ocr speed improved

measure_ocr_init.py

@@ -0,0 +1,41 @@
+#!/usr/bin/env python3
+"""
+Measure OCR initialization time.
+"""
+import time
+import sys
+import os
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+
+def measure_optimized_ocr_init():
+    from LightRAG_main.lightrag.optimized_ocr_processor import OptimizedOCRProcessor
+    start = time.time()
+    processor = OptimizedOCRProcessor(use_gpu=True, batch_size=4, max_workers=2)
+    elapsed = time.time() - start
+    print(f"OptimizedOCRProcessor initialization time: {elapsed:.2f} seconds")
+    print(f"OCR available: {processor.ocr_available}")
+    if processor.ocr_available:
+        # warm up model (already done in initialization)
+        pass
+    return processor
+
+def measure_simple_ocr_init():
+    from simple_ocr_processor import SimpleOCRProcessor
+    start = time.time()
+    processor = SimpleOCRProcessor()
+    elapsed = time.time() - start
+    print(f"SimpleOCRProcessor initialization time: {elapsed:.2f} seconds")
+    print(f"OCR available: {processor.available}")
+    return processor
+
+if __name__ == "__main__":
+    print("Measuring OCR initialization times...")
+    try:
+        measure_optimized_ocr_init()
+    except Exception as e:
+        print(f"Failed to measure OptimizedOCRProcessor: {e}")
+    try:
+        measure_simple_ocr_init()
+    except Exception as e:
+        print(f"Failed to measure SimpleOCRProcessor: {e}")