Character String Extraction Using Density Histograms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for extracting character strings from banknote images fail when the string and background have similar colors, or when additional characters or patterns are present, leading to inaccurate recognition.
Innovation Solution
A character string extraction method that sets a character frame on the image, calculates feature values for each character and surrounding region, and moves the frame to maximize a character string frame evaluation value, effectively distinguishing the target string from non-target characters and background patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If color information is used to distinguish character string from background, then extraction simplicity is improved, but extraction accuracy deteriorates when character and background have similar colors
Solution Approach 1:
The patent changes from using color information alone to using density information as the primary parameter for extraction. This parameter change allows accurate distinction between character and background even when their colors are similar, while maintaining computational simplicity through histogram-based density analysis.
Solution Approach 2:
The patent applies local quality analysis by examining density characteristics in specific regions (character portions and surrounding portions) rather than relying on global color properties. This enables adaptive extraction that works locally for each character region regardless of overall background color.
2Productivity
If histogram thresholding is used to extract character region, then extraction speed is improved, but extraction accuracy deteriorates when non-target characters or patterns are present
Solution Approach 1:
The patent segments the image into multiple character frames and processes each frame independently with its own histogram analysis. This segmentation allows accurate extraction of individual characters even when non-target characters or patterns are present in other regions, while maintaining overall processing efficiency through parallelizable operations.
Solution Approach 2:
The patent performs histogram thresholding on each character frame separately rather than on the entire image at once. This partial action approach focuses computational resources on relevant regions only, improving accuracy for each character while maintaining acceptable overall processing speed.
3Measurement precision
If character extraction window is moved to detect individual characters, then character position detection is improved, but processing complexity increases due to multiple scanning operations
Solution Approach 1:
The patent merges the character position detection and character extraction operations into a unified process. By setting character frames based on detected positions and extracting characters within these frames using the same density thresholding mechanism, the patent reduces processing complexity while maintaining detection precision.
Data Source
AI summary
In a character string extraction method, a character portion, a rim portion, a character frame, and a character string frame are set, a feature value of each image in the character portion and the rim portion is calculated for each character frame, a character string frame evaluation value is calculated based on the feature value for the character string frame, a position of the character string frame is moved on the paper sheet image, and the image in the character portion is extracted by using the character string frame at a position at which the character string frame evaluation value reaches a maximum.