DetectAreasMode enumeration

Determines the type of neural network used for areas detection.

Members

Member name	Description
LEAN	Prioritizes speed and reduces resource consumption by omitting support for complex layouts. Suitable only for simple images with a few lines of text without illustrations or formatting.
DOCUMENT_LAYOUT	Detects complete document layout, including text blocks, headings, columns, tables, images, charts, and formulas. Use this mode when the recognized structure must be preserved in DOCX or Markdown output.
MULTICOLUMN	Detects complete document layout, including multi-column text, tables, images, charts, and formulas. Kept for backward compatibility;
UNIVERSAL	Detects all blocks of text in the image, including sparse and irregular text on photos. A versatile option for most images, except for tables and multi-column layouts.
TABLE	Detects tabular structures in the image and extracts text from individual cells. Recommended for scanned spreadsheets, reports, and other table-based documents.
CURVED_TEXT	Automatically straightens curved lines of text in the image, improving recognition accuracy and allowing more text to be recovered and extracted. Requires significant processing power and RAM.
FORMULA	Detects all blocks with mathematical formulas.