Servers, systems, and methods for matching extracted data to a data source
AssigneeYouVersion, Inc.
InventorsRobert Lee Gruenewald, David Anthony Fedor, Timothy Todd Watson, Brian Daniel Herbert
Life of this patent
Estimating the term…
Loading patent…
AssigneeYouVersion, Inc.
InventorsRobert Lee Gruenewald, David Anthony Fedor, Timothy Todd Watson, Brian Daniel Herbert
Estimating the term…
Systems and methods for matching a captured image of a page to a characteristic database are disclosed. In some embodiments, a system receives a captured image of a book page from a computing device and executes a corrective normalization process to generate a normalized output. The system executes a cascaded processing sequence that applies processing operations in order of increasing computational intensity, escalating to more computationally demanding operations when a prior processing level produces an insufficient confidence level. In some embodiments, a first-level processing sequence generates a numerical identifier from pixel data extracted from a low-resolution representation of the normalized output. In some embodiments, a second-level processing sequence extracts page characteristics and abstracts them into a vector representation. In some embodiments, a third-level processing sequence generates one or more advanced representations including optical character recognition.
A system comprising one or more computers each comprising one or more processors and one or more non-transitory computer readable media, the one or more non-transitory computer readable media comprising program instructions stored thereon that when executed cause the one or more computers to:
Dependencies are inferred from the claim text.