Feature-Based OCR

PART V - Fragmentation

<< Previous  Main  Next >> 

Naive 3-Step OCR

Step 1: Input Image

 

 Step 2: Skeletonize Image

 
Step 3: Extract Features (goal state)

  • RED: end-point
  • YELLOW: tri-intersection
  • BLUE: quad-intersection
  

<Under Construction>

<< Previous  Main  Next >> 

Next: Character Bleeding

Questions? Comments? Send them here. Or return to deXXed.