G06V 30/41

Definition

(Für diese Definition ist die deutsche Übersetzung noch nicht abgeschlossen)

Document analysis, recognition and understanding by processing:

• Structured documents such as business forms, bank checks whose layout is provided with printed lines or input boxes, bounding boxes, checkboxes, straight lines or tables;
• Classification of document image content by identification of text regions, photographs, tables;
• Extracting and analysing the geometrical structure, e.g. the layout tree representation in which different entities such as paragraphs, images, etc. are represented as nodes of a tree or a graph;
• Extracting and analysing the logical structure, e.g. identification of the chapter headings, sections, columns, titles, paragraphs, captions, page numbers, or identification of the constituting elements such as authors, keywords, postal codes, money amounts;
• Document matching, e.g. by establishing the degree of (dis)similarity between two document images, one reference/template document image and one input document image.

Examples

Identification of the text and image regions; the image tiles are marked with an “I” and text tiles are marked with a “T”.

Extraction of document structure by analysis of its content, resulting in the identification of elements such as paragraphs, drawings, handwritten annotations.

Querverweise

Einschränkende Querverweise

Informative Querverweise