Spaces:

---

prithivMLmods

/

POINTS-Reader-OCR

like 13

Running on Zero

key changes made:

• New "Bounding Boxes" Tab: Added a new tab to the interface, positioned before the "PDF Preview" tab.
• Text and Coordinate Extraction: Includes a button to run the model with a prompt that retrieves OCR text along with its coordinates.
• Visualization: Displays recognized text with bounding box coordinates and generates an image showing the original document with red boxes around detected text.
• Integrated Logic: Encapsulated the logic for prompting the model, parsing coordinates, and drawing boxes into a dedicated function.
• UI Refinements: Unified the main image uploader for all functions and updated the "Clear All" button to also reset the new components.