Image Extraction

Overview

EdgeParse can detect and extract images embedded in PDF documents, reporting their bounding boxes and page locations.

Image Detection

Images are identified in the JSON output:

{
  "type": "image",
  "id": 5,
  "page number": 1,
  "bounding box": [72, 300, 540, 500]
}

CLI Usage

# Extract with image metadata
edgeparse document.pdf -f json

# Images are reported in the kids array with type "image"

Python Usage

import edgeparse
import json

json_str = edgeparse.convert("document.pdf", format="json")
data = json.loads(json_str)

for element in data["kids"]:
    if element["type"] == "image":
        print(f"Image on page {element['page number']}")
        print(f"  Bounding box: {element['bounding box']}")

Markdown Output

In Markdown format, images are represented as placeholders with their page and position information.