An_evaluation_of_classification_techniques_using_enhanced_Geometrical_Topological_Feature_Analysis
https://www.pirahansiah.com/farshid/portfolio/publications/Papers/An_evaluation_of_classification_techniques_using_enhanced_Geometrical_Topological_Feature_Analysis
PDF Download My Conference Paper
An Evaluation of Classification Techniques Using Enhanced Geometrical Topological Feature Analysis
1. Introduction
- Objective: Evaluation of classification techniques for the Malaysian License Plate Recognition (LPR) system.
- Applications of LPR:
- Law enforcement
- Border protection
- Vehicle theft detection
- Automatic toll collection
- Traffic control
2. Image Classification Techniques
- Artificial Immune Recognition System (AIRS):
- Mimics biological immune systems for pattern recognition.
- Neural Networks (NN):
- Machine learning technique that models the human brain to classify images.
- Bayesian Networks (BN):
- Probabilistic graphical models to classify data based on probability distributions.
- Support Vector Machine (SVM):
- Uses geometric representations to classify data.
3. Enhanced Geometrical Topological Feature Analysis
- Proposed Approach:
- Focuses on improving image classification accuracy for Malaysian license plates.
- Uses topological features from license plate characters and numbers.
- Input Features:
- Character shapes
- Geometrical features
4. Classification Error Analysis
- Character Error Analysis:
- Errors analyzed based on classification methods.
- Provides insights into which methods perform better in specific conditions.
5. Results
- Best Performing Technique: Support Vector Machine (SVM)
- Outperforms AIRS, Neural Networks, and Bayesian Networks in accuracy for Malaysian LPR.
- Performance Factors:
- Environmental conditions (weather, lighting)
- Character variability (font, size)
6. Conclusion
- Summary: SVM provides the most accurate classification for the Malaysian LPR system.
- Implications: Improved reliability for law enforcement, toll collection, and traffic control systems.