Method and system for generating intrinsic images using a smooth illumination constraint

1. An automated, computerized method for processing an image, comprising the steps of:

providing an image file depicting an image, in a computer memory;

organizing information relevant to the image according to a smooth-illumination constraint enforcing a Laplacian filter response such that ∇ 2 ƒ(x, y)=0; and

utilizing the smooth-illumination constraint to generate an intrinsic image from the image,

wherein the organizing includes applying the Laplacian filter response, ∇ 2 ƒ(x, y)=0, to a log-illumination image,

wherein the organizing includes approximating second derivatives in the x and y directions to represent ∇ 2 ƒ(x, y) with discrete centered differences to provide a Laplacian filtering kernel, wherein the organizing includes convolving the Laplacian filtering kernel with ƒ(x, y) to develop a linear system of constraints,

wherein the linear system of constraints is developed by applying the Laplacian filtering kernel to log-image data to generate a right-hand side of the linear system of constraints and building a left-hand matrix enforcing the same linear filtering relationship on unknown values.

2. The method of claim 1 wherein the information relevant to the image comprises pixels.

3. The method of claim 1 wherein the information relevant to the image comprises tokens.

4. The method of claim 1 wherein the intrinsic image comprises a material image.

5. The method of claim 1 wherein the intrinsic image comprises an illumination image.

6. The method as recited in claim 1 further comprising identifying regions of a single material reflectance.

7. The method as recited in claim 6 wherein the step of organizing comprises using the smooth-illumination constraint to constrain adjacent single material reflectance regions.

8. The method as recited in claim 1 wherein the Laplacian filter response is computed from the sum of the second derivatives in the x and y directions, as follows:

∇ 2 ƒ( x,y )∂ 2 ƒ( x,y )/∂ x 2 +∂ 2 ƒ( x,y )/∂ y 2 .

9. The method as recited in claim 1 wherein Laplacian filter response, ∇ 2 ƒ(x, y)=0, is met when the local variation of the two dimensional function ƒ(x, y) is planar.

10. A computer system which comprises:

a CPU; and

a memory storing an image file containing an image;

the CPU arranged and configured to execute a routine to organize information relevant to the image according to a smooth-illumination constraint enforcing a Laplacian filter response such that ∇ 2 ƒ(x, y)=0 and utilize the smooth-illumination constraint to generate an intrinsic image from the image,

wherein the organizing includes applying the Laplacian filter response, ∇ 2 ƒ(x, y)=0, to a log-illumination image,

wherein the organizing includes approximating second derivatives in the x and y directions to represent ∇ 2 ƒ(x, y) with discrete centered differences to provide a Laplacian filtering kernel, wherein the organizing includes convolving the Laplacian filtering kernel with ƒ(x, y) to develop a linear system of constraints,

wherein the linear system of constraints is developed by applying the Laplacian filtering kernel to log-image data to generate a right-hand side of the linear system of constraints and building a left-hand matrix enforcing the same linear filtering relationship on unknown values.

11. 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, in a computer memory, organize information relevant to the image according to a smooth-illumination constraint enforcing a Laplacian filter response such that ∇ 2 ƒ(x, y)=0 and utilize the smooth-illumination constraint to generate an intrinsic image from the image,

wherein the organizing includes applying the Laplacian filter response, ∇ 2 ƒ(x, y)=0, to a log-illumination image,

wherein the organizing includes approximating second derivatives in the x and y directions to represent ∇ 2 ƒ(x, y) with discrete centered differences to provide a Laplacian filtering kernel,

wherein the organizing includes convolving the Laplacian filtering kernel with ƒ(x, y) to develop a linear system of constraints,

wherein the linear system of constraints is developed by applying the Laplacian filtering kernel to log-image data to generate a right-hand side of the linear system of constraints and building a left-hand matrix enforcing the same linear filtering relationship on unknown values.