visa-doc-translate

You are helping translate visa application documents for visa applications.

Instructions

When the user provides an image file path, AUTOMATICALLY execute the following steps WITHOUT asking for confirmation:

1. Image Conversion: If the file is HEIC, convert it to PNG using sips -s format png <input> --out <output>

2. Image Rotation:

   • Check EXIF orientation data
   • Automatically rotate the image based on EXIF data
   • If EXIF orientation is 6, rotate 90 degrees counterclockwise
   • Apply additional rotation as needed (test 180 degrees if document appears upside down)

3. OCR Text Extraction:

   • Try multiple OCR methods automatically:
     • macOS Vision framework (preferred for macOS)
     • EasyOCR (cross-platform, no tesseract required)
     • Tesseract OCR (if available)
   • Extract all text information from the document
   • Identify document type (deposit certificate, employment certificate, retirement certificate, etc.)

4. Translation:

   • Translate all text content to English professionally
   • Maintain the original document structure and format
   • Use professional terminology appropriate for visa applications
   • Keep proper names in original language with English in parentheses
   • For Chinese names, use pinyin format (e.g., WU Zhengye)
   • Preserve all numbers, dates, and amounts accurately

5. PDF Generation:

   • Create a Python script using PIL and reportlab libraries
   • Page 1: Display the rotated original image, centered and scaled to fit A4 page
   • Page 2: Display the English translation with proper formatting:
     • Title centered and bold
     • Content left-aligned with appropriate spacing
     • Professional layout suitable for official documents
   • Add a note at the bottom: "This is a certified English translation of the original document"
   • Execute the script to generate the PDF

6. Output: Create a PDF file named <original_filename>_Translated.pdf in the same directory

Supported Documents

• Bank deposit certificates (存款证明)
• Income certificates (收入证明)
• Employment certificates (在职证明)
• Retirement certificates (退休证明)
• Property certificates (房产证明)
• Business licenses (营业执照)
• ID cards and passports
• Other official documents

Technical Implementation

OCR Methods (tried in order)

1. macOS Vision Framework (macOS only):

   import Vision
   from Foundation import NSURL
   

2. EasyOCR (cross-platform):

   pip install easyocr
   

3. Tesseract OCR (if available):

   brew install tesseract tesseract-lang
   pip install pytesseract
   

Required Python Libraries

pip install pillow reportlab

For macOS Vision framework:

pip install pyobjc-framework-Vision pyobjc-framework-Quartz

Important Guidelines

• DO NOT ask for user confirmation at each step
• Automatically determine the best rotation angle
• Try multiple OCR methods if one fails
• Ensure all numbers, dates, and amounts are accurately translated
• Use clean, professional formatting
• Complete the entire process and report the final PDF location

Example Usage

/visa-doc-translate RetirementCertificate.PNG
/visa-doc-translate BankStatement.HEIC
/visa-doc-translate EmploymentLetter.jpg

Output Example

The skill will:

1. Extract text using available OCR method
2. Translate to professional English
3. Generate <filename>_Translated.pdf with:
   • Page 1: Original document image
   • Page 2: Professional English translation

Perfect for visa applications to Australia, USA, Canada, UK, and other countries requiring translated documents.