Skip to main content
The document extraction function can recognize handwriting information in documents, including handwritten text content, position coordinates, and other detailed information. Handwriting recognition is significant for processing documents containing handwritten content (such as handwritten signatures, handwritten notes, etc.).

​
Handwriting Information Structure

Handwriting information is located in result.files[].data.handwritings[], with each handwriting item containing the following attributes:
  • page: Page number where handwriting content is located (starting from 0)
  • text: Recognized handwritten text content
  • position[]: Position coordinate information of handwriting content in the document

​
Handwriting Information Data Structure

{
  "page": 0,           // Page number
  "text": "March 1st",    // Handwritten text content
  "position": [
    {
      "page": 0,       // Page number
      "vertices": [    // Coordinates of four vertices [x1,y1,x2,y2,x3,y3,x4,y4]
        100, 200,      // Top-left corner
        200, 200,      // Top-right corner
        200, 250,      // Bottom-right corner
        100, 250       // Bottom-left corner
      ]
    }
  ]
}

​
Position Coordinate Structure

{
  "page": 0,           // Page number (starting from 0)
  "vertices": [        // Coordinates of four vertices [x1,y1,x2,y2,x3,y3,x4,y4]
    100, 200,          // Top-left corner
    200, 200,          // Top-right corner
    200, 250,          // Bottom-right corner
    100, 250           // Bottom-left corner
  ]
}
For detailed coordinate descriptions, please refer to the Coordinate System documentation.

​
Example Code

import requests
import json

def extract_handwritings(workspace_id, batch_number, app_id, secret_code):
    """Extract handwriting information from documents"""
    
    host = "https://docflow.textin.com"
    url = "/api/app-api/sip/platform/v2/file/fetch"
    
    resp = requests.get(
        f"{host}{url}",
        params={
            "workspace_id": workspace_id, 
            "batch_number": batch_number
        },
        headers={
            "x-ti-app-id": app_id, 
            "x-ti-secret-code": secret_code
        },
        timeout=60,
    )
    
    if resp.status_code != 200:
        print(f"Request failed: {resp.status_code}")
        return None
    
    data = resp.json()
    
    for file in data.get("result", {}).get("files", []):
        print(f"File name: {file.get('name')}")
        
        # Extract handwriting information
        handwritings = file.get("data", {}).get("handwritings", [])
        
        if handwritings:
            print(f"\n=== Handwriting Information ===")
            print(f"Number of handwriting items: {len(handwritings)}")
            
            for i, handwriting in enumerate(handwritings):
                page = handwriting.get("page", 0)
                text = handwriting.get("text", "")
                positions = handwriting.get("position", [])
                
                print(f"\nHandwriting item {i+1}:")
                print(f"  Page: Page {page+1}")
                print(f"  Content: {text}")
                
                # Display position information
                for j, pos in enumerate(positions):
                    pos_page = pos.get("page", 0)
                    vertices = pos.get("vertices", [])
                    print(f"  Position {j+1} (Page {pos_page+1}): {vertices}")
        else:
            print("No handwriting information found")
    
    return data

# Usage example
if __name__ == "__main__":
    workspace_id = "<your-workspace-id>"
    batch_number = "<your-batch-number>"
    app_id = "<your-app-id>"
    secret_code = "<your-secret-code>"
    
    result = extract_handwritings(workspace_id, batch_number, app_id, secret_code)

​
Return Data Example

{
  "code": 200,
  "result": {
    "files": [
      {
        "id": "202412190001",
        "name": "contract.pdf",
        "recognition_status": 1,
        "data": {
          "handwritings": [
            {
              "page": 0,
              "text": "John Smith",
              "position": [
                {
                  "page": 0,
                  "vertices": [100, 500, 150, 500, 150, 520, 100, 520]
                }
              ]
            },
            {
              "page": 0,
              "text": "December 19, 2024",
              "position": [
                {
                  "page": 0,
                  "vertices": [200, 500, 300, 500, 300, 520, 200, 520]
                }
              ]
            },
            {
              "page": 0,
              "text": "Agree to this clause",
              "position": [
                {
                  "page": 0,
                  "vertices": [100, 600, 200, 600, 200, 620, 100, 620]
                }
              ]
            },
            {
              "page": 1,
              "text": "Jane Doe",
              "position": [
                {
                  "page": 1,
                  "vertices": [100, 300, 150, 300, 150, 320, 100, 320]
                }
              ]
            }
          ]
        }
      }
    ]
  }
}