layoutReferences Array Elements in Vantage OCR JSON

Elements of the layoutReferences array in Vantage OCR JSON output map paragraphs to physical text blocks and locate each paragraph in the logical structure.

Elements of the layoutReferences array are objects that establish the mapping between a paragraph and text blocks of the document’s physical structure, as well as describing the location of the paragraph in the document’s logic structure. The main text of the document is split up into logic sections, where text can in turn be split up into columns. Elements that are not part of the main text (e.g. headings, headers/footers, row numbers, etc.) are not included in any logic section.

Property	Data type	Description
blockId *	string	The identifier of the text block or table cell containing the paragraph.
blockType *	string enum	The type of block the identifier of which is specified in the blockId property. Possible values: text, cell
sectionIndex *	integer	The index of the logic section on the paragraph page. A value of -1 indicates that the paragraph does not belong to any logic section. Logic sections do not include headings, footers, etc. Minimum value: -1
columnIndex *	integer	The index of a column in the paragraph’s logic section. A value of -1 indicates that the paragraph does not belong to any column. For example, columns do not include text boxes, notes, etc. Minimum value: -1
lineNumbering *	boolean	A value of true indicates that the paragraph is located inside a row numbering region.
parIndex *	integer	The index of the paragraph in a text block or table cell. Minimum value: 0
firstLine *	integer	The index of the first line of the paragraph in a text block or table cell. Minimum value: 0
lastLine *	integer	The index of the last row of the paragraph in a text block or table cell. Minimum value: 0

* Indicates a required property.