Image Reading Device Document Area Discrimination Using Pattern Cycle
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image reading devices face challenges in accurately detecting the document area using stripe patterns, as finer stripes require more vertices for detection, leading to potential errors due to difficulty in specifying vertices.
Innovation Solution
The image reading device employs a method where it considers three or more continuous pixels as a single unit and determines the presence or absence of a short cycle fluctuation in brightness values specific to the pattern. This allows for accurate discrimination between the document area and the background area based on the switch in the presence or absence of this fluctuation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the stripe pattern is made finer to increase vertex detection accuracy, then the document area detection accuracy is improved, but the vertex specification becomes difficult and detection errors increase
Solution Approach 1:
The invention extracts the periodicity information from the stripe pattern and uses it as a reference for document area detection. Instead of relying on vertex detection of fine stripes, the system extracts the cycle length of the pattern and uses this extracted information to identify document boundaries, thereby avoiding the vertex specification difficulty while maintaining detection accuracy.
Solution Approach 2:
The invention introduces an intermediary approach by using the periodicity of the stripe pattern as a mediator between the pattern itself and the document area detection. The periodicity serves as an intermediate reference that allows the system to detect document areas without directly detecting vertices, thus resolving the contradiction between detection accuracy and vertex specification difficulty.
2Measurement precision
If the number of vertices detected is increased by making stripes finer, then the document area detection accuracy is improved, but the reliability of vertex detection decreases
Solution Approach 1:
The invention extracts the periodicity information from the stripe pattern and uses it as a reference for document area detection. Instead of relying on vertex detection of fine stripes, the system extracts the cycle length of the pattern and uses this extracted information to identify document boundaries, thereby avoiding the vertex specification difficulty while maintaining detection accuracy.
Solution Approach 2:
The invention substitutes the mechanical vertex detection approach with a periodicity-based detection method. Instead of mechanically identifying vertices of fine stripes, the system uses the periodicity pattern to detect document areas, replacing the unreliable vertex detection mechanism with a more reliable periodicity recognition approach.
3Measurement precision
If a stripe pattern with specific cycle is used for background, then the pattern provides good contrast for document area detection, but the pattern itself may be mistaken for document content
Solution Approach 1:
The invention introduces an intermediary approach by using the periodicity of the stripe pattern as a mediator between the pattern itself and the document area detection. The periodicity serves as an intermediate reference that allows the system to detect document areas without directly detecting vertices, thus resolving the contradiction between detection accuracy and vertex specification difficulty.
Data Source
AI summary
The image reading device includes a reading section configured to read an image of a document and a background forming section that causes the reading section to read a background including a pattern with a specific cycle. A control section sets three or more continuous M-number of pixels in a direction with a cycle of the pattern in image data, determines the presence or absence of a short cycle fluctuation, which is a cyclical fluctuation of the brightness value specific to the pattern, for each of the single unit based on the brightness values of the M-number of pixels for each of the single unit, and discriminates between the document area and the background area based on the position at which the presence or absence of the short cycle fluctuation obtained from the image data switches.