Character Cutting via Projection Processing for Uneven Backgrounds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing character recognition methods face accuracy issues due to uneven background gray levels and non-uniform character spacing, leading to failed character cutting and subsequent erroneous recognition, especially when processing speed is a concern.
Innovation Solution
A method involving projection processing to distinguish between character and background areas by setting a straight line based on local maximum and minimum values in projected gray-level patterns, allowing for accurate character cutting even with complex patterns and shading, and enabling user-adjustment for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If projection processing is performed using accumulated image data in a single line, then processing speed is maintained at high level, but cutting out accuracy degrades when background gray level unevenness increases
Solution Approach 1:
The patent divides the projection processing into multiple independent line processing units, where each line is processed separately to generate projection data. This segmentation allows the system to handle background unevenness in each local region independently, improving cutting out accuracy without requiring complex global processing adjustments.
Solution Approach 2:
The patent applies different processing strategies to different regions (lines) of the character string image. By processing each line independently and adapting to local characteristics such as background gray level unevenness, the system achieves high cutting out accuracy while maintaining overall processing efficiency.
2Reliability
If character cutting out is performed without considering background unevenness, then processing time is reduced, but recognition accuracy degrades due to failed cutting out
Solution Approach 1:
The patent performs preliminary projection processing on multiple lines to generate projection data before the actual character cutting out. This preliminary action identifies character regions and prepares cutting parameters in advance, ensuring accurate cutting out while minimizing additional processing time during the main recognition task.
Solution Approach 2:
The patent uses projection data and local maximum/minimum value information as feedback to adjust the cutting out process. By continuously monitoring the projected patterns and adjusting cutting parameters based on this feedback, the system maintains high recognition accuracy without requiring excessive processing time.
3Productivity
If simple projection processing is used, then processing speed is maintained high, but cutting out fails when non-uniform pitch or small spaces exist
Solution Approach 1:
The patent introduces a new dimension of analysis by processing multiple lines simultaneously and using projected pattern data from different spatial dimensions. This multi-dimensional approach enables the system to detect character boundaries even in cases of non-uniform pitch or small spaces, improving cutting out accuracy without significantly reducing processing speed.
Solution Approach 2:
The patent uses projection data as an intermediary representation that bridges the original image and the cutting out process. This intermediary projection pattern simplifies the detection of character regions and spacing characteristics, enabling accurate cutting out while maintaining high processing speed through efficient data representation.
Data Source
AI summary
A method for cutting out, from a gray-scale image generated by capturing an image of a character string, each character in the character string for recognition, includes a first step of repeating projection processing for projecting a highest or lowest gray level in a line along a direction crossing the character string in the gray-scale image, onto an axis along the character string, with the lowest gray level selected when a character in the gray-scale image is darker than a background, the highest gray level selected when the character in the gray-scale image is brighter than the background, and a projection target position moved along the character string.