Currency discrimination
AssigneeMEI, Inc.
InventorFatiha Anouar
Life of this patent
Estimating the term…
Abstract
A reference data set used for classifying items of currency is established by obtaining at least one measured response from at least one known item of currency, projecting the measured response data from the known item of currency from a first space to a second space, and applying a reduction technique to the second space thereby reducing the complexity of the second space. The reference data set, which can be used for classifying items of currency, is thus represented by a subset of all measured responses obtained from the known item of currency.
Claims21
A method for implementation by one or more data processors forming part of at least one computing system and for establishing at least one reference data set used for classifying items of currency, the method comprising:
- obtaining at least one measured electromagnetic response from at least one known item of currency;
- projecting, using at least one data processor, the measured electromagnetic response from the at least one known item of currency from a first multi-dimensional space to a second multi-dimensional space;
- applying, using at least one data processor, at least one axis reduction technique to the second multi-dimensional space thereby selecting a subset of axes of the second space, wherein the at least one reference data set used for classifying items of currency is represented by a subset of all measured responses obtained from the at least one known item of currency;
- and
- applying, using at least one data processor, at least one variable reduction technique to the subset of axes of the second space, wherein the at least one variable reduction technique obtains an approximation of the original subset of axes of the second space.
A method of distinguishing between items of currency for implementation by a currency validation unit, the method comprising:
- sensing, using the currency validation unit, an inserted item of currency to obtain electromagnetic response information corresponding to a reduced set of measured electromagnetic responses on the inserted item of currency, the reduced set of measured electromagnetic responses defined by at least one reference data set for a known item of currency established using a multi-step reduction technique, wherein the multi-step reduction technique includes projecting the measured electromagnetic response information from a first multi-dimensional space to a second multi-dimensional space, selecting a subset of axes of the second multi-dimensional space, and applying a variable reduction technique to the subset of axes of the second multi-dimensional space to obtain an approximation of the subset of axes of the second multi-dimensional space;
- weighting data from the measured electromagnetic responses using weighting factors defined by at least one reference data set for each respective response;
- comparing the weighted response data set of the inserted item of currency to the at least one reference data set, and
- characterizing the inserted item of currency as a member of the class corresponding to the at least one known item of currency when the comparison is within a predefined tolerance or characterizing the inserted item of currency as not a member of the class corresponding to the at least one known item of currency when the comparison is outside of a predetermined tolerance.
An apparatus for discriminating items of currency comprising:
- a memory unit storing at least one reference data set for at least one known item of currency, the at least one reference data set representing measured electromagnetic responses from sensing the at least one known item of currency, the at least one measured electromagnetic response on the at least one known item of currency having been identified using a multi-step reduction technique, wherein the multi-step reduction technique includes projecting the measured electromagnetic response information from a first multi-dimensional space to a second multi-dimensional space, selecting a subset of axes of the second multi-dimensional space, and applying a variable reduction technique to the subset of axes of the second multi-dimensional space to obtain an approximation of the subset of axes of the second multi-dimensional space;
- a sensing unit to sense inserted items of currency, wherein the sensing unit is arranged to sense an inserted item of currency to obtain measured electromagnetic response information from the inserted item of currency based on the corresponding measured electromagnetic response information defined in the at least one reference data set;
- a processing unit to compare an output from the sensing unit to the at least one reference data, and to characterize the inserted item of currency as a member of the class corresponding to the at least one known item of currency if the comparison of the sensing unit output and the at least one reference data set falls within a predetermined tolerance, or to characterize the inserted item of currency as not a member of the class corresponding to the at least one known item of currency if the comparison of the sensing unit output and the reference data set falls outside of a predetermined tolerance.
Dependencies are inferred from the claim text.