Immutable chain of physical object provenance

A system for verifying the provenance of a collectible, comprising:

• one or more memories;
• and one or more processors communicatively coupled to the one or more memories, configured to:
• process an origin image of a collectible, wherein the processing of the origin image comprises:
• receiving an originator ID information;
• capturing the origin image of the collectible;
• receiving an origin geolocation associated with the collectible;
• storing the originator ID information, the origin geolocation, and the origin image in a datastore;
• and
• identification of an at least one feature of the collectible, wherein the identification of the at least one feature comprises:
• receiving a feature description of each feature identified on the origin image;
• receiving feature location for each feature identified on the origin image;
• and
• storing each feature description and each feature location in the datastore;
• generate a segment mask of the origin image to isolate the collectible from a background,
• wherein the generation of the segment mask comprises:
• transmitting the origin image and at least one prompt to a segmentation component to perform a segmentation analysis;
• receiving a segment image, a segment mask and a segment cutout from the segmentation component;
• storing the segment image, the segment mask and the segment cutout in the datastore;
• designating the segment mask as a golden mask and the origin image as a golden image;
• store the originator ID information, the origin geolocation, the golden image, and each prompt on a blockchain;
• process a candidate image of a candidate object, wherein the processing of the candidate image comprises:
• receiving a verifier ID information of a verifier submitting the candidate image;
• capturing the candidate image of the candidate object;
• receiving a candidate geolocation associated with the candidate object;
• storing the verifier ID information, the candidate geolocation, and the candidate image in the datastore;
• generate a candidate segment mask of the candidate image to isolate the candidate object from the background, wherein the generation of the candidate segment mask comprises:
• transmitting the candidate image and an at least one candidate prompt to the segmentation component to perform a segmentation analysis;
• receiving a candidate segment image, a candidate segment mask and a candidate segment cutout from the segmentation component;
• storing the candidate segment image, the candidate segment mask and the candidate segment cutout in the datastore;
• match the golden image and the candidate image to verify that the collectible and candidate object are the same physical object, wherein the matching comprises:
• submitting the golden image, golden mask, candidate image, and candidate segment mask to a matching component;
• receiving a match image and a confidence factor from the matching component;
• comparing the confidence factor with a confidence threshold;
• and
• if the confidence factor exceeds the confidence threshold, designating the candidate segment mask as the golden mask and the candidate image as a golden image;
• storing the verifier ID information, the candidate geolocation, the golden image, and the match image on the blockchain
• update the datastore and the blockchain when there has been an addition of at least one update feature of the collectible, wherein the updating comprises:
• receiving an updater ID information of an updater submitting an update image of an update object;
• capturing the update image of the update object;
• receiving an update geolocation associated with the update object;
• storing the updater ID information, the update geolocation, and the update image in the datastore;
• identification of an at least one update feature of the collectible, wherein the identification of the at least one update feature comprises:
• receiving an update feature description of each update feature identified on the update image;
• receiving an update feature location for each update feature identified on the update image;
• storing each update feature description and each update feature location in the datastore;
• generation of an update segment mask of the update image to isolate the update object from the background, wherein the generation of the update segment mask comprises:
• transmitting the update image, the golden mask, and each update feature location to the segmentation component to perform a segmentation analysis;
• receiving an update segment image, an update segment mask and an update segment cutout from the segmentation component;
• storing the update segment image, the update segment mask and the update segment cutout in the datastore;
• match the golden image and the update image to verify that the collectible and
• update object are the same physical object, wherein the matching comprises:
• submitting the golden image, golden mask, the update image, and the update segment mask to the matching component;
• receiving an update match image and an update confidence factor from the matching component;
• comparing the update confidence factor with an update confidence threshold;
• and
• if the update confidence factor exceeds the update confidence threshold, designating the update segment mask as the golden mask and the update image as a golden image;
• storing the updater ID information, the update geolocation, the golden image, and the update match image on the blockchain.

A non-transitory computer-readable medium with an executable program stored thereon, wherein the program instructs a processor connected to a network to perform operations comprising:

• processing an origin image of a collectible, wherein the processing of the origin image comprises:
• receiving an originator ID information;
• capturing the origin image of the collectible;
• receiving an origin geolocation associated with the collectible;
• storing the originator ID information, the origin geolocation, and the origin image in a datastore;
• and
• identifying an at least one feature of the collectible, wherein the identifying of the at least one feature comprises:
• receiving a feature description of each feature identified on the origin image;
• receiving feature location for each feature identified on the origin image;
• and
• storing each feature description and each feature location in the datastore;
• generating a segment mask of the origin image to isolate the collectible from a background, wherein the generation of the segment mask comprises:
• transmitting the origin image and at least one prompt to a segmentation component to perform a segmentation analysis;
• receiving a segment image, a segment mask and a segment cutout from the segmentation component;
• storing the segment image, the segment mask and the segment cutout in the datastore;
• designating the segment mask as a golden mask and the origin image as a golden image;
• storing the originator ID information, the origin geolocation, the golden image, and each prompt on a blockchain;
• processing a candidate image of a candidate object, wherein the processing of the candidate image comprises:
• receiving a verifier ID information of a verifier submitting the candidate image;
• capturing the candidate image of the candidate object;
• receiving a candidate geolocation associated with the candidate object;
• storing the verifier ID information, the candidate geolocation, and the candidate image in the datastore;
• generating a candidate segment mask of the candidate image to isolate the candidate object from the background, wherein the generation of the candidate segment mask comprises:
• transmitting the candidate image and an at least one candidate prompt to the segmentation component to perform a segmentation analysis;
• receiving a candidate segment image, a candidate segment mask and a candidate segment cutout from the segmentation component;
• storing the candidate segment image, the candidate segment mask and the candidate segment cutout in the datastore;
• matching the golden image and the candidate image to verify that the collectible and candidate object are the same physical object, wherein the matching comprises:
• submitting the golden image, golden mask, candidate image, and candidate segment mask to a matching component;
• receiving a match image and a confidence factor from the matching component;
• comparing the confidence factor with a confidence threshold;
• and
• if the confidence factor exceeds the confidence threshold, designating the candidate segment mask as the golden mask and the candidate image as a golden image;
• storing the verifier ID information, the candidate geolocation, the golden image, and the match image on the blockchain
• updating the datastore and the blockchain when there has been an addition of at least one update feature of the collectible, wherein the updating comprises:
• receiving an updater ID information of an updater submitting an update image of an update object;
• capturing the update image of the update object;
• receiving an update geolocation associated with the update object;
• storing the updater ID information, the update geolocation, and the update image in the datastore;
• identification of an at least one update feature of the collectible, wherein the identification of the at least one update feature comprises:
• receiving an update feature description of each update feature identified on the update image;
• receiving an update feature location for each update feature identified on the update image;
• storing each update feature description and each update feature location in the datastore;
• generating an update segment mask of the update image to isolate the update object from the background, wherein the generation of the update segment mask comprises:
• transmitting the update image, the golden mask, and each update feature location to the segmentation component to perform a segmentation analysis;
• receiving an update segment image, an update segment mask and an update segment cutout from the segmentation component;
• storing the update segment image, the update segment mask and the update segment cutout in the datastore;
• matching the golden image and the update image to verify that the collectible and
• update object are the same physical object, wherein the matching comprises:
• submitting the golden image, golden mask, the update image, and the update segment mask to the matching component;
• receiving an update match image and an update confidence factor from the matching component;
• comparing the update confidence factor with an update confidence threshold;
• and
• if the update confidence factor exceeds the update confidence threshold, designating the update segment mask as the golden mask and the update image as a golden image;
• storing the updater ID information, the update geolocation, the golden image, and
• the update match image on the blockchain.