With the widespread use of mobile phones and scanners to photograph and upload documents, the need for extracting the information trapped in unstructured document images such as retail receipts, insurance claim forms and financial invoices is becoming more acute. A major hurdle to this objective is that these images often contain information in the form of tables and extracting data from tabular sub-images presents a unique set of challenges. This includes accurate detection of the tabular region within an image, and subsequently detecting and extracting information from the rows and columns of the detected table. While some progress has been made in table detection, extracting the table contents is still a challenge since this involves more fine grained table structure(rows & columns) recognition. Prior approaches have attempted to solve the table detection and structure recognition problems independently using two separate models. In this paper, we propose TableNet: a novel end-to-end deep learning model for both table detection and structure recognition. The model exploits the interdependence between the twin tasks of table detection and table structure recognition to segment out the table and column regions. This is followed by semantic rule-based row extraction from the identified tabular sub-regions. The proposed model and extraction approach was evaluated on the publicly available ICDAR 2013 and Marmot Table datasets obtaining state of the art results. Additionally, we demonstrate that feeding additional semantic features further improves model performance and that the model exhibits transfer learning across datasets. Another contribution of this paper is to provide additional table structure annotations for the Marmot data, which currently only has annotations for table detection.

TableNet: Deep Learning Model for End-to-end Table Detection and Tabular Data Extraction from Scanned Document Images

Message management services can include processing an email to identify relevant content from among all content in the email and converting the identified relevant content into a simple, easy-to-read format. For example, message management services can apply multiple parsing strategies to an email. Each strategy can attempt to parse the email to identify relevant content and to output results that include any identified relevant content and an associated confidence score. The results having the highest confidence score can be selected. The identified relevant content included in the selected results can be converted into a flat string with stylings and links for displaying relevant content of the email in a simple, easy-to-read format.

Identifying relevant content in email

TableNet: Deep Learning model for end-to-end Table detection and Tabular data extraction from Scanned Document Images

Methods to select and extract tabular data among the optical character recognition returned strings to automatically process documents, including documents containing academic transcripts.

Methods for automatic structured extraction of data in OCR documents having tabular data

Described are methods and systems for determining authenticity. For example, the method may include providing an object of authentication, capturing characteristic data from the object of authentication, deriving authentication data from the characteristic data of the object of authentication, and comparing the authentication data with an electronic database comprising reference authentication data to provide an authenticity score for the object of authentication. The reference authentication data may correspond to one or more reference objects of authentication other than the object of authentication.

System and method for authentication

A borderless table detection engine and associated method for identifying borderless tables appearing in data extracted from a fixed format document. Due to the lack of visible borders, reliable automated detection of a borderless table is difficult. The borderless table detection engine uses whitespace, rather than content, to detect borderless table candidates. Applying heuristic analysis, the borderless table detection engine discards borderless table candidates with a layout that lacks sufficient characteristics of a table and is unlikely to be a valid borderless table.

Borderless table detection engine

Detection of East Asian layout features and reconstruction of East Asian layout features is provided. Vertically written text in the fixed format document is detected and rotated for layout analysis. After layout analysis, the rotated text is rotated back and restructured in a flow format document. When a plurality of characters is written horizontally in a vertical line of text, vertically overlapping text runs are detected, designated as horizontal-in-vertical text, and are restructured as horizontal-in-vertical text in a flow format document. Lines of text are analyzed for attributes of a ruby line and are designated as ruby text, associated with corresponding text in a ruby base line, and restructured as ruby text in a flow format document. Text in a fixed format document is analyzed for detection of a particular East Asian language so that a font for the language is designated in a flow format document.

Detection and Reconstruction of East Asian Layout Features in a Fixed Format Document

A fixed format document conversion engine and associated method for converting a fixed format document into a flow format document. The fixed format document conversion engine includes a sequence of layout analysis engines and semantic analysis engines to analyzes the base physical layout information obtained from the fixed format document to enrich, modify, and classify the physical layout information into progressively more advanced physical layout information and, ultimately, semantic layout information. The semantic layout information is mapped and serialized into a selected flow format document with a high level of flowability.

Fixed format document conversion engine

A vector graphics classification engine and associated method for classifying vector graphics in a fixed format document is described herein and illustrated in the accompanying figures. The vector graphics classification engine defines a pipeline for categorizing vector graphics parsed from the fixed format document as font, text, paragraph, table, and page effects, such as shading, borders, underlines, and strikethroughs. Vector graphics that are not otherwise classified are designated as basic graphics. By sequencing the detection operations in a selected order, misclassification is minimized or eliminated.

Vector Graphics Classification Engine

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.

Milos Raskovic

Papers

Borderless table detection engine

Detection and Reconstruction of East Asian Layout Features in a Fixed Format Document

Fixed format document conversion engine

Vector Graphics Classification Engine

Formula Detection Engine