[0001] A computer-implemented method for adding texture to a digital image, wherein the computer has a digital image, comprising the method the following steps: - transforming the digital image into an image R that emulates the retinal output; - adding noise n<r> to the image R in order to obtain a noisy image R<r>; - applying to the noisy image R<r> the inverse of the transformation applied to the digital image in order to obtain a final image O.
1. A computer-implemented method for adding texture to a digital image, wherein the computer has a digital image, comprising the method the following steps: - transforming the digital image into an image R that emulates the retinal output; - adding noise n<r> to the image R in order to obtain a noisy image R<r>; - applying to the noisy image R<r> the inverse of the transformation applied to the digital image in order to obtain a final image O. 2. A method according to the previous claim wherein the transformation of the digital image into an image R that emulates the retinal output involves obtaining an emulation of the response of photoreceptors (I<P>) to the light stimulus of the digital image. 3. A method according to the previous claims wherein the value of the emulation of the response of photoreceptors for each pixel x of the digital image is obtained via the function P: I p x = P I x = I x n I x n + I s n <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0059" /> wherein I denotes one colour channel of the digital image, n controls the slope of the function and is a value in the following range: 0,5 ≤ n ≤ 1 and Is is the semi-saturation constant and is a value in the following range: 0,1 ≤ I<s> ≤ 0,5. 4. A method according to the previous claim wherein n is 0,74 and I<s> is 0,18. 5. A method according to any of the previous claims wherein the transformation of the digital image into an image R involves also convolving the image I<p> with a kernel K which has a center-surround form. 6. A method according to the previous claim wherein the kernel K is defined by K = F − 1 1 0.81 + 0.2 F G K <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0060" /> wherein F is the Fourier transform, GK is a 2D Gaussian kernel with standard deviation equal to 1/3 of the maximum of the height of the digital image dimension or of the width of the digital image dimension. 7. A method according to any of the previous claims wherein the noise n<r> is computed as: n r = a G c − G s * I N <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0061" /> where I<N> is an image with white Gaussian noise of zero mean and standard deviation of value 1, G<c> and G<s> are Gaussian kernels and a is a constant in the following range: 0 ≤ a ≤ 1. 8. A method according to the previous claim wherein G<c> and G<s> are symmetric Gaussian kernels; G<c> has a standard deviation σc in the following range: 0.02 ≤ σc ≤ 2, and Gs has a standard deviation σS in the following range: 0.01 ≤ σS ≤ 2, and a is in the following range 0.01 ≤ a ≤ 0.5. 9. A method according to the previous claim wherein σC is 0.7 and σS is 1.5 and a is 0.015. 10. A method according to claim 7 wherein G<c> and G<S> are non-symmetric Gaussian kernels; G<c> has the following covariance matrix Σ C = σ cxx 2 0 0 σ cyy 0 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0062" /> wherein σ cxx 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0063" /> is a value in the following range 0.02 ≤ σ cxx 2 ≤ 1 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0064" /> and σ cyy 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0065" /> is a value in the following range 0.02 ≤ σ cyy 2 ≤ 1 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0066" /> and G<s> has the following covariance matrix Σ S = σ sxx 2 0 0 σ syy 2 ∑S wherein σ sxx 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0068" /> is a value in the following range 0.01 ≤ σ sxx 2 ≤ 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0069" /> and σ syy 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0070" /> is a value in the following range 0.01 ≤ σ syy 2 ≤ 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0071" /> and a is in the following range 0.01 ≤ a < 0.5. 11. A method according to the previous claim wherein σ cxx 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0072" /> is 0.2, σ cyy 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0073" /> is 0.05, σ sxx 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0074" /> is 1, σ syy 2 <img class="EMIRef" id="397b523b-f7de-4d81-a528-5c9b3ec21607-ib0075" /> is 0.25 and a is 0.05. 12. A data processing device comprising means for carrying out the steps of any of the methods of claims 1 to 11.