Matching prints with feature collections

A machine determining how well fingerprints match-comprising:

• a processor:
• a memory storing information representing fingerprints, and storing one or more machine readable instructions, which when implemented cause the processor to implement the method including at least:
• comparing, by the processor, a first pair of features from one a first fingerprint with a first pair of features from a second fingerprint, deriving a transformation that transforms the first pair of features of the first fingerprint to more closely match the first pair of features of the second fingerprint, and storing the transformation in the memory of the machine;
• repeating, by the machine, the comparing and the deriving of the transformation for each of a randomly selected subset of a larger set of pairs of features on the first fingerprint and a set of pairs of features on the second fingerprint, therein forming a set of transformations;
• determining, by the machine, a degree to which each of the transformations of the set of transformations match one another, and finding a subset of the sets of transformations that is the largest group of transformations that match one another within a predetermined tolerance;
• computing a matching score for each transformation in file subset;
• combining, by the machine, the matching scores of each of the transformations of the subset to form a matching score representing a degree to which the first fingerprint matches the second fingerprint;
• and if the matching score for the fingerprints is above a threshold, indicating, by the machine, that the first and second fingerprint match.

A method of measuring how well fingerprints match comprising:

• for each of a set of pairs of features on a first fingerprint determining a cross foliation of at least one feature of one pair of features on the first fingerprint and a cross foliation of at least one feature of a second pair of features;
• determining whether the cross foliation of the at least one feature of the one pair of features on one fingerprint and the cross foliation of the at least one feature of the second pair of feature match within a tolerance;
• comparing, by a machine having a processor and memory unit, one or more pairs of features from the first fingerprint with one or more pairs of features from the second fingerprint wherein the pairs including a pair of features from the first fingerprint and a pair or features from the second fingerprint, that have been compared, if the one or more pairs of features from the first fingerprint have a cross foliation that matches a cross foliation of a feature of a pair of features on the second fingerprint, the one or more pairs of fingerprints, and compared based on one or more of the following:
• the distance between the pairs of features on the first finger print;
• the number of valleys between the pairs of features on the first finger print;
• angles of features in the first fingerprint and the second fingerprint;
• predetermined feature types of features in the pairs of features of the first fingerprint;
• the method also includes at least deriving a transformation that matches one pair of features from the first fingerprint with another pair of features from the second fingerprint;
• repeating, by the machine, the comparing and the deriving of the transformation for each of a set of pairs of features on the first fingerprint and a set of pairs of features on the second fingerprint, therein forming a set of transformations;
• determining, by the machine, a degree to which each of the transformations of the set of transformations match one another, and finding a subset of the set of transformations that is the largest group of transformations that match one another within a predetermined tolerance;
• computing a matching score for each transformation in the subset;
• combining, by the machine, the matching scores of each of the transformations of the subset to form a matching score representing a degree to which the first fingerprint matches the second fingerprint;
• and
• if the matching score for the fingerprints is above a threshold, indicating, by the machine, that the first and second fingerprint match.