Weighted entropy minimization for optimizing a log-chromaticity normal for use in an image process
View Patent ↗In a first exemplary embodiment of the present invention, an automated, computerized method is provided for processing an image. According to a feature of the present invention, the method comprises the steps of providing an image file depicting an image defined by image locations, in a computer memory, generating a bi-illuminant chromaticity plane in a log color space for representing the image locations of the image in a log-chromaticity representation for the image, providing an estimate for an orientation of the bi-illuminant chromaticity plane and optimizing the estimate of an orientation as a function of a minimization of a weighted entropy.
1. An automated, computerized method for processing an image, comprising the steps of:
providing an image file depicting an image defined by image locations, in a computer memory;
generating a bi-illuminant chromaticity plane in a log color space for representing the image locations of the image in a log-chromaticity representation for the image;
providing an estimate for an orientation of the bi-illuminant chromaticity plane; and
optimizing the estimate of an orientation as a function of a minimization of a weighted entropy, wherein the weighted entropy is based upon a distance function between the image locations and a sub-set of the image locations used to calculate the estimate for an orientation.
2. The method of claim 1 wherein the distance function is based upon a spatial distance and a spectral distance related to the image locations and the sub-set of image locations used to calculate the estimate for an orientation.
3. The method of claim 1 wherein the step of optimizing the estimate of an orientation as a function of a minimization of a weighted entropy is carried out by executing a search algorithm for establishing a selection of orientations, the search algorithm being initialized with the estimate for an orientation, and for each one of the selection of orientations, calculating a weighted entropy, and selecting a selected orientation corresponding to the one of the selection of orientations with a minimum calculated entropy.
4. The method of claim 1 wherein the image locations comprise pixels.
5. The method of claim 1 wherein the image locations comprise tokens.
6. A computer system which comprises:
a CPU; and
a memory storing an image file containing an image defined by image locations;
the CPU arranged and configured to execute a routine to:
generate a bi-illuminant chromaticity plane in a log color space for representing the image locations of the image in a log-chromaticity representation for the image,
provide an estimate for an orientation of the bi-illuminant chromaticity plane and
optimize the estimate of an orientation as a function of a minimization of a weighted entropy, wherein the weighted entropy is based upon a distance function between the image locations and a sub-set of the image locations used to calculate the estimate for an orientation.
7. A computer program product, disposed on a non-transitory computer readable media, the product including computer executable process steps operable to control a computer to:
provide an image file depicting an image defined by image locations, in a computer memory,
generate a bi-illuminant chromaticity plane in a log color space for representing the image locations of the image in a log-chromaticity representation for the image,
provide an estimate for an orientation of the bi-illuminant chromaticity plane and
optimize the estimate of an orientation as a function of a minimization of a weighted entropy, wherein the weighted entropy is based upon a distance function between the image locations and a sub-set of the image locations used to calculate the estimate for an orientation.
8. The computer program product of claim 7 wherein the distance function is based upon a spatial distance and a spectral distance related to the image locations and the sub-set of image locations used to calculate the estimate for an orientation.
9. The computer program product of claim 7 wherein the process step to optimize the estimate of an orientation as a function of a minimization of a weighted entropy is carried out by executing a search algorithm for establishing a selection of orientations, the search algorithm being initialized with the estimate for an orientation, and for each one of the selection of orientations, calculating a weighted entropy, and selecting a selected orientation corresponding to the one of the selection of orientations with a minimum calculated entropy.
10. The computer program product of claim 7 wherein the image locations comprise pixels.
11. The computer program product of claim 7 wherein the image locations comprise tokens.