Single-image specular reflection separation
View Patent ↗Systems and methods are discussed to separate the specular reflectivity and/or the diffuse reflectivity from an input image. Embodiments of the invention can be used to determine the specular chromaticity by iteratively solving one or more objective functions. An objective function can include functions that take into account the smooth gradient of the specular chromaticity. An objective function can take into account the interior chromatic homogeneity of the diffuse chromaticity and/or the sharp changes between chromaticity. Embodiments of the invention can also be used to determine the specular chromaticity of an image using a pseudo specular-free image that is calculated from the input image and a dark channel image that can be used to iteratively solve an objective function(s).
1. A method comprising:
receiving an input image at a computer device, wherein the input image comprises a red channel, a green channel, and a blue channel;
calculating a dark channel image from the input image, wherein each pixel value of the dark channel image comprises a lowest pixel value of a corresponding pixel of the red channel, a corresponding pixel of the green channel, and a corresponding pixel of the blue channel;
calculating a pseudo specular-free image as a difference between the input image and the dark channel image;
calculating a specular chromaticity of the input image from the pseudo specular-free image by:
computing a cluster analysis on the pseudo specular-free image that clusters each pixel in the pseudo specular-free image into clusters having similar chromaticity, wherein the cluster analysis determines a chromaticity index for each pixel and a chromaticity vector that indicates an average chromaticity of each cluster represented by the chromaticity index, and
solving an objective function to determine the specular chromaticity of the input image, wherein the objective function specifies that either (i) diffuse reflectivity is relatively constant within each cluster, or (ii) diffuse reflectivity changes sharply between clusters; and
providing the calculated specular chromaticity of the input image.
2. The method according to claim 1 , further comprising calculating a chromaticity of the pseudo specular-free image, and wherein the specular chromaticity of the input image is calculated using the chromaticity of the pseudo specular-free image.
3. The method according to claim 1 , further comprising:
solving an additional objective function to determine a specular reflectivity of the input image from the pseudo specular-free image, wherein the additional objective function specifies that specular reflectivity is smooth across edges within the input image; and
providing the determined specular reflectivity of the input image.
4. The method according to claim 1 , further comprising:
solving an additional objective function to determine a specular reflectivity of the input image from the pseudo specular-free image, wherein the additional objective function specifies that diffuse reflectivity is relatively constant for the same chromaticity; and
providing the determined specular reflectivity of the input image.
5. The method according to claim 1 , further comprising:
solving an additional objective function to determine a specular reflectivity of the input image from the pseudo specular-free image, wherein the additional objective function specifies that diffuse reflectivity is relatively constant within each cluster; and
providing the determined specular reflectivity of the input image.
6. The method according to claim 1 , further comprising:
solving an additional objective function to determine a specular reflectivity of the input image from the pseudo specular-free image, wherein the additional objective function specifies that diffuse reflectivity changes sharply between clusters; and
providing the determined specular reflectivity of the input image.
7. A system for determining a specular reflectivity comprising:
a digital storage device; and
a processor configured to:
receive an input image and store the input image in the digital storage device, wherein the input image comprises a red channel, a green channel, and a blue channel for each pixel in the input image;
calculate a dark channel image of the input image, wherein the dark channel image comprises a value for each pixel, each value corresponding to a lowest pixel value of a corresponding red channel pixel, a corresponding green channel pixel, and a corresponding blue channel pixel of the input image;
calculate a pseudo specular-free image as a difference between the input image and the dark channel image of the input image;
calculate the specular reflectivity of the input image from the pseudo specular-free image by:
computing a cluster analysis on the pseudo specular-free image that clusters each pixel in the pseudo specular-free image into clusters having similar chromaticity, wherein the cluster analysis determines a chromaticity index for each pixel and a chromaticity vector that indicates an average chromaticity of each cluster represented by the chromaticity index, and
solving an objective function to determine the specular reflectivity of the input image; wherein the objective function specifies that either (i) diffuse reflectivity is relatively constant within each cluster, or (ii) diffuse reflectivity changes sharply between clusters; and
provide the calculated specular reflectivity of the input image.