Banknote Overlap Detection Using Local Texture Windows
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing overlapping banknote detection methods inaccurately determine banknotes as overlapped due to attachments, slight curvature, or aging, which affects the recognition capabilities of intelligent banknote detectors and sorters.
Innovation Solution
A method and device that acquire front and back images of banknotes, measure the coupling degree using local feature attribute windows, and determine overlapping conditions based on tone texture periods, structure texture periods, textons, and texture definitions, to accurately assess whether banknotes are overlapped.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If thickness signal-based detection is used to identify overlapping banknotes, then detection speed is improved, but measurement precision deteriorates due to false positives from attachments, curvature, or aging
Solution Approach 1:
The banknote image is divided into multiple local regions (local feature attribute windows) for independent analysis. Each region's texture characteristics are separately measured and compared, allowing the system to identify whether different parts of the banknote are at different heights, which indicates overlapping. This segmentation approach enables precise localization of overlapping areas without being affected by attachments or curvature in other regions.
Solution Approach 2:
The patent analyzes texture characteristics including tone texture (color variations) and structure texture (pattern variations) in different regions of the banknote. By detecting changes in these visual properties across local regions, the system can identify height differences caused by overlapping while distinguishing them from color variations due to attachments or aging.
2Measurement precision
If high density and accuracy of thickness sensor signals are required, then detection precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex mechanical thickness sensing systems with a simplified optical imaging system. By capturing images of the banknote and analyzing texture characteristics in local regions, the system achieves accurate overlapping detection without requiring high-density thickness sensor arrays. This substitution of mechanical sensing with optical analysis reduces device complexity while maintaining or improving detection accuracy.
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
A detection method and a detection apparatus for overlapped notes. The detection method comprises: collecting a front image and a back image of banknotes to be recognized, and transmitting the images (101); respectively acquiring front banknote information and back banknote information about the banknotes to be recognized according to the front image and the back image, and initializing a corresponding local characteristic attribute window, wherein the local characteristic attribute window is a template of a pre-set recognition region containing corresponding characteristic attributes of an image that needs to be recognized (102); performing coupling degree measurement on the transmitted image of the front image and the back image according to the local characteristic attribute window to obtain a measurement result (103); judging whether the measurement result meets a pre-set banknote overlapping condition (104); if so, determining that the banknotes to be recognized are overlapped banknotes (105); and if not, determining that the banknotes to be recognized are not overlapped banknotes (106). The detection method for overlapped notes solves the problem of misjudgement that is easily caused where there are attachments on banknotes, the banknotes slightly arch and old banknotes become thick.