Magnetic Character Recognition Waveform Shifting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for magnetic ink character recognition face challenges in accurately reading characters with variations in printing quality and deformation, leading to high waveform deformation and increased rejection rates, especially when characters are poorly printed or deformed.
Innovation Solution
A method and apparatus that prepares standard waveforms for magnetic ink characters, generates regeneration waveforms, segments character waveforms, compares them with standard waveforms, shifts waveforms to calculate a coefficient of coincidence, and selects the candidate character with the greatest coincidence value, using additional standard waveforms for thick-lined, thin-lined, and tilted characters to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a magnetic signal is used for character recognition by comparing waveform with standard waveform, then character recognition can be performed, but waveform deformation due to printing variations and character alterations causes false recognition and increased rejection rate
Solution Approach 1:
The patent applies preliminary action by performing waveform normalization and preprocessing before comparison. The system pre-processes both the input waveform and standard waveforms to eliminate printing variations and deformations, including adjusting for line thickness, tilt, and magnetic force variations. This preliminary preparation ensures that the subsequent comparison operates on standardized data, reducing false recognition caused by physical character variations.
Solution Approach 2:
The patent employs parameter changes by dynamically adjusting comparison parameters based on detected waveform characteristics. The system modifies comparison thresholds and parameters according to the specific waveform features, such as peak positions, amplitude variations, and time intervals. This adaptive parameter adjustment allows the recognition system to accommodate different printing qualities and character deformations while maintaining high accuracy.
2Reliability
If expansion and contraction of standard waveform is performed to increase collation ratio, then poor collation ratio can be improved, but excessive waveform deformation makes correction impossible and causes heavy operation load
Solution Approach 1:
The patent applies partial action by performing waveform expansion and contraction only on standard waveforms that fall within a predefined deformation range. The system calculates a deformation index for each standard waveform and selectively applies correction only when the index indicates moderate deformation. This selective approach avoids the heavy computational load of processing all waveforms while still improving collation ratio for cases where correction is beneficial and possible.
Solution Approach 2:
The patent uses copying by creating multiple candidate standard waveforms with different deformation levels for comparison. Instead of heavily deforming a single standard waveform, the system generates several copies with varying expansion and contraction ratios. The input waveform is then compared against these multiple copies, and the best match is selected. This approach distributes the computational effort across simpler operations rather than requiring one complex deformation correction.
3Ease of operation
If absolute values of differences at all corresponding points between objective recognition waveform and standard waveform are within predefined range, then character recognition can be performed, but this assumption fails when waveform deformation exists
Solution Approach 1:
The patent applies dynamics by making the comparison process adaptive rather than static. The system dynamically adjusts the predefined range thresholds based on the actual waveform characteristics being compared. When waveform deformation is detected, the comparison algorithm automatically expands the acceptable difference range for corresponding points. This dynamic adjustment maintains the simplicity of the comparison operation while increasing adaptability to handle various printing variations and character deformations.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables high-accuracy magnetic character recognition even with waveform deformation, reducing false recognition and operation load by shifting waveforms to optimize coincidence values and using additional standard waveforms for deformed characters.
Implementation Method 1
reading magnetic ink characters, printed with a magnetic ink, by using a magnetic head so as to obtain a magnetic regeneration signal
Data Source
AI summary
A method for magnetic character recognition is disclosed. The method may include preparing a standard waveform that is used as a datum in an operation of reading magnetic ink characters, generating a regeneration waveform from a character string of the magnetic ink characters printed on a surface of an information recording medium, segmenting a character waveform of each of the magnetic ink characters from the regeneration waveform, comparing the character waveform segmented through the segmentation process with the standard waveform, and selecting a plurality of the standard waveforms of candidate characters in accordance with a comparison result of the comparison process. A read character may be identified with the candidate character that has the greatest value among all the coefficient values of coincidence.