Convert recognition results to markdown format

Process elements of the same type in batches

Dolphin-V1.5 layout string parsing function Parse layout string to extract bbox and category information Supports multiple formats: 1

Save JSON and markdown outputs

Process regular text content, preserving paragraph structure

Convert PDF pages to images Args: pdf_path: Path to PDF file target_size: Target size for the longest dimension Returns:

Process a single image (either from file or converted from PDF page) Args: image: PIL Image object model: DOLPHIN model insta

Process layout for a single image Args: pil_image: PIL Image object model: DOLPHIN model instance save_dir: Directory

Create necessary output directories

Handle figure content

Handle formula-specific content

Convert section headings to appropriate markdown format

Convert list items to markdown list format

Convert table content to markdown format

Process mathematical formulas in text by iteratively finding and replacing formulas. - Identify inline and block formulas - R

Remove newline in heading

Assign colors to elements in order Args: num_elements: Number of elements to assign colors to Returns: List of c

Vectorized IoU matrix calculation [N, N] Args: boxes: List of bounding boxes in [x1, y1, x2, y2] format Returns:

Check if bounding boxes have significant overlaps, indicating a distorted/photographed document If more than 60% of boxes have overlaps (IoU

from [202,217,921,325][para][author] extract para and author

Extract and clean table tags from HTML string

Get a visually pleasing color palette for layout visualization Returns: List of BGR color tuples (semi-transparent, good for overlay)

Parse documents with two stages - Handles both images and PDFs

Process a single element image (text, table, formula) Args: image_path: Path to the element image model: HFModel model instan

Parse all document elements with parallel decoding

Process layout detection for image or PDF Args: input_path: Path to input image or PDF model: DOLPHIN model instance

Save combined results for multi-page PDF with both JSON and Markdown Args: all_page_results: List of results for all pages pdf_pa

Save cropped figure to local file system Args: pil_crop: PIL Image object of the cropped figure save_dir: Base directory to save

Visualize layout detection results on the image Args: image_path: Path to the input image layout_results: List of (bbox, labe

Initialize the Hugging Face model Args: model_id_or_path: Path to local model or Hugging Face model ID

Initialize the Hugging Face model Args: model_id_or_path: Path to local model or Hugging Face model ID

Initialize the Hugging Face model Args: model_id_or_path: Path to local model or Hugging Face model ID

Crop margins from image