Image Reading Streak Correction Using Color Sensor Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image reading apparatuses face challenges in correcting streak images caused by dust or foreign matter on the platen, which can lead to unwanted deletion of characters, especially when reading forms like invoices, as existing streak image correction methods may erase characters or fail to remove streaks effectively.
Innovation Solution
The apparatus employs a reading unit with color sensors to detect abnormal pixels and perform correction processing, operating in different modes to either correct all colors or focus on specific colors based on the presence of streaks, ensuring accurate image data generation and minimizing character loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If streak image correction is performed to remove foreign matter artifacts, then image quality is improved, but character information may be deleted
Solution Approach 1:
The patent applies different correction strategies to different regions of the image based on local characteristics. The processor determines whether to perform correction at each pixel location by analyzing color component ratios and comparing against reference values, thereby applying correction only where appropriate and preserving character information where needed.
Solution Approach 2:
The patent changes parameters such as color component ratios (R/G, G/B, R/B) and compares them against reference ranges to dynamically determine correction necessity. By monitoring these parameter changes across different image regions, the system selectively applies streak correction while preserving legitimate character information.
2Loss of information
If streak image correction is not performed, then character information is preserved, but streak images remain in the reading image
Solution Approach 1:
The system evaluates each image region locally using color component analysis to determine whether streak correction is needed. By comparing color ratios at each pixel against reference values, the system achieves local quality optimization - removing streaks where present while preserving characters where appropriate.
Solution Approach 2:
The patent implements a feedback mechanism where the processor continuously monitors color component ratios and compares them against reference ranges. This feedback loop enables dynamic decision-making about correction application, allowing the system to adapt to different image regions and maintain both character integrity and image quality.
3Measurement precision
If correction processing is applied to all colors, then streak images are effectively removed, but processing complexity increases
Solution Approach 1:
The patent applies correction selectively based on local color analysis rather than uniformly across all pixels. By calculating color component ratios (R/G, G/B, R/B) and comparing against reference ranges, the system identifies only those regions requiring correction, thereby reducing overall processing complexity while maintaining effective streak removal where needed.
Solution Approach 2:
The system performs partial correction action by applying streak removal only to specific color components and specific image regions where streaks are detected. This partial action approach avoids the excessive processing complexity of applying full correction to all pixels while still achieving effective streak removal in problematic areas.
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 removes streak images while preserving character information, allowing for clear recognition of characters in scanned images, particularly in forms and invoices, by using processor-driven correction modes tailored to the presence and location of dust.
Implementation Method 1
the image reading apparatus receives the reflected light by an image pick up element to convert it into an electrical signal
Data Source
AI summary
An image reading apparatus includes an original tray on which an original is to be stacked; a conveyance unit configured to convey the original stacked on the original tray; a reading unit including a reading sensor having a light receiving element to receive light of a first color and a light receiving element to receive light of a second color that is different from the first color, wherein the reading unit is configured to read an image of the original conveyed by the conveyance unit by using the reading sensor to generate image data which represents a reading result of the reading unit; at least one processor configured to: determine a first abnormal position which is a position in a first direction of an abnormal pixel of the first color in the image represented by the image data, wherein, in the image represented by the image data.