Method for identification of person by recognition of a digital fingerprint

1. A method of identifying a person by fingerprint/toeprint recognition, comprising the steps of:

producing a digital photograph having a digitized image of a fingerprint/toeprint or part of a fingerprint/toeprint present on a surface of an object;

analyzing said digitized image of the fingerprint/toeprint;

detecting characteristic points of the fingerprint/toeprint in the digitized image;

exchanging detected digital data of the detected characteristic points with a data bank in which digital data of the characteristic points of a multiplicity of fingerprints/toeprints are stored in a memory, said stored digital data corresponding to plane images of the multiplicity of fingerprints/toeprints;

comparing the detected digital data of the abovementioned detected characteristic points with the digital data stored in the memory of the data bank; and

identifying a person having said fingerprint/toeprint as a result of the above comparing step,

said method further comprising, where the fingerprint/toeprint is present on a curved surface, the steps of

selecting a pre-established model of a curved semicylindrical or semiconical surface of revolution having a shape corresponding or close to a shape of the curved surface or of a portion of the curved surface on which the fingerprint/toeprint is present,

sending information about respective positions of two diametrically opposed generatrices of the curved surface which are visible in the digitized image to an algorithm processing device, whereby the algorithm processing device deduces from the information relevant geometrical characteristics of the curved surface, and

transforming, using the algorithm processing device for a plane projection, the digitized image into a corrected digitized image with a distortion level below a predetermined threshold, said corrected digitized image showing the characteristic points of said fingerprint/toeprint in a plane,

whereby said corrected digitized image is used for detecting the characteristic points of the fingerprint/toeprint in the detecting step and for identifying the person in the indentifying step.

2. The method as claimed in claim 1 , wherein the transforming step includes the step of associating, by use of the algorithm processing device, each point in the digitized image of the curved fingerprint/toeprint present on the curved surface with a projected point lying in a projection plane such that the linear distance of said projected point in the projection plane relative to the projection of the axis of said curved surface is equal to the curvilinear distance of the point in the digitized image relative to the projection of said axis onto said curved surface.

3. The method as claimed in claim 2 , wherein said transforming step includes, at any point P in the projection plane, determining with the algorithm processing device a projected point P 1 such that:

O 1 ·P 1 =r· cos(π/2− O 1 ·P/r )

where O 1 is the projection of the axis of the curved surface onto said projection plane, and

where r is an estimated radius of the curved surface, and then associating the projected point P 1 in the plane with a point P 2 on the curved surface such that the projected point P 1 is the projection of the point P 2 on the projection plane.

4. The method as claimed in claim 1 , where the fingerprint/toeprint present on a surface of complex shape,

wherein said transforming step includes the steps of

breaking the digitized image of said surface of complex shape down into partial images of surfaces of simple shape,

processing each partial image in relation to the shape of the respective surface in order to obtain corrected partial images, and

juxtaposing the corrected partial images so as to obtain a corrected full image of the fingerprint/toeprint like a mosaic.