Image Perspective Transformer

This project provides a PySide6 GUI for marking four points on an image, applying a perspective transform, and exporting a clean crop for documents (e.g., LaTeX). It supports interactive point editing, live preview, and automatic downscaling to a maximum side length for consistent output sizes.

Features

• Smart image view: Fit-to-screen display with accurate coordinate mapping.
• Fast point selection: Click up to four points to define the transform region.
• Editable points: Drag points to refine corners; lines and area shading update live.
• Area highlight: The selected quadrilateral is shaded for clear visual feedback.
• Detail magnifier: Hold Ctrl near a point for a 2x zoom lens on the corner.
• Preview + export: Preview the transformed image and save to a chosen path.
• Max-side downscale: Cap output size via --max-side (default 1024) for LaTeX-friendly images.
• Shortcuts: Ctrl+Q to quit, Ctrl+S to save, Ctrl+P to preview, Ctrl+Enter to apply.

Ensure the following are installed on your system:

• Python 3.8+

Install Steps

1. Install from PyPI:

pip install perspective-transformer

2. Run the application:

perspective-transformer <image_path> [output_path] [--max-side 1024]

How to Use

1. Run the Application:

   • Replace image.jpg in the code with the path to your image or place your image in the same directory as the script.

   • Start the application:

     perspective-transformer <image_path> [output_path] [--max-side 1024]
     

2. Mark Points:

   Click on the image to select four points in any order. These points define the region for perspective transformation. Drag any point to adjust.

3. Confirm Points:

   After selecting four points, click "Select Points" (or Ctrl+Enter) to apply the transformation.

4. Preview Transformed Image:

   Click "Preview Transformed Image" to see the transformed result in the same window. If the result is unsatisfactory, click "Clear Points" to reset and reselect points.

5. Save Transformed Image:

   When satisfied, click "Write Transformed Image" (or Ctrl+S) to save the output as transformed_image.jpg or the specified output path.

License

This project is open-source and can be modified and used for personal or educational purposes. Attribution to the original creator is appreciated. (MIT License)

Enjoy transforming your images! 🚀