Character Recognition via Dot Arrangement Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional optical character recognition systems experience a significant decline in recognition rate for characters displayed on computer screens due to coarser dot resolution compared to printed characters, especially when small characters are used, leading to low recognition accuracy.
Innovation Solution
A character recognition system that includes a storage device with a recognition dictionary containing information on various character arrangements, a character recognition unit for comparing dot arrangements, and a font check unit to identify display font and size, allowing for accurate recognition by matching image data from the video memory with pre-stored dictionary entries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional optical character recognition techniques are used for recognition of characters displayed on a display, then the system can recognize printed characters with high accuracy, but the recognition rate sharply declines for displayed characters due to coarser dot resolution
Solution Approach 1:
The patent changes the recognition parameters from continuous optical patterns to discrete dot arrangement patterns. By representing characters as grids of dots and comparing dot positions rather than continuous image data, the system adapts to the lower resolution of display devices while maintaining recognition accuracy.
Solution Approach 2:
The patent segments character images into discrete dot elements arranged in grids. Instead of treating characters as continuous images, the system divides them into individual dot positions that can be independently compared, making recognition robust to resolution variations.
2Productivity
If small characters are used to increase the amount of data displayed, then more information can be shown on the display, but character recognition rate becomes extremely low
Solution Approach 1:
The patent adds a structural dimension to character recognition by organizing dots into grid patterns with defined spatial relationships. This dimensional structure allows recognition of small characters by comparing their dot arrangement patterns rather than relying on pixel density, enabling accurate recognition even at reduced sizes.
3Measurement precision
If a comprehensive recognition dictionary is created to cover various fonts and sizes, then recognition accuracy improves, but system complexity and memory requirements increase
Solution Approach 1:
The patent changes the dictionary structure from storing complete character images to storing dot arrangement patterns. This parameter change reduces memory requirements while maintaining recognition accuracy across different fonts and sizes, as the dot grid representations are more compact than full image data.
Data Source
AI summary
A character recognition system for recognizing character(s) displayed on display(s) of computer(s). In one embodiment, memory device(s) store recognition dictionaries, and there may be a separate recognition dictionary for each display font and character size combination. Such a recognition dictionary may contain information pertaining to arrangement of constituent dots making up each character of a corresponding display font and character size. A character recognition unit may identify target character(s) displayed on display(s) by searching such recognition dictionary or dictionaries and extracting therefrom character(s) for which information pertaining to arrangement of constituent dots making up extracted character(s) is a perfect match for target character(s) displayed on display(s).