Methods and systems for automatic object recognition and authentication
A system and a method of receiving object data representing one or more discriminating characteristics of a physical object or group of physical objects is described herein. The method includes: processing the object data by means of a machine-learning-based object recognition process to obtain discriminating data representing one or more collision resistant virtual representations of the physical object or group of physical objects; comparing at least one of the discriminating data and an original hash value derived therefrom by application of a pre-determined cryptographic hash function thereto with corresponding reference data stored in one or more data repositories with restricted access; and, if said comparison with the reference data results in a match, outputting digitally signed identification data comprising said hash value.
1. A method of training a system for automatic object recognition, the method comprising:
receiving object data representing one or more discriminating characteristics of a physical object or group of physical objects;
processing the object data by means of a machine-learning-based object recognition process to obtain original discriminating data by generating, by means of performing one or more predefined modification operations, a plurality of collision resistant virtual representations of said physical object or group of physical objects, such that at least two of these virtual representations represent the same physical object or group of physical objects, but each at a different condition, wherein said different conditions relate to different age points that represent aging effects in a lifetime of said physical object or group of physical objects; and
storing reference data comprising the original discriminating data and an original hash value derived therefrom by application of a pre-determined cryptographic hash function thereto into one or more data repositories with restricted access.
2. The method of claim 1 , wherein storing the reference data comprises storing the original discriminating data into a first one of the data repositories and storing identification data comprising the original hash value into a separate second one of the data repositories such that neither the first nor the second data repository stores both the original discriminating data and the corresponding original hash value.
3. The method of claim 1 , wherein storing the reference data comprises storing combined data, which is formed as a combination of the reference data and an additional adjunct value, which is defined independent from said reference data, into the one or more data repositories with restricted access.
4. The method of claim 1 , further comprising iteratively re-training the system for automatic object authentication based on additional object data representing discriminating characteristics of the physical object or group of physical objects, for which respective original reference data has already been previously generated and stored in the one or more data repositories during a previous iteration step of training the system, wherein the additional object data serves as input data for said machine-learning-based object recognition process.
5. The method of claim 1 , wherein storing the reference data comprises one or more of the following steps:
storing the reference data in a digitally signed form into at least one of said data repositories;
storing or causing one or more nodes of a blockchain environment to store the identification data in a digitally signed form into a block of a blockchain related to said blockchain environment;
storing or causing one or more nodes of a blockless distributed ledger environment to store the identification data in a digitally signed form into at least one node of said blockless distributed ledger environment;
storing the identification data in a digitally signed form into a storage of a public/private-key environment.
6. The method of claim 1 , wherein the method is performed by a system comprising two or more separate devices, which collectively perform the method, and which are communicatively connected to each other by one or more data links being protected against unauthorized interception.
7. A system for automatic object recognition, the system being configured to perform the method of claim 1 .
8. A non-transitory computer-readable medium having stored thereon computer-executable instructions which, when executed on one or more processors of a system for automatic object recognition, causes the system to perform the method according to claim 1 .
9. The method of claim 1 , wherein said different conditions further relate to different environmental conditions to which said physical object or group of physical objects might be exposed during a detection of respective one or more discriminating characteristics thereof in a subsequent object authentication process.