Character Recognition Using Reference Line Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing character recognition techniques that scan images with varying window sizes impose a heavy processing load, compromising recognition accuracy and efficiency.
Innovation Solution
A character recognition system that uses a two-step approach: first, it identifies character regions and sets reference lines for alignment, then performs recognition within limited scanning regions defined by these lines, reducing processing load while maintaining accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If character recognition is performed by scanning all over an extraction target region while changing the size of a window for character recognition, then recognition accuracy is improved, but processing load increases heavily
Solution Approach 1:
The patent segments the character recognition process into two distinct stages: a first recognition method that provides preliminary character identification, and a second recognition method that performs detailed recognition only in regions where characters are likely to exist (determined by reference lines from the first stage). This segmentation allows the system to maintain high recognition accuracy through the second stage while reducing overall processing load by limiting its scope.
Solution Approach 2:
The first recognition method performs preliminary character identification across the entire extraction target region before the second recognition method is applied. This preliminary action establishes reference lines that guide the subsequent detailed recognition process, allowing the system to prepare recognition regions in advance and avoid unnecessary scanning in areas without characters.
2Measurement precision
If the window size for character recognition is changed repeatedly to improve accuracy, then recognition precision is improved, but the time required for processing increases
Solution Approach 1:
The patent divides the recognition process into two temporal stages: the first recognition method executes quickly to establish reference framework (reference lines), and the second recognition method executes more slowly with varying window sizes but only within limited regions. This temporal segmentation reduces total processing time compared to applying the slow, high-precision method across the entire image.
Solution Approach 2:
The second recognition method applies partial action by performing detailed recognition with varying window sizes only in specific regions defined by reference lines from the first stage, rather than applying the computationally intensive process to the entire extraction target region. This partial application maintains precision where needed while reducing overall processing time.
Data Source
AI summary
A character recognition device includes an acquisition means configured to acquire an image containing a character region, a first recognition means configured to recognize a character from the character region by a first recognition method, a setting means configured to set reference lines along an alignment direction of the characters and passing through a specified position in each character, a second recognition means configured to recognize a character by a second recognition method, the second recognition method being a method that recognizes a character from an image within a recognition window by scanning in a recognition target region in an image while changing a size of the recognition window, and configured to set a position or a height in a vertical direction of the recognition window based on the reference lines, and an output means configured to output a word composed of characters recognized by the second recognition means.