Image Data Extraction

WebToNative On-Device Image → Structured Data — QA Console

Checking...
WebToNative Image Data Extraction API is not available. Open this page inside the WebToNative WebView.

1. Image Source

📷 Camera
📷️ Gallery
</> Base64

Camera: capture a new image using the native camera.

2. Schema Builder

Maximum nesting depth: 5 · Maximum fields: 50 · Max field name length: 100 · Max description length: 500 · Max schema JSON size: 32 KB

Use a Custom Schema JSON File

Upload a .json file containing the schema object. The file is read locally in this page; it is not uploaded to a server.

No custom schema file selected.

Schema Validation Tests

Each button fills the raw override box with a deliberately edge-case schema. The native API decides the actual validation result (expected: UNSUPPORTED_SCHEMA for invalid cases).

UI Options

Request Preview


    
  


  
  

    
    

      Status
      Checking...
    

    

      Callback Count (this request)
      0
    

  


  
  

    
Result

    
Status—

    
Engine—

    
Execution Time—

    
Error Code—

    
Error Message—

    
Extracted Data

    
No result yet.

    
Raw Response

    
No result yet.

    
    
  


  
  

    
Testing Controls

    

      
      
      
      
      
      
    

  


  
  

    
Concurrency Test Results

    
Fires two extraction requests back-to-back. Expect one to succeed/fail normally and the other to potentially return EXTRACTION_IN_PROGRESS.

    
Request 1not run

    
Request 2not run

  


  
  

    
Request History (in-memory only)

    

      

        
TimeModeStatusEngineExec ms

        
      

    

    
  


  
  

    

      Debug Information
      
Current Modecamera

      
Schema Field Count0

      
Schema JSON Size0 bytes

      
Request Start Time—

      
Request End Time—

      
Total Callback Count0

      
Last Error Code—

      
Last Engine Used—

      
Last Execution Time—

    

  


  
  

    
Activity Log (mirrors console, no sensitive data)

    

  


  
  

    
Implementation Note

    

      Current implementation pipeline: Image → OCR → recognized text → on-device GenAI(text) → structured result. This is not multimodal image GenAI.


      GenAI availability depends on device/OS support. When GenAI is unavailable or fails, OCR fallback may be used.
    

  


  
  

    
Schema Limits (V1)

    
Primitive typesstring, number, integer, boolean

    
Complex typesobject, array

    
Maximum fields50

    
Maximum nesting depth5

    
Maximum description length500 chars

    
Maximum field-name length100 chars

    
Maximum schema JSON size32 KB

    
Missing valuesnull

    
Confidence fieldnot available in V1

    
Documented Error Codes

    
CAMERA_PERMISSION_DENIED · USER_CANCELLED · INVALID_REQUEST · INVALID_IMAGE · UNSUPPORTED_SCHEMA · EXTRACTION_FAILED · EXTRACTION_IN_PROGRESS · INTERNAL_ERROR

  


  
  

    
QA Checklist (in-memory only)

    

  


  
  

    
Security / Privacy

    
This page never uploads images, never sends extracted data anywhere, never writes Base64 or extracted data to localStorage, and never logs Base64 content. Everything lives in runtime memory only and is lost on reload.