Text Image Enhancement for OCR via Profile-Based Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional OCR engines face performance degradation due to poor image quality, especially in images with varying contrast, illumination, and skew, leading to increased processing time and reduced accuracy.
Innovation Solution
The method involves detecting and filtering the background by calculating background and text color profiles, using stroke recovery techniques to enhance text quality, and applying thresholding and interpolation processes to improve text integrity and contrast, ensuring dark text on a white background.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If conventional OCR engine processes low quality images, then processing time increases, but OCR accuracy degrades
Solution Approach 1:
The patent applies preliminary image enhancement actions before OCR processing by detecting text regions, enhancing contrast within those regions, and correcting skew. This preprocessing prepares low-quality images in advance, ensuring they meet quality standards before OCR processing, thus maintaining both accuracy and reducing processing time.
Solution Approach 2:
The patent enhances image quality locally rather than globally by identifying text regions and applying contrast enhancement and skew correction specifically to those regions. This localized approach improves OCR accuracy where needed while minimizing unnecessary processing elsewhere, thereby reducing overall processing time.
2Loss of time
If conventional OCR engine processes images with poor contrast and illumination, then processing time increases, but text-background separation reliability deteriorates
Solution Approach 1:
The patent applies contrast enhancement locally to text regions identified through region detection. By calculating contrast metrics for detected text regions and applying enhancement only where needed, the patent improves text-background separation accuracy without requiring global image processing, thus reducing processing time.
3Manufacturing precision
If stroke width is below threshold, then text quality is poor, but OCR accuracy decreases
Solution Approach 1:
The patent detects stroke width as a quality parameter and applies enhancement operations when stroke width falls below a threshold. By monitoring this parameter and triggering appropriate enhancement, the patent maintains text quality within acceptable ranges, ensuring OCR accuracy without unnecessary processing.
Data Source
Figure 1~4
Figure 5~9
Figure 10(A)~11
AI summary
A method for enhancing a textual image for undergoing optical character recognition begins by receiving an image that includes native lines of text. A background line profile is determined which represents an average background intensity along the native lines in the image. Likewise, a foreground line profile is determined which represents an average foreground background intensity along the native lines in the image. The pixels in the image are assigned to either a background or foreground portion of the image based at least in part on the background line profile and the foreground line profile. The intensity of the pixels designated to the background portion of the image is adjusted to a maximum brightness so as to represent a portion of the image that does not include text.