Frequency domain based MICR reader

1. A frequency domain based magnetic ink character recognition (MICR) system, comprising:

a computer device having a memory and processor, comprising:

a segmentation system for segmenting inputted MICR data into sets of temporal data for inputted characters;

a Fourier system for generating a set of Fourier components from temporal data for an inputted character, wherein the set of Fourier components comprises six components including a fundamental and next five harmonics;

a normalization system for normalizing the set of Fourier components to generate a normalized set of Fourier components, wherein the normalized set of Fourier components include magnitudes that sum to a predetermined value; and

a matching system for comparing the normalized set of Fourier components with each of a set of reference waveforms to determine an identity of the inputted character, wherein the set of reference waveforms comprises a corresponding six components including a corresponding fundamental and corresponding next five harmonics, and wherein the comparing, comprises:

calculating a set of difference value between each component in the normalized set of Fourier components and an associated component in a reference waveform;

weighting each difference value such that a difference value calculated for a relatively higher Fourier component receives a relatively lower weight;

summing each of the weighted difference values;

comparing concavities between the normalized set of Fourier components and each set of reference waveforms; and

assigning values based upon a degree of concavity matching.

2. The frequency domain MICR system of claim 1 , wherein the inputted MICR data is obtained from a single gap MICR reader.

3. The frequency domain MICR system of claim 1 , wherein the Fourier system filters out a set of higher frequency Fourier components.

4. The frequency domain MICR system of claim 1 , wherein the matching system includes a reference waveform for each character in a character set.

5. The frequency domain MICR system of claim 4 , wherein the matching system calculates a difference between the normalized set of Fourier components and each reference waveform.

6. The frequency domain MICR system of claim 4 , further comprising a reference generation system for generating each of the reference waveforms.

7. A computer program product stored on a computer readable non-transitory storage medium for implementing a frequency domain based magnetic ink character recognition (MICR) system, comprising:

program code configured for generating a set of Fourier components from temporal MICR data for an inputted arbitrary character, wherein the set of Fourier components comprises six components including a fundamental and next five harmonics;

program code configured for normalizing the set of Fourier components to generate a normalized set of Fourier components; and

program code configured for comparing the normalized set of Fourier components with each of a set of reference waveforms to determine an identity of the inputted arbitrary character, wherein the set of reference waveforms comprises a corresponding six components including a corresponding fundamental and corresponding next five harmonics, and wherein the comparing, comprises:

calculating a set of difference value between each component in the normalized set of Fourier components and an associated component in a reference waveform;

weighting each difference value such that a difference value calculated for a relatively higher Fourier component receives a relatively lower weight;

summing each of the weighted difference values;

comparing concavities between the normalized set of Fourier components and each set of reference waveforms; and

assigning values based upon a degree of concavity matching.

8. The computer program product of claim 7 , further comprising program code configured for segmenting inputted MICR data into discrete sets of temporal data for individual characters.

9. The computer program product of claim 7 , wherein the normalized set of Fourier components include magnitudes that sum to a predetermined value.

10. The computer program product of claim 7 , wherein a reference waveform is provided for each character in a character set.

11. The computer program product of claim 10 , wherein the program code configured for comparing the normalized set of Fourier components with each of a set of reference waveforms calculates a difference between the normalized set of Fourier components and each reference waveform.

12. The computer program product of claim 10 , further comprising program code configured for generating each of the reference waveforms.

13. A method of implementing a frequency domain based magnetic ink character recognition (MICR) system, comprising:

generating a set of Fourier components from temporal MICR data for an inputted arbitrary character using a computer device, wherein the set of Fourier components comprises six components including a fundamental and next five harmonics;

normalizing the set of Fourier components to generate a normalized set of Fourier components; and

comparing the normalized set of Fourier components with each of a set of reference waveforms to determine an identity of the inputted arbitrary character, wherein the set of reference waveforms comprises a corresponding six components including a corresponding fundamental and corresponding next five harmonics, and wherein the comparing comprises:

calculating a set of difference value between each component in the normalized set of Fourier components and an associated component in a reference waveform;

weighting each difference value such that a difference value calculated for a relatively higher Fourier component receives a relatively lower weight;

summing each of the weighted difference values;

comparing concavities between the normalized set of Fourier components and each set of reference waveforms; and

assigning values based upon a degree of concavity matching.

14. The method of claim 13 , further comprising the initial steps of:

collecting MICR data from a single gap MICR reader; and

segmenting collected MICR data into discrete sets of temporal data for individual characters.

15. The method of claim 13 , wherein the normalized set of Fourier components include magnitudes that sum to a predetermined value.

16. The method of claim 13 , wherein a reference waveform is provided for each character in a character set.

17. The method of claim 13 , further comprising the step of generating each of the reference waveforms.

18. A method for deploying an application for implementing a frequency domain based magnetic ink character recognition (MICR) system, comprising:

providing a computer infrastructure being operable to:

generate a set of Fourier components from temporal MICR data for an inputted arbitrary character, wherein the set of Fourier components comprises six components including a fundamental and next five harmonics;

normalize the set of Fourier components to generate a normalized set of Fourier components; and

compare the normalized set of Fourier components with each of a set of reference waveforms to determine an identity of the inputted arbitrary character, wherein the set of reference waveforms comprises a corresponding six components including a corresponding fundamental and corresponding next five harmonics, and wherein the comparing comprises:

calculating a set of difference value between each component in the normalized set of Fourier components and an associated component in a reference waveform;

weighting each difference value such that a difference value calculated for a relatively higher Fourier component receives a relatively lower weight;

summing each of the weighted difference values;

comparing concavities between the normalized set of Fourier components and each set of reference waveforms;

assigning values based upon a degree of concavity matching.