Document Edge Detection via Density Fluctuation Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies struggle to accurately detect document boundaries in scenarios where the document has non-straight ends or is placed on complex backgrounds, such as wood grain finishes or electronic paper, which complicates the differentiation between document edges and background patterns.
Innovation Solution
An image processing apparatus and method that extracts edge pixels, groups them, and uses approximated line detection to identify document edges, even in cases with non-straight ends, by connecting edge pixels in horizontal and vertical directions to define document boundaries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If simple color backing is used for document reading, then document boundary detection is improved, but adaptability to different background types deteriorates
Solution Approach 1:
The patent changes the detection parameter from color-based differentiation to density fluctuation-based differentiation. By detecting fluctuations in density of image data rather than relying on color contrast between document and backing, the system achieves accurate boundary detection across diverse background types including wood grain finishes and electronic paper displays.
Solution Approach 2:
The patent replaces the optical/color-based detection mechanism with a density fluctuation analysis mechanism. Instead of using color contrast to identify boundaries, the system analyzes variations in pixel density values and their fluctuations, substituting the detection approach to work independently of background color or pattern.
2Measurement precision
If ruled line extraction is used for boundary detection, then detection accuracy is improved, but applicability to documents without ruled lines deteriorates
Solution Approach 1:
The patent extracts and removes ruled lines from the image data as a preprocessing step. By detecting and eliminating the influence of ruled lines before boundary detection, the system prevents ruled lines from interfering with boundary detection accuracy, while the subsequent density fluctuation method ensures documents without ruled lines are also handled correctly.
Solution Approach 2:
The patent performs preliminary processing to detect and remove ruled line influences before the actual boundary detection takes place. This preliminary action eliminates potential interference from ruled lines, ensuring that the subsequent density fluctuation-based detection works accurately for both ruled and non-ruled documents.
3Ease of manufacture
If straight line approximation is used for document ends, then detection simplicity is improved, but accuracy for non-straight ends deteriorates
Solution Approach 1:
The patent replaces straight line approximation with curve approximation methods. By detecting density fluctuations along curved paths and using curve fitting techniques, the system accurately identifies document boundaries even when ends are non-straight, such as with open books or curled paper, while maintaining computational feasibility.
Data Source
AI summary
There are provided an image processing apparatus, line detection method and a computer-readable, non-transitory medium that can precisely detect boundaries of a document from a read image. The image processing apparatus includes an edge pixel extractor for extracting edge pixels from an input image which includes a document, a grouping unit for forming mutually adjoining edge pixels in the extracted edge pixels into a group, and an approximated line detector for detecting a line which connects edge pixels which are positioned at the two ends in a horizontal direction and vertical direction among edge pixels which are included in the group as a line which approximates an end of the document.