Document Image Correction Mesh Using Boundary Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing document image correction methods assume a cylindrical document surface, leading to persistent distortion due to the non-planar shape of actual documents.
Innovation Solution
A device and method that extract boundaries and lines in document images, filter them to match predetermined shapes, extend and adjust these lines to construct a correcting mesh, and use this mesh to accurately correct the document image, accounting for both document boundaries and content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a cylindrical model is used for document correction, then the correction process is simplified, but distortion remains in the corrected image
Solution Approach 1:
The patent divides the document image into multiple regions based on extracted lines and boundaries, creating a segmented correction approach. The correcting mesh is constructed by dividing the image into grid cells, with each cell being corrected independently based on local geometric transformations, thereby achieving high precision without requiring a complex global model
Solution Approach 2:
The patent transitions from 2D image coordinates to 3D spatial coordinates by introducing depth information through line extraction and boundary detection. The correcting mesh operates in a transformed coordinate system that accounts for document curvature, allowing accurate correction of non-planar surfaces while maintaining computational feasibility
2Productivity
If only document boundaries are used for correction, then the correction process is faster, but content distortion cannot be eliminated
Solution Approach 1:
The patent applies different correction strategies to different regions of the document. Content regions are identified and corrected using local geometric transformations based on extracted text lines and graphical elements, while boundary regions use edge-based correction. This localized approach ensures both processing efficiency and content accuracy
Solution Approach 2:
The patent implements an iterative correction process where the correcting mesh is refined based on feedback from detected lines and boundaries. The system repeatedly extracts features, evaluates correction accuracy, and adjusts the mesh parameters until convergence, ensuring that both boundaries and content regions are accurately corrected
Data Source
AI summary
Embodiments of the present invention provide a device and method for correcting a document image and a scanner. Wherein the device includes: an extracting unit configured to extract boundaries in a first direction of the document image and extract lines in the first direction according to a content of the document image, a filtering unit configured to filter the extracted lines, a constructing unit configured to extend and adjust the filtered lines and construct a correcting mesh, and a correcting unit configured to perform correcting according to the correcting mesh. An accurate correcting model can be constructed by taking document boundaries and document contents into account in constructing the correcting model and constructing a correcting mesh after filtering, extending and adjusting the extracted lines, thereby effectively eliminating distortion in the document image.