Character Recognition Line Separation via Distance Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional character recognition methods fail to accurately separate lines of text when the image is rotationally offset or when the space between text lines is narrow, leading to inaccurate detection of character strings.
Innovation Solution
The method involves setting a search area in a gray-scale image and accumulating image data across lines extending from multiple points along the axes, identifying local maximum or minimum values to determine the lines passing between character strings, and using these values to correctly detect and separate text lines, even with rotational offsets or narrow spacing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If conventional projection processing is used to separate lines of text, then processing is simple when text is aligned with axes, but separation accuracy deteriorates when image is rotationally offset or lines are narrowly spaced
Solution Approach 1:
The patent transitions from conventional 2D projection processing to 3D spatial analysis by introducing depth information through distance transformation. The algorithm creates a 3D representation where pixel values represent distances to text boundaries, enabling accurate line separation even when text lines are narrowly spaced or rotationally offset by analyzing spatial relationships in three dimensions rather than relying on simple 2D projections
Solution Approach 2:
The patent changes the parameter being analyzed from simple pixel intensity (conventional projection) to distance from text boundaries (distance transformation). This parameter transformation allows the system to identify text line boundaries more accurately by measuring spatial distance fields rather than relying on intensity projections, thereby improving separation accuracy for rotated or closely-spaced lines
2Measurement precision
If image capture location is prevented from overlapping background portions, then text line separation is accurate for aligned text, but adaptability deteriorates for rotationally offset text
Solution Approach 1:
The patent creates a universal text line separation method that functions effectively for both axis-aligned and rotationally offset text, as well as for narrowly spaced lines. The distance transformation approach provides a unified solution that adapts to various text orientations and spacing conditions without requiring separate processing paths, thereby achieving multi-functionality and broad adaptability across different text configurations
3Measurement precision
If narrow space between text lines is handled, then separation accuracy improves, but processing complexity increases
Solution Approach 1:
The patent replaces complex mechanical or geometric analysis methods with a field-based approach using distance transformation. Instead of attempting to directly analyze and differentiate narrowly spaced lines through complex geometric operations, the system substitutes a computational field method that naturally represents spatial relationships, simplifying the processing while improving accuracy for closely spaced text lines
Data Source
AI summary
An image processing method of processing a gray-scale image of character strings arranged in a plurality of lines of text, for character recognition, includes setting at least a part of a range, in which the character strings in the gray-scale image area are distributed, as a search area, and setting, for each of a plurality of points arranged along one of two axes, representing coordinates in the image, corresponding to an arrangement of the line of text, a plurality of lines extending across the search area, with the point as a starting point, and accumulating, for each of the lines, pixel data in the line within the search area.