layout-analyzer

The layout-analyzer skill performs document layout analysis for images and PDFs using surya, an advanced document understanding system. It helps identify the structural regions of complex documents, such as text blocks, tables, figures, headings, captions, formulas, headers, footers, and footnotes. This solves the problem of understanding how a document page is organized before extracting, reading, or processing its content.

Its main capabilities include layout detection, structure analysis, text region detection, and reading order detection. The skill can analyze document images, detect layout element types, return bounding boxes with confidence scores, and determine the sequence in which detected regions should be read. It also includes examples for using surya predictors such as DetectionPredictor, LayoutPredictor, ReadingOrderPredictor, and OCRPredictor in Python.

This skill is useful for users working with scanned documents, PDFs, research papers, reports, forms, or visually complex pages. It is especially relevant for document processing workflows that need to separate paragraphs, tables, figures, headings, and other layout components before further OCR, extraction, or analysis. Target users include technical teams, document automation specialists, data extraction engineers, and anyone needing structured document understanding.