Raster Boundary Curve Fitting With Corner-Based Path Simplification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image systems struggle to balance efficiency and accuracy in vectorizing raster images, resulting in complex vector images that are difficult to edit and inefficient for web-based applications due to excessive curves and points, leading to large file sizes and slow loading times.
Innovation Solution
A path simplification system that uses corner detection and a directed graph to determine line segments and candidate vector paths, generating a vector image by selecting paths that minimize costs based on cornerness scores, thereby reducing the number of vertices and paths.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image systems vectorize raster images with high accuracy, then the vector image preserves image quality, but the resulting vector image contains excessive curves and points leading to large file sizes and slow loading times
Solution Approach 1:
The patent extracts and removes redundant vertices from the vectorized paths while preserving the essential shape characteristics. The path simplification process identifies and eliminates unnecessary control points that contribute to file size without significantly impacting visual fidelity, directly resolving the contradiction between maintaining vectorization accuracy and reducing vector image complexity
Solution Approach 2:
Instead of starting with a simplified path and adding details, the patent inverts the approach by starting with a highly accurate but complex vectorized path and systematically removing elements. This inversion allows the system to begin with guaranteed accuracy and progressively simplify, ensuring that complexity reduction does not compromise essential image quality
2Manufacturing precision
If conventional image systems vectorize raster images with detailed paths, then the vector image accurately represents the original image, but the file sizes increase and loading times slow down
Solution Approach 1:
The patent extracts redundant geometric information from overly detailed paths, removing vertices that contribute minimally to boundary accuracy but significantly to file size. The simplification algorithm identifies and eliminates these excess elements while preserving the essential shape characteristics, directly addressing the contradiction between boundary accuracy and file size
Solution Approach 2:
The patent applies parameter changes by adjusting the tolerance and simplification thresholds to optimize the balance between boundary accuracy and file size. By dynamically adjusting these parameters based on image characteristics, the system achieves appropriate levels of detail without excessive complexity, resolving the contradiction between manufacturing precision and quantity of substance
3Reliability
If conventional image systems produce complex vector images, then the vector image maintains high fidelity to the original raster image, but the images are difficult to edit and inefficient for web-based applications
Solution Approach 1:
The patent extracts and removes redundant vertices and control points that make vector images difficult to edit while preserving the essential shape characteristics. By reducing the number of editable elements from excessive amounts to a manageable subset, the system improves ease of operation while maintaining sufficient image fidelity through the preservation of key geometric features
Solution Approach 2:
The patent inverts the conventional approach by starting with high-fidelity vector paths and systematically simplifying them, rather than starting with simplified paths and adding detail. This ensures that editability improvements do not compromise essential image fidelity, as the simplification process preserves critical shape characteristics while removing excessive complexity
Data Source
AI summary
Methods, systems, and non-transitory computer readable storage media are disclosed for vectorizing raster images using path simplification via corner detection and a directed graph. The disclosed system determines a plurality of line segments from one or more paths along boundaries of segmented regions of a raster image. The disclosed system determines cornerness scores corresponding to portions of the one or more paths at a set of vertices by comparing the portions at the set of vertices to different shapes including at least a corner shape. The disclosed system also fits a plurality of candidate vector paths to input paths corresponding to pairs of the set of vertices. Additionally, the disclosed system generates a vector image comprising a set of vector paths fitted to the boundaries of the segmented regions that minimize costs according to the cornerness scores.