Magnetic Ink Character Recognition End Noise Exclusion
Solution Overview
Problem
Conventional magnetic ink character recognition systems face challenges in maintaining high character recognition rates due to paper bending-induced 'end noise' when increasing paper conveyance speed, as existing noise removal technologies do not account for this issue.
Innovation Solution
The system includes a conveyance module, an acquisition module, an excluding module, and a recognition module that exclude a predetermined exclusion section corresponding to the end portion of the paper from the magnetic detection signal, allowing for accurate recognition of magnetic ink characters by isolating and removing 'end noise'.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the conveyance speed of the paper is increased to improve processing speeds, then productivity is improved, but paper bending occurs at the end portion of the paper causing end noise and deterioration in character recognition rate
Solution Approach 1:
The magnetic detection signal is segmented into multiple sections along the paper conveyance direction. The exclusion section corresponding to the end portion of the paper is identified and separated from the effective character recognition sections. This segmentation allows the system to process only the valid signal portions, eliminating end noise interference while maintaining high conveyance speeds.
Solution Approach 2:
The harmful end noise component is extracted and removed from the magnetic detection signal by identifying and excluding the signal section corresponding to the paper's end portion. This extraction process isolates the useful character information from the harmful end noise, enabling accurate recognition even at high conveyance speeds where paper bending occurs.
2Object-affected harmful factors
If conventional noise removal technology is used to remove magnetic noise from the conveyance motor, then magnetic noise is reduced, but end noise from paper bending is not addressed and character recognition rate deteriorates
Solution Approach 1:
Different noise removal approaches are applied to different sections of the magnetic detection signal. Conventional noise removal techniques are applied to sections containing motor noise, while a specific exclusion approach is applied to the end section of the signal where paper bending occurs. This localized differentiation allows effective handling of both noise types without compromising character recognition.
Solution Approach 2:
The system preliminarily identifies the exclusion section corresponding to the paper's end portion before performing character recognition. By pre-marking and excluding this problematic section, the system prevents end noise from interfering with the recognition process, ensuring high accuracy even when paper bending occurs at high conveyance speeds.
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 effectively improves the character recognition rate by eliminating end noise from the magnetic detection signal, thereby enhancing the accuracy of magnetic ink character recognition.
Implementation Method 1
a magnetic head 14 which reads the magnetic ink character MC
Data Source
AI summary
In accordance with an embodiment, a magnetic ink character recognition apparatus comprises a magnetic head; a conveyance module configured to relatively convey a medium on which a magnetic ink character is printed with respect to the magnetic head; an acquisition module configured to acquire a magnetic detection signal of the medium read by the magnetic head; an excluding module configured to exclude a predetermined exclusion section including a reading result of an end portion of the medium from a signal section of the magnetic detection signal; and a recognition module configured to recognize the magnetic ink character based on the magnetic detection signal of the remaining signal section except for the exclusion section.