[@Tayseer] Thanks for asking! In recent versions of OCR Reader, we introduced new fine-tuning options to improve recognition accuracy and flexibility.
Previously, the software treated the entire input image as a single block of text. While this worked well for simple documents, it was limited when dealing with more complex layouts.
The new version has the following modes:
* Read One Text Block – Processes the image as a single, uniform block of text (similar to the old behavior).
* Find Scattered Text – Designed for images with randomly placed or separated lines of text.
* Auto-Detect Layout and Orientation – Automatically detects text orientation, columns, and overall layout for more accurate OCR on complex documents.
For a full list of changes and improvements, you can also check the version history here:
https://vovsoft.com/version-history/ocr-reader/