Character_Recognition_Based_on_Global_Feature

https://www.pirahansiah.com/farshid/portfolio/publications/Papers/Character_Recognition_Based_on_Global_Feature

spotify

PDF Download My Conference Paper

Character Recognition Based on Global Feature Extraction

1. Introduction

• Objective: Propose a combination of two feature extraction techniques for character recognition.
• Key Techniques:
  • Gray Level Co-occurrence Matrix (GLCM)
  • Edge Direction Matrix (EDMS)
• Challenges:
  • Selecting the best feature extraction technique for various character recognition tasks.

2. Feature Extraction Techniques

• Gray Level Co-occurrence Matrix (GLCM):
  • Focuses on texture and pixel relationships in the image.
  • Extracts texture features for character recognition.
• Edge Direction Matrix (EDMS):
  • Emphasizes edges and directions in character images.
  • Extracts shape features.

3. Proposed Method

• Combination of GLCM and EDMS:
  • A hybrid approach that leverages both texture and edge features.
  • Aims to improve accuracy over using either GLCM or EDMS alone.

4. Classification Techniques

• Classifiers Used:
  • Neural Networks (NN)
  • Bayesian Networks (BN)
  • Decision Tree Classifiers
• Objective: Find the best classifier to complement the hybrid feature extraction method.

5. Experimental Results

• Datasets:
  • Binary character images of different sizes used for testing.
• Performance:
  • The hybrid method (GLCM + EDMS) outperforms individual feature extraction techniques.
  • Results show improved accuracy in character recognition tasks.

6. Conclusion

• Key Findings:
  • The combination of GLCM and EDMS provides better recognition performance than using either technique alone.
• Implications:
  • More reliable character recognition systems for real-world applications.