OCR/openspec/changes/archive/2025-12-11-improve-ocr-track-algorithm/tasks.md

1. Algorithm Changes (gap_filling_service.py)

1.1 IoA Implementation

• 1.1.1 Add _calculate_ioa() method alongside existing _calculate_iou()
• 1.1.2 Modify _is_region_covered() to use IoA instead of IoU
• 1.1.3 Update deduplication logic to use IoA

1.2 Dynamic Threshold Strategy

• 1.2.1 Add element-type-specific thresholds as class constants
• 1.2.2 Modify _is_region_covered() to accept element type parameter
• 1.2.3 Apply different thresholds based on element type (TEXT: 0.6, TABLE: 0.1, FIGURE: 0.8)

1.3 Boundary Shrinking

• 1.3.1 Add optional shrink_pixels parameter to coverage detection
• 1.3.2 Implement bbox shrinking logic (inward 1-2 px)

2. OCR Data Source Changes

2.1 Extract overall_ocr_res from PP-StructureV3

• 2.1.1 Modify pp_structure_enhanced.py to extract overall_ocr_res from result
• 2.1.2 Convert dt_polys + rec_texts + rec_scores to TextRegion format
• 2.1.3 Store extracted OCR in result dict for gap filling

2.2 Update Processing Orchestrator

• 2.2.1 Add option to use overall_ocr_res as OCR source
• 2.2.2 Skip separate Raw OCR inference when using PP-StructureV3's OCR
• 2.2.3 Maintain backward compatibility with explicit Raw OCR mode

3. Configuration Updates

3.1 Add Settings (config.py)

• 3.1.1 Add gap_filling_ioa_threshold_text: float = 0.6
• 3.1.2 Add gap_filling_ioa_threshold_table: float = 0.1
• 3.1.3 Add gap_filling_ioa_threshold_figure: float = 0.8
• 3.1.4 Add gap_filling_use_overall_ocr: bool = True
• 3.1.5 Add gap_filling_shrink_pixels: int = 1

4. Testing

4.1 Unit Tests

• 4.1.1 Test IoA calculation with known values
• 4.1.2 Test dynamic threshold selection by element type
• 4.1.3 Test boundary shrinking edge cases

4.2 Integration Tests

• 4.2.1 Test with scan.pdf (current problematic file)
• 4.2.2 Compare results: old IoU vs new IoA approach
• 4.2.3 Verify no duplicate text rendering in output PDF
• 4.2.4 Verify table content is not duplicated outside table bounds

5. Documentation

• 5.1 Update spec documentation with new algorithm
• 5.2 Add inline code comments explaining IoA vs IoU