Image Reading Apparatus Dirt Detection via Shading Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image reading apparatuses face challenges in accurately detecting dirt substances on the imaging device, leading to noise lines in captured images, especially when dirt substances like paper dust or glue adhere to the glass surface, affecting image quality and requiring improved detection methods.
Innovation Solution
The implementation of an image reading apparatus and system that generates a white reference image and a document image, utilizing processing steps to detect dirt substances through shading correction, allowing for accurate identification and correction of noise lines by comparing image data from both sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If dirt substance detection is performed only from the white reference image, then the detection process is simple, but the detection accuracy is insufficient because dirt substances may be missed or false detections occur
Solution Approach 1:
The detection process is segmented into two independent detection stages: first detection from the white reference image, and second detection from the corrected document image. Each stage uses different image data and correction parameters, allowing comprehensive dirt detection while maintaining manageable complexity in each individual detection step.
Solution Approach 2:
Shading correction data acts as an intermediary element between the white reference image and the final document image. The correction data generated from the white reference image is applied to the document image to create a corrected version that enhances dirt visibility, enabling more accurate second detection without directly modifying the original document image.
2Measurement precision
If dual processing (first processing from white reference image and second processing from corrected image) is performed, then dirt substance detection accuracy is improved, but processing time increases
Solution Approach 1:
Shading correction is performed as a preliminary action before the second dirt detection process. The correction data is generated in advance from the white reference image and applied to the document image, preparing the corrected image for more effective dirt detection without adding significant processing time during the critical detection phase.
Solution Approach 2:
The detection process maintains continuity by using the corrected document image that already contains enhanced dirt visibility from the shading correction. The second detection process continues the useful action of dirt identification without restarting the imaging or correction process, reducing redundant processing time.
3Reliability
If shading correction is applied to the document image, then image quality is improved for detection, but processing complexity increases
Solution Approach 1:
Shading correction data serves as an intermediary that bridges the white reference image and the document image. Instead of directly modifying the document image with complex transformations, the correction data (which contains shading information) is applied as a separate processing layer, simplifying the overall processing architecture while maintaining image quality.
Solution Approach 2:
The shading correction creates a corrected copy of the document image rather than modifying the original. This copying approach allows the original document image to remain unchanged while providing an enhanced version for detection purposes, reducing processing complexity by working with a derived copy instead of repeatedly processing the original.
Data Source
AI summary
An image reading apparatus includes an imaging device for generating a white reference image of a white reference member and a document image of a document and a periphery of the document, and a processor for performing first processing for detecting a dirt substance from the white reference image, generating data for shading correction based on the white reference image, correcting the document image using the data for shading correction to generate a correction image, and performing second processing for detecting a dirt substance from the correction image. One of the first processing or the second processing is performed using a dirt substance detection result of the other one of the first processing or the second processing.