CIS Faulty Line Detection and Template Segmentation for Paper Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated teller machines (ATMs) face a decrease in acceptance rate and customer experience due to the 'CIS faulty line phenomenon', where a faulty light emitting element or dirty transparent piece of the contact image sensor (CIS) causes random black lines in images, affecting paper medium identification.
Innovation Solution
A device and method that includes an image data acquiring part, faulty line detecting part, image cutting part, standard template storage, comprehensive analysis part, and new template generation to cut the standard template based on the faulty line's position, generating sub-templates for unaffected regions, and performing template match identification using these sub-templates to avoid faulty line influence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If template match identification is performed using the complete image including faulty lines, then the identification process is simple, but the acceptance rate decreases and identification accuracy deteriorates
Solution Approach 1:
The patent divides the image into multiple line segments and identifies faulty lines separately. By segmenting the image processing into line-level analysis and region-level identification, the system can exclude faulty lines while maintaining overall process simplicity. This resolves the contradiction by adding minimal complexity at the line segmentation stage to prevent major reliability losses in the identification stage.
Solution Approach 2:
The patent extracts and removes faulty lines from the image before performing template match identification. By taking out the harmful faulty line data, the identification process uses only clean image data, maintaining high acceptance rate and accuracy without requiring complex fault tolerance mechanisms in the identification algorithm itself.
2Device complexity
If the complete image is used for template match identification, then processing is straightforward, but identification accuracy deteriorates due to faulty lines
Solution Approach 1:
The patent performs preliminary detection and removal of faulty lines before the template match identification process. This preliminary action ensures that the identification algorithm receives clean image data without faulty lines, maintaining high identification accuracy while keeping the overall processing complexity low through a simple two-stage approach.
3Reliability
If faulty lines are detected and handled, then acceptance rate improves, but processing time increases
Solution Approach 1:
The patent segments the image into horizontal lines and processes each line independently for fault detection. This segmentation allows for efficient parallel processing and minimizes the time overhead of faulty line detection, as the system only needs to analyze one-dimensional line data rather than the complete two-dimensional image, thus improving acceptance rate without significant time penalty.
4Measurement precision
If the standard template is cut into sub-templates based on faulty lines, then identification accuracy improves, but device complexity increases
Solution Approach 1:
The patent segments both the image and the standard template into corresponding regions that exclude faulty line areas. By cutting the template into sub-templates that align with the valid image regions, the system maintains high identification accuracy through precise template matching while keeping the complexity manageable through systematic segmentation rather than complex adaptive processing.
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
Significantly improves the acceptance rate of paper medium identification by effectively handling faulty lines, ensuring accurate identification results and maintaining device performance.
Implementation Method 1
an image data acquiring part of the device for identifying the paper medium is configured to read the paper medium optically, acquire grayscale image data
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A paper medium identifying device (100) and an identifying method. The paper medium identifying device (100) comprises an image data obtaining unit(10), a faulty wire detecting unit(20), an image dividing unit(30), a standard template data storage unit (40), a comprehensive analyzing unlit(50), a new template, generating unit (60)and a judging unit(70). The paper medium identifying device (100) divides the standard template into new sub-templates by dividing the template from a faulty wire position as margin, and then matches the sub-templates with a papery medium image which being identified so as to avoid the influence of faulty wires on the template, match identification and improve the acceptance rate of the papery medium identifying device.