Add local OCR recognition and search for image content (supporting multiple platforms)

[fightnvrgp]

Describe feature

I would like to suggest adding a local OCR (Optical Character Recognition) recognition feature for images in Trilium, as well as the ability to search image content based on the extracted text. This would greatly enhance the usability of image notes and allow users to quickly find relevant information contained in images.

For the OCR recognition engine, it would be ideal to not only support Tesseract but also to integrate platform-specific OCR engines where available. For example:

• On macOS/iOS: Vision framework
• On Android: Android's native OCR API
• On Windows: Windows built-in OCR engine
• On Linux: Tesseract (and others if available)

Implementing this feature would enable seamless text extraction and searching capabilities for images across all major platforms, making Trilium more versatile for users who rely on image notes.

Additional Information

This feature should allow users to perform OCR locally without relying on cloud APIs, ensuring privacy and performance. Additionally, the search functionality should be able to index and search the extracted text from images just like regular note content for unified results. Supporting multiple OCR engines per platform will also make the feature more robust and accessible to all users.

[eliandoran]

There's already a PR in #5834 but it needs more work to be complete.
Either way, the feature request is a duplicate.