Formula Detection Engine for Fixed Format Document Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing document converters fail to accurately convert fixed format documents into flow format documents, especially when dealing with complex elements like mathematical formulas, resulting in limited flowability and requiring substantial manual reconstruction.
Innovation Solution
A formula detection engine is employed to identify and process mathematical formulas in fixed format documents by detecting seeds, grouping elements, eliminating overlaps, and reconstructing formulas as flowable elements, ensuring proper formatting and layout in the output document.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If base techniques are used to preserve visual fidelity of fixed format document layout, then layout accuracy is improved, but formula detection accuracy deteriorates
Solution Approach 1:
The patent segments the document processing into distinct phases: initial formula detection using specialized algorithms, followed by layout preservation techniques. Formula seeds are identified and grouped into formula areas before general layout processing occurs, allowing formulas to be detected and marked with unique identifiers that preserve both detection accuracy and visual fidelity.
Solution Approach 2:
The patent performs preliminary formula detection and seeding before applying layout preservation techniques. By identifying formula seeds and creating formula areas in advance, the system ensures that formulas are properly detected and structured before the layout preservation process modifies the document structure, thus maintaining both detection accuracy and visual fidelity.
2Adaptability or versatility
If formula areas are expanded to group all nearby elements, then formula completeness is improved, but overlap with normal text increases
Solution Approach 1:
The patent applies local quality by making formula area expansion adaptive rather than uniform. The expansion distance is adjusted based on local characteristics: expanding sufficiently to include all formula elements while contracting to avoid normal text. Formula seeds are grouped based on their specific spatial relationships, allowing each formula area to have customized boundaries that prevent text overlap.
Solution Approach 2:
The patent implements dynamic formula area boundaries that adjust based on the distribution of formula elements and proximity to normal text. The expansion process is iterative and adaptive, allowing formula areas to grow to include necessary elements while dynamically responding to text boundaries to prevent harmful overlaps.
3Ease of operation
If substantial manual reconstruction is performed to achieve useful flow format document, then document usability is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary formula detection, seeding, and area grouping automatically during the conversion process. By establishing formula areas and identifying mathematical elements before the user receives the document, the system eliminates the need for substantial manual reconstruction, thus improving usability while minimizing processing time.
Solution Approach 2:
The patent enables the document conversion system to self-serve by automatically detecting formulas, grouping elements, and structuring flow format output without requiring user intervention. The system autonomously identifies formula seeds, expands formula areas, and produces usable flow format documents, eliminating manual reconstruction efforts.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A formula detection engine and associated method. The formula detection engine locates formulas within a fixed format document portion by identifying formula seeds. The formula detection engine creates and expands a boundary around the formula seed to define a formula area. To eliminate overlap with surrounding normal text, the formula area is divided into multiple formula areas based on vertical position and horizontal spacing between the formula elements. After being vertically ordered, horizontally overlapping formula areas are merged to reconstruct the formula as a flowable element.