determining, with respect to an estimated projection geometry, a function space of possible impurity artifacts,
identifying a minimum distance between an estimated artifact formation model and one of the possible impurity artifacts of the determined function space, and
constructing, based on the determined function space and on the identified minimum distance, a dust removal solution; and
removing, based on the constructed dust removal solution, the identified artifact from the image.
2. The system of claim 1 where the estimated projection geometry is based on a displacement of a filter from a sensor.
3. The system of claim 1 where the removing comprises performing an iterative optimization that solves for a proportion of scene radiance that passes through the impurity and reaches a sensor, and for an impurity-free appearance of a portion of the image.
4. The system of claim 1 where the removing is in response to distinguishing the artifact from an image feature based on a color consistency.
5. The system of claim 4 where the distinguishing comprises assuming that the impurity is composed of a material that is monochromatic or of uniform color.
6. The system of claim 4 where the distinguishing comprises detecting candidate impurity regions.
7. A method performed on a computing device, the method comprising:
identifying an artifact in an image, where the artifact results from an impurity, and where the identifying comprises:
determining, with respect to an estimated projection geometry, a function space of possible impurity artifacts,
identifying a minimum distance between an estimated artifact formation model and one of the possible impurity artifacts of the determined function space, and
constructing, based on the determined function space and on the identified minimum distance, a dust removal solution; and
removing, based on the constructed dust removal solution, the identified artifact from the image.
8. The method of claim 7 where the estimated projection geometry is based on a displacement of a filter from a sensor.
9. The method of claim 7 where the removing comprises performing an iterative optimization that solves for a proportion of scene radiance that passes through the impurity and reaches a sensor, and for an impurity-free appearance of a portion of the image.
10. The method of claim 7 where the removing is in response to distinguishing the artifact from an image feature based on a color consistency.
11. The method of claim 10 where the distinguishing comprises assuming that the impurity is composed of a material that is monochromatic or of uniform color.
12. The method of claim 10 where the distinguishing comprises detecting candidate impurity regions.
13. At least one computer-readable storage device storing computer-executable instructions that, when executed by a computing device, cause the computing device to perform a method comprising:
identifying an artifact in an image, where the artifact results from an impurity, and where the identifying comprises:
determining, with respect to an estimated projection geometry, a function space of possible impurity artifacts,
identifying a minimum distance between an estimated artifact formation model and one of the possible impurity artifacts of the determined function space, and
constructing, based on the determined function space and on the identified minimum distance, a dust removal solution; and
removing, based on the constructed dust removal solution, the identified artifact from the image.
14. The at least one computer-readable storage device of claim 13 where the estimated projection geometry is based on a displacement of a filter from a sensor.
15. The at least one computer-readable storage device of claim 13 where the removing comprises performing an iterative optimization that solves for a proportion of scene radiance that passes through the impurity and reaches a sensor, and for an impurity-free appearance of a portion of the image.
16. The at least one computer-readable storage device of claim 13 where the removing is in response to distinguishing the artifact from an image feature based on a color consistency.
17. The at least one computer-readable storage device of claim 16 where the distinguishing comprises assuming that the impurity is composed of a material that is monochromatic or of uniform color.
18. The at least one computer-readable storage device of claim 16 where the distinguishing comprises detecting candidate impurity regions.