2D Pattern Decoding on Non-Uniform Backgrounds via Relative Parameter Differences
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional two-dimensional data matrix patterns are not robust enough, especially when applied to non-uniform backgrounds, and their detection is challenging due to the absence of guiding patterns like L-shapes or QR code squares, leading to difficulties in decoding, especially when print quality is poor or the substrate is uneven.
Innovation Solution
A method for reading and decoding two-dimensional patterns by converting digital images into a one-dimensional color space and using a sliding window to detect reference patterns, where each symbol's state is determined by the difference in parameter values of its elements, allowing for robust detection and decoding even on non-uniform backgrounds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional two-dimensional data matrix patterns are used on non-uniform backgrounds, then the identification function is provided, but the detection reliability deteriorates due to substrate unevenness and poor print quality
Solution Approach 1:
The patent changes the parameter used for detection from absolute intensity values to relative parameter value differences within symbols. By comparing the difference between parameter values of different elements within each symbol rather than using fixed thresholds, the system becomes invariant to substrate unevenness and print quality variations, thereby improving detection reliability under harmful conditions
Solution Approach 2:
Instead of trying to make the pattern invariant to background variations by using robust encoding, the patent inverts the approach by making the detection method adaptive to local variations. The reader adjusts its detection parameters based on local background characteristics, effectively inverting the problem from 'making the code robust' to 'making the reader adaptive'
2Measurement precision
If conventional thresholding methods are used for detection, then the decoding process is simplified, but the measurement precision deteriorates on non-uniform backgrounds
Solution Approach 1:
The patent replaces conventional fixed thresholding with a method that uses relative parameter value differences. For each symbol, the system calculates the difference between parameter values of its constituent elements and compares this difference to a threshold. This parameter change from absolute to relative measurement maintains detection simplicity while dramatically improving precision on non-uniform backgrounds
Solution Approach 2:
The detection process is segmented into symbol-level operations where each symbol is processed independently by comparing internal element differences. This segmentation allows the complex problem of non-uniform background detection to be broken down into simple, repeatable unit operations, maintaining low overall complexity while achieving high precision
3Reliability
If robust encoding patterns with multiple representations are used, then the reliability improves, but the device complexity increases for detection and decoding
Solution Approach 1:
Rather than increasing pattern complexity to achieve robustness, the patent inverts the approach by simplifying the detection logic while maintaining pattern flexibility. The reader uses a unified detection method based on parameter differences that works across all symbol types, reducing detection complexity while preserving the robustness benefits of multiple representations
Data Source
Figure 1A
Figure 1B~1C
Figure 2
AI summary
The invention relates to a method and device for detecting and decoding a two- dimensional encoded pattern, e.g. of the Robust Encoding Pattern type (REP), even in case the pattern is applied on a non-uniform colored background of a surface of a substrate. The operations for decoding such an encoded pattern involve detecting the states of its symbols through a test pattern, checking that local differences between parameter values of the elements of the symbols are consistent with an allowed representation of the test pattern to assign via a one-to-one mapping the states of the symbols of the test pattern to the corresponding symbols of the encoded pattern, and decoding the encoded pattern based on the assigned representation of its symbols.