	US 12,229,786 B2
	Method for identifying authenticity of an object
Tuomas Kannas, Espoo (FI); Oskari Heikel, Espoo (FI); Hemmo Latvala, Espoo (FI); Nicola Piccinini, Espoo (FI); and Olli-Heikki Paloheimo, Espoo (FI)
Assigned to TRUEMED OY, Espoo (FI)
Appl. No. 17/787,607
Filed by TRUEMED OY, Espoo (FI)
PCT Filed Dec. 22, 2020, PCT No. PCT/FI2020/050863 § 371(c)(1), (2) Date Jun. 21, 2022, PCT Pub. No. WO2021/130413, PCT Pub. Date Jul. 1, 2021.
Claims priority of application No. 20196125 (FI), filed on Dec. 23, 2019.
Prior Publication US 2022/0414902 A1, Dec. 29, 2022
Int. Cl. G06V 20/00 (2022.01); G06Q 30/018 (2023.01); G06T 7/33 (2017.01); G06V 10/75 (2022.01); G06V 10/82 (2022.01)

CPC G06Q 30/018 (2013.01) [G06T 7/337 (2017.01); G06V 10/7515 (2022.01); G06V 10/82 (2022.01); G06V 20/95 (2022.01)]

17 Claims

1. A method for identifying authenticity of an object, wherein the method comprises the steps of:

a) maintaining, in an identification server system, a reference image of an original object, the reference image is provided to represent an equivalent original object;

b) receiving, in the identification server system, one or more input images of the object to be identified;

c) generating, by the identification server system, a target image from the at least one of the one or more input images;

d) aligning, by the identification server system, the target image to the reference image by distorting the target image; and

e) analysing, by the identification server system, the aligned target image in relation to the reference image for identifying authenticity of the object,

wherein step e) comprises:

maintaining, in the identification server system, a machine learning identification algorithm or an identification neural network, and comparing, by the identification server system, the aligned target image to the reference image by utilizing the machine learning identification algorithm or the identification neural network;

providing, in the identification server system, a first machine learning identification algorithm, or a first identification neural network, trained to determine differences between the aligned target image and the reference image, and a second machine learning identification algorithm, or a second identification neural network, specific to the reference image and trained to analyse the differences determined by the first machine learning identification algorithm, or the first identification neural network, in relation to the reference image;

processing, by the identification server system, the aligned target image and the reference image with the first machine learning identification algorithm, or with the first identification neural network, and generating, by the identification server system, a first difference vector by utilizing the first machine learning identification algorithm, or the first identification neural network; and

processing, by the identification server system, the first difference vector with the second machine learning identification algorithm, or with the second identification neural network, specific to the reference image for identifying authenticity of the object.