Means for using microstructure of materials surface as a unique identifier
The present application concerns the visual identification of materials or documents for tracking or authentication purposes. It describes methods to automatically authenticate an object by comparing some object images with reference images, the object images being characterized by the fact that visual elements used for comparison are non-disturbing for the naked eye. In some described approaches it provides the operator with visible features to locate the area to be imaged. It also proposes ways for real-time implementation enabling user friendly detection using mobile devices like smart phones.
1. A method to authenticate a product comprising a parameter settings phase, an acquisition phase and an identification phase;
the parameter setting phase comprising:
selecting an imaging lighting orientation, an imaging resolution and a region of interest on the reference product surface;
the acquisition phase comprising the steps of:
lighting, with an imaging device, the surface of a first reference product item with the imaging lighting orientation;
capturing, with the imaging device, an image of a microstructure of the region of interest on the reference product surface, wherein the image has sufficient resolution to show a least one product surface microstructure feature at the pixel level;
computing, with the imaging device, an image feature to characterize the captured image; and
storing, with the imaging device, the image feature in a database;
the identification phase comprising the steps of:
lighting, with an imaging device, the surface of a product to authenticate with the imaging lighting orientation;
capturing, with the imaging device, an image of a microstructure of the region of interest on the surface of the product to authenticate, wherein the image has sufficient resolution to show a least one microstructure feature at the pixel level;
computing, with the imaging device, an initial image feature to characterize the captured image of the region of interest on the surface of the product to authenticate;
querying, with the imaging device, for a reference product item, a reference image feature in the database;
comparing the initial image feature and the reference image feature, with an image matching technique;
responsive to the comparison, computing, with the imaging device, a refined image feature to characterize the captured image;
measuring a correlation metric between the refined image feature and the reference image feature; and
identifying a match between the product to authenticate and the reference product item when the correlation metric is above a predefined threshold.
2. The method of claim 1 , wherein the image feature is a matrix of values corresponding to a color component sampled along a uniform orthogonal grid.
3. The method of claim 1 , wherein the image feature is a 1D profile.
4. The method of claim 1 , wherein the image feature is a compressed image.
5. The method of claim 1 , wherein the image feature is a set of invariant features.
6. The method of claim 1 , wherein the image feature is a set of Fourier coefficients.
7. The method of claim 1 , wherein the image matching technique is cross-correlation.
8. The method of claim 1 , wherein the image matching technique is applied in the transform domain.
9. The method of claim 1 , wherein the image matching technique is applied in the spatial domain.
10. The method of claim 1 , wherein the image matching technique is a contour matching technique.
11. The method of claim 1 , wherein the image matching technique is a color matching technique.
12. The method of claim 1 , wherein the image matching technique is a fingerprint minutiae matching technique.
13. The method of claim 1 , wherein the refined image feature is compensated for translation.
14. The method of claim 1 , wherein the refined image feature is compensated for rotation.
15. The method of claim 1 , wherein the refined image feature is a flattened version of the captured image.
16. The method of claim 1 , wherein the refined image feature is a combination of several cross-correlations.
17. The method of claim 1 , wherein the refined image feature has a larger size than the initial image feature.
18. The method of claim 1 , wherein the correlation metric is the Signal-to-Noise Ratio (SNR) of the cross-correlation between the refined image feature and the reference image feature.