Reading Device Backside Reflection Suppression Using Dual-Wavelength Light
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image reading devices face challenges in effectively removing backside reflection from documents, especially when reading duplex-print documents, as the reflection light from the background plate is absorbed by the print on the back side, causing backside reflection in the read image.
Innovation Solution
A reading device is equipped with a first and second light source having different transmittances, which irradiate the document, and a comparison unit that detects data differences between the images obtained from both light sources to correct areas with significant differences, thereby suppressing backside reflection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single light source is used to read documents, then the reading process is simple and fast, but backside reflection cannot be effectively removed from the read image
Solution Approach 1:
The reading process is segmented into two separate reading operations: one using a first light source (visible light) to capture the front side image, and another using a second light source (infrared light) to capture the backside reflection image. This segmentation allows independent optimization of each reading operation for its specific purpose, resolving the contradiction between reading speed and image quality by performing both readings in sequence rather than requiring a complex simultaneous dual-light system
Solution Approach 2:
The patent introduces an intermediary processing step where the backside reflection image (captured with infrared light) is used as reference data to remove backside reflection artifacts from the front side image (captured with visible light). This intermediary approach allows the system to achieve high-quality front side images by using the infrared image as a mediator to identify and remove unwanted backside reflections
2Measurement precision
If separate images are acquired to remove backside reflection, then backside reflection can be removed from the read image, but processing time increases
Solution Approach 1:
The patent implements continuous useful action by using the infrared light reading not as a separate post-processing step, but as a concurrent operation that provides real-time backside reflection data. The infrared image is captured continuously during the reading process, and this data is immediately used to remove backside reflection from the front side image, maintaining continuous productive action rather than sequential batch processing
Solution Approach 2:
The patent changes the light wavelength parameter from visible light to infrared light for the backside reflection detection. This parameter change allows the system to penetrate the document and detect backside reflections without interfering with the visible light reading process, enabling simultaneous or near-simultaneous operation that reduces overall processing time while maintaining image quality
3Device complexity
If luminance histogram is used to remove backside reflection, then processing is simplified, but accuracy decreases for documents with varying luminance distribution
Solution Approach 1:
The patent replaces the mechanical/mathematical approach of luminance histogram analysis with an optical approach using infrared light. Instead of processing the visible light image through complex histogram calculations and thresholding operations, the system uses infrared light to directly illuminate and detect backside reflections, substituting a simpler optical measurement for a complex image processing algorithm
Solution Approach 2:
The patent creates a copy of the backside reflection information using infrared light, which then serves as a template or reference for removing backside reflections from the visible light image. This copying approach allows the system to preserve the original visible light image while creating and using a separate infrared-based reference image for accurate backside reflection identification and removal
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
The solution allows for accurate detection and correction of backside reflection in a single reading operation, reducing processing time and improving image quality by distinguishing between areas with and without backside reflection.
Implementation Method 1
a first light source (13a), a second light source (13b), the first light source irradiates an object to be read, with first light having a first transmittance to the object
Implementation Method 2
the second light source irradiates the object with second light having a second transmittance to the object, the second transmittance being different from the first transmittance
Implementation Method 3
The first reading unit reads the first light of the first light source from the object to output first image data
Implementation Method 4
The second reading unit reads the second light of the second light source from the object to output second image data
Data Source
AI summary
A reading device includes a first light source, a second light source, a first reading unit, a second reading unit, a comparison unit, and a correction unit. The first source irradiates an object to be read, with first light having a first transmittance to the object. The second source irradiates the object with second light having a second transmittance to the object. The first reading unit reads the first light from the object to output first image data. The second reading unit reads the second light from the object to output second image data. The comparison unit compares a data difference between the first and second image data. In a case where the data difference is equal to or larger than a threshold, the correction unit corrects data of an image area having the data difference equal to or larger than the threshold in the first or second image data.