Re-identification of rough gemstones

1. A method of re-identifying a rough gemstone, the method comprising:

providing a three-dimensional (3D) model of a first rough gemstone;

generating a series of virtual two-dimensional (2D) silhouette images of the 3D model;

processing each 2D image of the series of virtual 2D silhouette images to obtain a dataset associated with the first rough gemstone;

comparing the dataset of the first rough gemstone with an existing dataset of a rough gemstone; and

where the dataset of the first rough gemstone and the existing dataset match each other, re-identifying the first rough gemstone as the same rough gemstone from which the existing dataset was obtained.

2. The method as claimed in claim 1 , further comprising capturing a virtual 2D silhouette image of the 3D model from each of a plurality of virtual viewpoints in a geodesic arrangement around the 3D model.

3. The method as claimed in claim 1 , wherein processing each 2D image of the series of virtual 2D silhouette images comprises converting each said 2D image into a set of geometric shapes that characterise a silhouette of the first rough gemstone present in said 2D image.

4. The method as claimed in claim 3 , wherein the set of geometric shapes is based upon one or more of a longest chord of the silhouette, an orthospread, and a convex hull deviance.

5. The method as claimed in claim 1 , wherein the dataset comprises a series of numerical features derived from each of the 2D images of the series of virtual 2D silhouette images.

6. The method as claimed in claim 5 , wherein the numerical features derived from each of the 2D images comprise thirty-two features to which four statistical operations are applied.

7. The method as claimed in claim 5 or 6 , wherein the numerical features include a measurement of one or more of blockiness, symmetry, normalised inverted longest chord, Hu's moments, orthonormal dispersion ratio, convex hull area ratio, standard convex hull deviance, normalised convex hull deviance, convex hull deviance ratio, normalised missing perimeter lengths, concave segment perimeter ratio, area and perimeter, longest chord, and orthospread.

8. The method as claimed in claim 7 , wherein the measurement is obtained from one or more of the rough gemstone and a convex hull of the rough gemstone.

9. The method as claimed in claim 5 , further comprising calculating one or more of the maximum, minimum, mean and coefficient of variance of each of the series of numerical features.

10. The method as claimed in claim 1 , wherein one or more of: the dataset associated with the first rough gemstone; the existing dataset; and a result of the comparison between the datasets forms part of a distributed ledger, such as a blockchain.

11. The method as claimed in claim 1 , wherein providing the 3D model of the first rough gemstone comprises:

obtaining a series of 2D silhouette images of the rough gemstone using one or more physical image capture devices; and

processing said series of 2D silhouette images to generate the 3D model of the rough gemstone.

12. The method as claimed in claim 11 , wherein said one or more physical image capture devices are stationary while the series of 2D silhouette images are obtained.

13. The method as claimed in claim 11 , further comprising illuminating the rough gemstone with collimated lighting while obtaining the series of 2D silhouette images.

14. The method as claimed in claim 1 , wherein the 3D model is a surface model of the rough gemstone.

15. The method as claimed in claim 1 , further comprising obtaining the existing dataset as part of a process of sorting the output of a mine, and wherein re-identifying the rough gemstone comprises confirming that the rough gemstone is originated from the mine.

16. The method as claimed in claim 1 , wherein the rough gemstone is a rough diamond.

17. The method as claimed in claim 1 , further comprising comparing the dataset of the first rough gemstone with a plurality of existing datasets of rough gemstones stored in a database.