Content-Based Image Classification: Efficient Machine Learning Using Robust Feature Extraction Techniques

Cover
Half Title
Title Page
Copyright Page
Dedication
Contents
Preface
Author
1. Introduction to Content-Based Image Classification
	1.1. Prelude
	1.2. Metrics
		1.2.1. Precision
		1.2.2. True Positive (TP) Rate/Recall
		1.2.3. Misclassification Rate (MR)
		1.2.4. F1-Score
		1.2.5. Accuracy
		1.2.6. False Positive (FP) Rate
		1.2.7. True Negative (TN) Rate
		1.2.8. False Negative (FN) Rate
	1.3. Classifiers
		1.3.1. KNN Classifier
		1.3.2. Random Forest Classifier
		1.3.3. ANN Classifier
		1.3.4. SVM Classifier
	1.4. Datasets Used
		1.4.1. Wang Dataset
		1.4.2. Caltech Dataset
		1.4.3. Corel Dataset
		1.4.4. Oliva Torralba (OT-Scene) Dataset
	1.5. Organization of the Book
	Chapter Summary
	References
2. A Review of Handcrafted Feature Extraction Techniques for Content-Based Image Classification
	2.1. Prelude
	2.2. Extraction of Features with Color Contents
	2.3. Extraction of Features with Image Binarization
	2.4. Extraction of Features with Image Transforms
	2.5. Extraction of Features with Morphological Processing
	2.6. Extraction of Features with Texture Content
	2.7. Fusion of Features Extracted with Multiple Techniques
	2.8. Techniques of Classification
	2.9. Logic-Based Algorithms
		2.9.1. Decision Trees
		2.9.2. Learning a Set of Rules
		2.9.3. Perceptron-Based Techniques
			2.9.3.1 Single-Layer Perceptrons
			2.9.3.2 Multilayer Perceptrons
		2.9.4. Statistical Learning Algorithm
		2.9.5. Support Vector Machine
	Chapter Summary
	References
3. Content-Based Feature Extraction: Color Averaging
	3.1. Prelude
	3.2. Block Truncation Coding
	3.3. Feature Extraction Using Block Truncation Coding with Color Clumps
	3.4. Code Example (MATLAB®)
	3.5. Coding Exercise
	3.6. Feature Extraction Using Sorted Block Truncation Coding for Content-Based Image Classification
	3.7. Code Example (MATLAB)
	3.8. Coding Exercise
	3.9. Comparison of Proposed Techniques
	3.10. Comparison with Existing Techniques
	3.11. Statistical Significance
	Chapter Summary
	References
4. Content-Based Feature Extraction: Image Binarization
	4.1. Prelude
	4.2. Feature Extraction Using Mean Threshold Selection
		4.2.1. Feature Extraction with Multilevel Mean Threshold Selection
	4.3. Code Example (MATLAB®)
	4.4. Coding Exercise
	4.5. Feature Extraction from Significant Bit Planes Using Mean Threshold Selection
	4.6. Code Example (MATLAB)
	4.7. Coding Exercise
	4.8. Feature Extraction from Even and Odd Image Varieties Using Mean Threshold Selection
	4.9. Code Example (MATLAB)
	4.10. Coding Exercise
	4.11. Feature Extraction with Static and Dynamic Ternary Image Maps Using Mean Threshold Selection
	4.12. Code Example (MATLAB)
	4.13. Feature Extraction Using Local Threshold Selection
	4.14. Code Example (MATLAB)
	4.15. Coding Exercise
	4.16. Comparing the Discussed Techniques for Performance Evaluation
	4.17. Comparison with Existing Techniques
	4.18. Statistical Significance
	Chapter Summary
	References
5. Content-Based Feature Extraction: Image Transforms
	5.1. Prelude
	5.2. Generating Partial Energy Coefficient from Transformed Images
	5.3. Code Example (MATLAB®)
	5.4. Coding Exercise
	5.5. Computational Complexity for the Image Transforms
	5.6. Feature Extraction with Partial Energy Coefficient
		5.6.1. Discrete Cosine Transform
		5.6.2. Walsh Transform
		5.6.3. Kekre Transform
		5.6.4. Discrete Sine Transform
		5.6.5. Discrete Hartley Transform
	5.7. Evaluation of the Proposed Techniques
	5.8. Comparison with Existing Techniques
	5.9. Statistical Significance
	Chapter Summary
	References
6. Content-Based Feature Extraction: Morphological Operators
	6.1. Prelude
	6.2. Top-Hat Transform
	6.3. Code Example (MATLAB®)
	6.4. Coding Exercise
	6.5. Bottom-Hat Transform
	6.6. Code Example (MATLAB)
	6.7. Coding Exercise
	6.8. Comparison of Proposed Techniques
	6.9. Comparison with Existing Methods
	6.10. Statistical Significance
	Chapter Summary
	References
7. Content-Based Feature Extraction: Texture Components
	7.1. Prelude
	7.2. Feature Extraction by Vector Quantization Codebook Representation Using Linde-Buzo-Grey (LBG) Algorithm
	7.3. Code Example (MATLAB®)
	7.4. Coding Exercise
	7.5. Feature Extraction by Gray Level Co-occurrence Matrix (GLCM)
	7.6. Code Example (MATLAB)
	7.7. Coding Exercise
	7.8. Evaluation of Proposed Techniques
	7.9. Comparison with Existing Methods
	7.10. Statistical Significance
	Chapter Summary
	References
8. Fusion-Based Classification: A Comparison of Early Fusion and Late Fusion Architecture for Content-Based Features
	8.1. Prelude
	8.2. Image Preprocessing
	8.3. Feature Extraction with Image Binarization
	8.4. Feature Extraction Applying Discrete Cosine Transform (DCT)
	8.5. Classification Framework
		8.5.1. Method 1
		8.5.2. Method 2
	8.6. Classification Results
	Chapter Summary
	References
9. Future Directions: A Journey from Handcrafted Techniques to Representation Learning
	9.1. Prelude
	9.2. Representation Learning-Based Feature Extraction
	9.3. Code Example (MATLAB®)
	9.4. Image Color Averaging Techniques
	9.5. Binarization Techniques
	9.6. Image Transforms
	9.7. Morphological Operations
	9.8. Texture Analysis
	9.9. Multitechnique Feature Extraction for Decision Fusion-Based Classification
	9.10. Comparison of Cross Domain Feature Extraction Techniques
	9.11. Future Work
	References
10. WEKA: Beginners' Tutorial
	10.1. Prelude
	10.2. Getting Started with WEKA
	References
Index