Irregular Character Orientation for OCR Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current OCR technologies face difficulties in recognizing and processing irregularly arranged characters within images, such as those oriented in arbitrary directions or positions, which limits their ability to convert handwritten or printed text into machine-encoded formats for electronic editing, searching, and machine processing.
Innovation Solution
A text structuring system that detects irregularly arranged characters, modifies their direction to a proper orientation, and then recognizes them, allowing the image to be rebuilt with the characters in a machine-encoded format, thereby facilitating their recognition and utilization in machine processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current OCR technologies are used to recognize irregularly arranged characters, then the recognition process can be performed, but the recognition accuracy is poor and training time is excessive
Solution Approach 1:
The patent applies preliminary action by detecting irregularly arranged characters and pre-processing them (extracting, rotating, and resizing) before OCR recognition. This pre-processing step prepares the characters in advance, converting them into a standardized format that OCR algorithms can recognize accurately without requiring extensive training, thus resolving the contradiction between recognition accuracy and training time.
2Measurement precision
If irregularly arranged characters are processed without pre-processing, then the processing speed is maintained, but the recognition accuracy deteriorates
Solution Approach 1:
The patent applies segmentation by dividing the image processing into distinct stages: detecting irregularly arranged characters, extracting them from the image, rotating them to proper orientation, resizing to standard dimensions, and finally performing OCR recognition. This segmented approach allows each step to be optimized independently, maintaining overall processing speed while achieving high recognition accuracy through specialized pre-processing operations.
3Reliability
If traditional OCR methods are applied directly to irregularly arranged characters, then the system complexity is low, but the recognition reliability is poor
Solution Approach 1:
The patent introduces an intermediary pre-processing module between the image input and OCR recognition. This intermediary component detects irregularly arranged characters, extracts them, rotates them to proper orientation, and resizes them to standard dimensions before passing them to the OCR engine. This intermediary layer significantly improves recognition reliability by ensuring characters are in the correct format, while adding only moderate system complexity through well-defined processing steps.
Data Source
AI summary
Aspects of the present disclosure relate to processing irregularly arranged characters. An image is received. An irregularly arranged character within the image is detected. A direction of the irregularly arranged character is modified to a proper direction to obtain a properly oriented character. The properly oriented character is recognized to obtain a first identified character. The image is then rebuilt by replacing the irregularly arranged character with the first identified character, the first identified character in a machine-encoded format.