Optical Character Recognition Localization via Inter-Character Spacing Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical character recognition (OCR) processes face challenges with images suffering from perspective distortion, curved surfaces, low contrast, noise, character variation, skew, variable spacing, and background variation, leading to segmentation errors and recognition failures.
Innovation Solution
An image processing system that partitions images into regions, determines characteristic values to identify lines of characters, calculates probable inter-character spacing, and uses this information to iteratively recognize characters, employing a trained classifier for precise character recognition even under poor acquisition conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional OCR segmentation is applied to images with perspective distortion, curved surfaces, noise, and variable spacing, then the system attempts to recognize characters, but segmentation errors occur leading to recognition failures
Solution Approach 1:
The patent divides the image processing task into distinct stages: (1) detecting candidate character regions using trained classifiers, (2) determining inter-character spacing through transition analysis, (3) identifying starting character positions, and (4) recognizing characters in identified regions. This segmented approach allows each stage to be optimized independently, improving overall reliability under poor image conditions.
Solution Approach 2:
The system performs preliminary analysis by detecting transitions between foreground and background regions before character recognition. It identifies candidate character regions and determines probable inter-character spacing in advance, which prepares the system for more accurate recognition even when images suffer from distortion, noise, or variable spacing.
2Reliability
If the classification sub-process is performed throughout the entire image, then all characters may be recognized, but computational cost becomes excessively high
Solution Approach 1:
The patent extracts only the necessary portions of the image for classification by first identifying candidate character regions through transition detection. The classification sub-process is then applied only to these extracted candidate regions rather than the entire image, significantly reducing computational energy while maintaining recognition completeness.
Solution Approach 2:
The system applies different processing strategies to different regions: candidate character regions undergo full classification analysis, while non-character regions are excluded from intensive processing. This local quality approach ensures computational resources are focused where needed, balancing completeness with energy efficiency.
3Adaptability or versatility
If images are captured from curved surfaces or at various angles, then more characters can be accessed, but perspective distortion and character skew complicate segmentation
Solution Approach 1:
The patent analyzes transitions in candidate character regions to determine probable inter-character spacing and identifies starting character positions based on these transition patterns. This parameter-based approach adapts to various image orientations and perspectives without requiring complex geometric corrections, maintaining segmentation effectiveness across diverse acquisition conditions.
Data Source
AI summary
Systems and methods for processing images to recognize characters. Such may include locating background in an image which separate lines of characters, and then locating background which separate characters. The inter-character spaces within a line may be used to determine a probable inter-character spacing for the characters. Within each detected line, the character having an inter-character spacing most similar to the probable inter-character spacing may be set as a starting character for classification. Using the probable inter-character spacing and the location of the starting character, the location of a character adjacent to the starting character in a first direction may be determined and the character classified. Such process may repeat in the first direction until a first end of the line is reached. The process may then move to a character adjacent the starting character in a second direction and repeat until a second end of the line is reached.