Dual-Candidate Edge Detection for Flare-Resistant Document Scanning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing technologies face challenges in accurately detecting edges between document regions and background regions due to phenomena like flare and vertical line noise, which can lead to incorrect edge detection and reduced precision.
Innovation Solution
An image processing apparatus and method that employs a dual-candidate detection system, using both pattern matching and differential calculation to identify edge points, where the second candidate is determined as an edge point only if it is positioned more inside the document region than the first candidate, thereby reducing incorrect detections caused by factors other than flare.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single edge detection method is used, then the detection process is simple, but the detection precision deteriorates due to flare and noise
Solution Approach 1:
The edge detection process is segmented into two independent candidate detection methods: pattern matching for detecting first candidates and differential calculation for detecting second candidates. Each method operates independently to identify potential edge points, allowing the system to leverage the strengths of both approaches while mitigating their individual weaknesses against flare and noise
Solution Approach 2:
The results from both pattern matching and differential calculation are merged through a selection mechanism that chooses the appropriate candidate based on positional relationships. This combination allows the system to achieve higher detection precision by integrating multiple detection strategies rather than relying on a single method
2Reliability
If pattern matching is used for edge detection, then it can detect edges in binarized images, but it produces incorrect detections due to flare
Solution Approach 1:
Differential calculation acts as an intermediary verification mechanism for pattern matching results. By introducing this additional detection layer, the system can identify and eliminate false positives caused by flare, as the differential method responds differently to flare artifacts compared to actual edge transitions
Solution Approach 2:
The system uses feedback from the differential calculation results to validate or reject pattern matching candidates. When the differential detection does not confirm a pattern matching candidate, the system can identify this as a false positive and correct the detection, thereby improving reliability in the presence of flare
3Measurement precision
If differential calculation is used for edge detection, then it can detect brightness changes, but it produces incorrect detections due to noise
Solution Approach 1:
Pattern matching serves as an intermediary validation mechanism for differential calculation results. The pattern matching approach, which operates on binarized images with predefined edge patterns, can distinguish between genuine edges and noise artifacts, providing a filtering effect that reduces false detections from vertical line noise
Solution Approach 2:
The system uses feedback from pattern matching to verify differential calculation candidates. When differential detection identifies a potential edge but pattern matching does not confirm it with the expected pattern, the system can reject this as a false positive caused by noise, thereby improving detection precision
Data Source
AI summary
Provided are an image processing apparatus, an image processing method, a computer-readable medium storing a computer program and an image processing system improving edge detection precision at the time of flare occurrence. The image processing apparatus includes an input unit for taking image data including a document region, a first candidate detector for detecting a first candidate of an edge point constituting a boundary line of the document region by scanning binarized image of the image data with a predetermined pattern along a line, a second candidate detector for detecting a second candidate of an edge point based on differential value of pixels adjoining each other, and an edge point determination unit for determining the second candidate as an edge point when the second candidate is positioned more inside of the document region than the first candidate, and otherwise determines the first candidate as an edge point.