18071465. LEARNING A FORM STRUCTURE simplified abstract (Microsoft Technology Licensing, LLC)
Contents
- 1 LEARNING A FORM STRUCTURE
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 LEARNING A FORM STRUCTURE - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Unanswered Questions
- 1.11 Original Abstract Submitted
LEARNING A FORM STRUCTURE
Organization Name
Microsoft Technology Licensing, LLC
Inventor(s)
LEARNING A FORM STRUCTURE - A simplified explanation of the abstract
This abstract first appeared for US patent application 18071465 titled 'LEARNING A FORM STRUCTURE
Simplified Explanation
The system described in the patent application learns the structure of a form from a single image without user annotation, groups text entries based on detected lines, measures distances and angles between text entry locations, and identifies optimal pairing solutions for typewritten and handwritten text entries.
- The system learns form structure from a single image without user input.
- Text entries are grouped based on detected lines in the form.
- Distances and angles between text entry locations are measured.
- Possible pairing solutions are represented in a bipartite graph.
- Optimal pairing solutions are identified by minimizing standard deviation of distances and/or circular standard deviation of angles.
Potential Applications
This technology could be applied in document processing, form recognition, and data entry automation.
Problems Solved
This technology solves the problem of efficiently pairing typewritten and handwritten text entries in forms without manual intervention.
Benefits
The system streamlines the process of analyzing and processing forms containing both typewritten and handwritten text entries, increasing efficiency and accuracy.
Potential Commercial Applications
One potential commercial application of this technology could be in document management software for businesses.
Possible Prior Art
One possible prior art for this technology could be existing form recognition systems that require manual annotation for text entry pairing.
Unanswered Questions
How does the system handle forms with complex layouts or overlapping text entries?
The patent application does not provide details on how the system deals with forms that have intricate designs or text entries that overlap.
What is the computational complexity of the system when processing large volumes of forms?
The patent application does not discuss the scalability of the system when handling a high volume of forms for processing.
Original Abstract Submitted
A system learns the structure of a form. The structure of the form can be learned from a single image (e.g., a photograph that includes the form) without user annotation. The form includes typewritten and handwritten text entries. The system groups text entries in the form based on lines detected in the form. The system then measures a distance and an angle between two text entry locations in the group of text entries. The group of text entries, the distances, and the angles can be captured in a bipartite graph. The bipartite graph represents possible pairing solutions where a typewritten text entry is paired with a handwritten text entry. The system identifies an optimal pairing solution, from the possible pairing solutions, using the distances and angles. The optimal pairing solution is identified by minimizing the standard deviation of the distances and/or by minimizing the circular standard deviation of the angles.
