Method of OCR template enhancement by pixel weighting
First Claim
1. The method of matching an unknown input pixel symbol with a library of enhanced pixel templates by recognition enhancement of a library of L unenhanced pixel images (I1 I2 I3 . . . Ij . . . IL) with respect to a pre-existing group of G unknown input pixel symbols (S1 S2 S3 . . . Si . . . SG) for providing a library of G recognition enhanced pixel templates (T1 T2 T3 . . . Ti . . . TG) of the G pixel symbols, comprising the steps of:
- providing a library of L unenhanced pixel images (I1 I2 I3 . . . Ij . . . IL) to be enhanced to provide a library of G recognition enhanced pixel templates (T1 T2 T3 . . . Ti . . . TG);
providing a pre-existing group of G unknown input pixel symbols (S1 S2 S3 . . . Si . . . SG);
comparing each of the G unknown input pixel symbols with each of the L pixel images to obtain G×
L comparisons forming G sets of L comparisons (C1 C2 C3 . . . Cj . . . CL), one set of L comparisons for each unknown input pixel symbol Si of the G pixel symbols, each set of comparisons having a comparison Ci for each pixel image Ij ;
identifying the primary comparison C* from the L comparisons within each of the G sets of comparisons having the closest comparison with the unknown input pixel symbol Si for that set of comparisons forming a collection of G primary comparisons C* (C1 * C2 * C3 * . . . Ci * . . . CL *);
identifying the secondary comparison C** from the L-1 remaining comparisons within each of the G sets of comparisons having the next closest comparison with the unknown input pixel symbol Si for that set of comparisons forming a collection of G secondary comparisons C** (C1 ** C2 ** C3 ** . . . Ci ** . . . CL **), to provide G pairs of identified comparisons C* and C**, one pair from each of the G sets of comparisons;
determining G recognition margins (M1 M2 M3 . . . Mi MG), one recognition margin between each pair of identified comparisons C* and C**;
selecting the single pair of identified comparisons C* and C** having the smallest recognition margin M* of all of the G pairs of identified comparisons from the G sets of comparisons;
identifying the single pair of pixel images I* and I** corresponding to the pair of identified comparisons C* and C**;
weighting certain pixels of the closest pixel image I* and the next closest pixel image I** corresponding to the selected pair of identified comparisons C* and C** in order to incrementally increase the recognition margin M* therebetween causing the pixel images to become the closest pixel template T* or the next closest pixel template T** or both;
iterating the comparing, identifying, determining, selecting, and weighting steps until the library of pixel images has become a library of enhanced symbol templates (T1 T2 T3 . . . Ti . . . TG) which have been recognition enhanced with respect to the pre-existing group of G unknown input pixel symbols (S1 S2 S3 . . . Si . . . SG), at least some templates of which have weighted pixel abberations not present in the corresponding pixel symbol of the pre-existing group of G unknown input pixel symbols; and
matching an unknown input pixel symbol of the group of G pixel symbols (S1 S2 S3. . . Si . . . SG) with the library of enhanced templates (T1 T2 T3 . . . Ti . . . TG) by comparing the unknown pixel symbol with each of the enhanced pixel templates in the library of enhanced templates and selecting the enhanced template with the closest comparison.
2 Assignments
0 Petitions
Accused Products
Abstract
A library of L unenhanced images (pixel bit maps) is enhanced for optical character recognition (OCR) with respect to a pre-existing group of G input symbols (pixel bit maps) for creating a library of G recognition enhanced templates (pixel bit maps) of the G input symbols. The enhancement is accomplished by comparing each image of the library with each symbol of the group, and weighting the images with the highest potential for confusion. A primary comparison C* and a secondary comparison C** are identified from the L comparisons within each of the G sets of comparisons. A recognition margin is determined between each pair of identified comparisons C* and C**. The single pair of identified comparisons C* and C** is selected forming the smallest recognition margin M*. The single pair of images Ij * and Ij ** underlying the pair of identified comparisons C* and C** is identified. Certain pixels of the closest pixel image Ij * and the next closest pixel image Ij ** are weighted in order to increase the recognition margin M* therebetween. The steps are iterated until the library of pixel images has become a library of sufficiently enhanced symbol templates. The weighted pixel abberations generated in the enhanced templates are not present in the corresponding pixel symbol of the pre-existing group of G pixel symbols.
24 Citations
28 Claims
-
1. The method of matching an unknown input pixel symbol with a library of enhanced pixel templates by recognition enhancement of a library of L unenhanced pixel images (I1 I2 I3 . . . Ij . . . IL) with respect to a pre-existing group of G unknown input pixel symbols (S1 S2 S3 . . . Si . . . SG) for providing a library of G recognition enhanced pixel templates (T1 T2 T3 . . . Ti . . . TG) of the G pixel symbols, comprising the steps of:
-
providing a library of L unenhanced pixel images (I1 I2 I3 . . . Ij . . . IL) to be enhanced to provide a library of G recognition enhanced pixel templates (T1 T2 T3 . . . Ti . . . TG); providing a pre-existing group of G unknown input pixel symbols (S1 S2 S3 . . . Si . . . SG); comparing each of the G unknown input pixel symbols with each of the L pixel images to obtain G×
L comparisons forming G sets of L comparisons (C1 C2 C3 . . . Cj . . . CL), one set of L comparisons for each unknown input pixel symbol Si of the G pixel symbols, each set of comparisons having a comparison Ci for each pixel image Ij ;identifying the primary comparison C* from the L comparisons within each of the G sets of comparisons having the closest comparison with the unknown input pixel symbol Si for that set of comparisons forming a collection of G primary comparisons C* (C1 * C2 * C3 * . . . Ci * . . . CL *); identifying the secondary comparison C** from the L-1 remaining comparisons within each of the G sets of comparisons having the next closest comparison with the unknown input pixel symbol Si for that set of comparisons forming a collection of G secondary comparisons C** (C1 ** C2 ** C3 ** . . . Ci ** . . . CL **), to provide G pairs of identified comparisons C* and C**, one pair from each of the G sets of comparisons; determining G recognition margins (M1 M2 M3 . . . Mi MG), one recognition margin between each pair of identified comparisons C* and C**; selecting the single pair of identified comparisons C* and C** having the smallest recognition margin M* of all of the G pairs of identified comparisons from the G sets of comparisons; identifying the single pair of pixel images I* and I** corresponding to the pair of identified comparisons C* and C**; weighting certain pixels of the closest pixel image I* and the next closest pixel image I** corresponding to the selected pair of identified comparisons C* and C** in order to incrementally increase the recognition margin M* therebetween causing the pixel images to become the closest pixel template T* or the next closest pixel template T** or both; iterating the comparing, identifying, determining, selecting, and weighting steps until the library of pixel images has become a library of enhanced symbol templates (T1 T2 T3 . . . Ti . . . TG) which have been recognition enhanced with respect to the pre-existing group of G unknown input pixel symbols (S1 S2 S3 . . . Si . . . SG), at least some templates of which have weighted pixel abberations not present in the corresponding pixel symbol of the pre-existing group of G unknown input pixel symbols; and matching an unknown input pixel symbol of the group of G pixel symbols (S1 S2 S3. . . Si . . . SG) with the library of enhanced templates (T1 T2 T3 . . . Ti . . . TG) by comparing the unknown pixel symbol with each of the enhanced pixel templates in the library of enhanced templates and selecting the enhanced template with the closest comparison. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28)
-
Specification