Method of generating a multiscale contrast enhanced image

1. A computer-implemented method for enhancing the contrast of an image that is represented by a digital signal wherein

a. said digital signal is decomposed into a multi-scale representation comprising at least two detail images representing detail at multiple scales and approximation images of which the detail images are derived, an approximation image at a scale representing the grey values of said image in which all details at that scale have been omitted,

b. translation difference images are computed by pixel-wise subtracting the values of an approximation image at scale s and the values of translated versions of said approximation image,

c. the values of said translation difference images are non-linearly modified,

d. an amplification image is computed at at least one scale as the ratio of 2 images wherein the first image is computed by forming a weighted sum of said modified translation difference images at the same or smaller scale and the second image is created by forming a weighted sum of un-enhanced translation difference images at the same or smaller scale, said un-enhanced difference images being different from said detail images,

e. an enhanced multi-scale detail representation is computed by modifying at at least one scale the detail image by applying the amplification image at that scale,

f. an enhanced image representation is computed by applying a reconstruction algorithm inverting the multi-scale decomposition to the enhanced multi-scale detail representation, wherein said non-linear modification of the values of the translation difference image is steered by the values of an orientation map which comprises for each pixel on a specific scale a local direction of interest or wherein values of weights that are applied to the images that are summed up to form said amplification image are steered by the values of an orientation map which comprises for each pixel on a scale a local direction of interest.

2. A computer-implemented method for enhancing the contrast of an image, the method comprising:

a. decomposing a digital signal into a multi-scale representation comprising at least two detail images representing detail at multiple scales and approximation images of which the detail images are derived, an approximation image at a scale representing the grey values of said image in which all details at that scale have been omitted,

b. computing translation difference images by pixel-wise subtracting the values of an approximation image at scale s and the values of translated versions of said approximation image,

c. modifying the values of said translation difference images non-linearly,

d. computing an amplification image at at least one scale as the ratio of two images wherein the first image is computed by forming a weighted sum of said modified translation difference images at the same or smaller scale and the second image is created by forming a weighted sum of un-enhanced translation difference images at the same or smaller scale, said un-enhanced difference images being different from said detail images,

e. computing an enhanced multi-scale detail representation by modifying at at least one scale the detail image by applying the amplification image at that scale,

f. computing an enhanced image representation by applying a reconstruction algorithm inverting the multi-scale decomposition to the enhanced multi-scale detail representation,

wherein said non-linear modification of the values of the translation difference image is steered by the values of an orientation map which comprises for each pixel on a specific scale a local direction of interest or wherein values of weights that are applied to the images that are summed up to form said amplification image are steered by the values of an orientation map which comprises for each pixel on a scale a local direction of interest.

3. A method according to claim 1 wherein said orientation map is deduced from filtered approximation images.

4. A method according to claim 1 wherein a translation difference image at a specific scale is computed out of an approximation image at the same scale.

5. A method according to claim 1 wherein all of said translation difference images are computed out of the original image.

6. A method according to claim 1 wherein a translation difference image at a scale k is computed out of an approximation image at scale m, wherein m represents a scale between scale 1 and scale k−1.

7. A method according to claim 1 wherein the center difference images being weighted sums of translation difference images are identical to the multi-scale detail images or a close approximation of the multi-scale detail images.

8. A method according to claim 1 wherein said image is a mammographic image.

9. A method according to claim 1 wherein said image is a CT image.

10. A computer program product embodied on a non-transitory computer readable medium adapted to carry out a method of enhancing the contrast of an image that is represented by a digital signal when run on a computer, the method comprising:

a. said digital signal is decomposed into a multi-scale representation comprising at least two detail images representing detail at multiple scales and approximation images of which the detail images are derived, an approximation image at a scale representing the grey values of said image in which all details at that scale have been omitted,

b. translation difference images are computed by pixel-wise subtracting the values of an approximation image at scale s and the values of translated versions of said approximation image,

c. the values of said translation difference images are non-linearly modified,

d. an amplification image is computed at least one scale as the ratio of 2 images wherein the first image is computed by forming a weighted sum of said modified translation difference images at the same or smaller scale and the second image is created by forming a weighted sum of un-enhanced translation difference images at the same or smaller scale, said un-enhanced difference images being different from said detail images,

e. an enhanced multi-scale detail representation is computed by modifying at at least one scale the detail image by applying the amplification image at that scale,

f. an enhanced image representation is computed by applying a reconstruction algorithm inverting the multi-scale decomposition to the enhanced multi-scale detail representation,

wherein said non-linear modification of the values of the translation difference image is steered by the values of an orientation map which comprises for each pixel on a specific scale a local direction of interest or wherein values of weights that are applied to the images that are summed up to form said amplification image are steered by the values of an orientation map which comprises for each pixel on a scale a local direction of interest.

11. A non-transitory computer readable medium for enhancing the contrast of an image that is represented by a digital signal comprising computer executable program code adapted to carry out a method comprising:

g. said digital signal is decomposed into a multi-scale representation comprising at least two detail images representing detail at multiple scales and approximation images of which the detail images are derived, an approximation image at a scale representing the grey values of said image in which all details at that scale have been omitted,

h. translation difference images are computed by pixel-wise subtracting the values of an approximation image at scale s and the values of translated versions of said approximation image,

i. the values of said translation difference images are non-linearly modified,

j. an amplification image is computed at at least one scale as the ratio of 2 images wherein the first image is computed by forming a weighted sum of said modified translation difference images at the same or smaller scale and the second image is created by forming a weighted sum of un-enhanced translation difference images at the same or smaller scale, said un-enhanced difference images being different from said detail images,

k. an enhanced multi-scale detail representation is computed by modifying at at least one scale the detail image by applying the amplification image at that scale,

l. an enhanced image representation is computed by applying a reconstruction algorithm inverting the multi-scale decomposition to the enhanced multi-scale detail representation,

wherein said non-linear modification of the values of the translation difference image is steered by the values of an orientation map which comprises for each pixel on a specific scale a local direction of interest or wherein values of weights that are applied to the images that are summed up to form said amplification image are steered by the values of an orientation map which comprises for each pixel on a scale a local direction of interest.

12. A method according to claim 2 , further comprising deducing said orientation map from filtered approximation images.

13. A method according to claim 2 , further comprising computing a translation difference image at a specific scale out of an approximation image at the same scale.

14. A method according to claim 2 , wherein all of said translation difference images are computed out of the original image.

15. A method according to claim 2 , wherein a translation difference image at a scale k is computed out of an approximation image at scale m, wherein m represents a scale between scale 1 and scale k−1.

16. A method according to claim 2 , wherein the center difference images being weighted sums of translation difference images are identical to the multi-scale detail images or a close approximation of the multi-scale detail images.

17. A method according to claim 2 , further comprising creating said translation difference image from a mammographic image.

18. A method according to claim 2 , further comprising creating said translation difference image from a computed tomography image.