Authentication and identification of physical objects using immutable physical code

A computer-program product tangibly embodied in a non-transitory machine-readable storage medium, including instructions configured to cause a data processing apparatus to perform operations, for machine authentication of a physical object using a computer vision model and machine learning (ML) algorithms for comparing attributes of the physical object, the operations comprising:

• capturing an image of the physical object with computer vision;
• focusing the computer vision on the physical object based on information of the physical object in the image;
• capturing, in response to focusing the computer vision, a first analysis image of the physical object with the computer vision, wherein the first analysis image is covering less area than the image;
• segmenting the first analysis image into a plurality of select patches, wherein a first set of features is determined from a select patch of the plurality of select patches;
• determining the first set of features associated with the first analysis image using a ML model, wherein the ML model comprises a first set of functions for the first set of features;
• generating a first feature-vector for the physical object based on the first set of features comparing the first feature-vector with a second feature-vector, wherein the second feature-vector is generated from a second set of features that is associated with a second analysis image;
• and
• generating, in response to comparing the first feature-vector with the second feature-vector, a notification associated with a status for machine authentication of the physical object, wherein the select patch has a length of at least 1 cm and a width of at least 1 cm, the first analysis image has a resolution of at least 800 pixels per inch (ppi), and a count of features in the first set of features is at least ten.