About Attrasoft ImageFinder Software Requirements Install the Software Information and Support Statement of Copyright Restriction 1. INTRODUCTION 1.1 What will the ImageFinder do? 1.2 Biometrics 1.3 Software Requirements 1.4 Customized Software 1.5 Trails 2. IMAGEFINDER OVERVIEW 2.1 ImageFinder Internal Structures 2.2 Image Preprocessing 2.3 Normalization 2.4 Feature Recognition 2.5 ABM Matching Engine 2.6 Multi-Layered ABM 2.7 Chapter Overview 3. FACEFINDER 3.1 Introduction 3.1.1 Expectation 3.1.2 FaceFinder Input 3.1.3 FaceFinder 3.1.4 Restrictions 3.2 Identification 3.2.1 Operation 3.2.2 Template Matching Output 3.2.3 Checking Identification Results 3.2.4 Analysis 3.2.5 Examples 3.2.6 Identify with the Template File 3.3 Verification 3.3.1 Operation 3.3.2 Example 4. DOCUMENTFINDER 4.1 Why Image Matching? 4.1.1 Expectations 4.2 Run DocumentFinder 4.2.1 Input 4.2.2 Training 4.2.3 Matching 4.3 Restrictions 4.4 Examples 4.5 Simple Matching 4.5.1 Operation 4.5.2 File Input Examples 4.5.3 Dir Input Examples 4.6 Three-Step Matching 4.6.1 Matching Procedure 4.6.2 Example “BioAPI” 4.7 Parameters 4.8 Checking the Results 4.8.1 B1_matchlist.txt 4.8.2 Example “Abm54” 4.8.3 Example “All” 4.9 Query Set Vs Target Set 4.9.1 Query Set against Target Set 4.9.2 Examples 5. IMAGEEXAMINER 5.1 Introduction 5.2 Operations 5.2.1 Data 5.2.2 Commands 5.2.3 Parameters 5.3 Getting Started 5.3.1 The Problem 5.3.2 Good Match 5.3.3 Bad Match 5.4 Multiple Matches 5.4.1 Multiple Good Matches 5.4.2 Multiple Bad Matches 5.4.3 Limitation of the Software 6. IMAGE PROCESSING 7. BIOFILTERS 7.1 Test-Driving the Label Matching Problem, Unsupervised N:N Matching 7.2 BioFilter Overview 7.3 Unsupervised 1:N Matching 7.4 Training 7.5 Supervised N:N Matching 7.6 Supervised 1:N Matching 7.7 Checking the Results 7.8 File Input 7.9 Analysis 7.10 Summary 8. NEURAL FILTERS 8.1 Test-Driving the Label Matching Problem 8.2 Neural Filter Overview 8.3 Training 8.4 N:N Matching 8.5 1:N Matching 8.6 Checking the Results 8.7 Analysis 8.8 Summary 9. NEURALNET FILTERS 9.1 Test-Driving the Logo Matching Problem 9.2 NeuralNet Filter Overview 9.3 Training 9.4 N:N Matching 9.5 1:N Matching 9.6 Summary 10. PARAMETERS 10.1 Overview 10.2 Filter Parameters 10.3 Image Processing 10.3.1 Edge Filters 10.3.2 Threshold Filters 10.3.3 Clean Up Filters 10.3.4 Starting the Selection 10.4 Normalization Filter 10.4.1 Reduction Filters 10.4.2 Parameters 10.5 BioFilter 10.6 Neural Filters 10.7 NeuralNet Filter 10.7.1 Symmetry 10.7.2 Translation Type 10.7.3 Scaling Type 10.7.4 Rotation Type 10.7.5 Area of Interest (AOI) 10.7.6 Blurring 10.7.7 Sensitivity 10.7.8 Internal/External Weight Cut 10.7.9 L/S Segments 10.7.10 Image Type 10.7.11 Output 10.7.12 Segment 10.7.13 Bypass 10.7.14 Summary 11. BATCH JOB 11.1 Creating Batch Code 11.2 Sample Batch 11.3 Overview 11.4 Batch Execution Code 11.5 BioFilter Examples 11.6 NeuralFilter Examples 11.7 NeuralNet Filter Examples 12. NEURALNET FILTER AND SHORT-SEARCH 12.1 Overview 12.2 Parameters 12.3 Short-Search, Long-Search, and File- Search 12.4 ImageFinder Operations for Short- Search 12.5 ImageFinder Operations for Short- Search (Advanced) 12.6 TradeMark Retrieval 12.6.1 United Way - Rotation Symmetry 12.6.2 Tabasco - Rotation Symmetry 12.6.3 Mr. Potato - Scaling Symmetry 12.6.4 Monopoly - Scaling Symmetry 12.6.5 Chemical Compound 12.7 Stamp Recognition 12.7.1 Example 1 12.7.2 Example 2 13. NEURALNET FILTER LONG-SEARCH 13.1 Long-Search File Structure 13.2 Trademark Example 13.3 Keyword-Search vs Content-Based Search 14. NEURALNET FILTER PARAMETER ADVISOR 14.1 Parameter Advisor 14.2 License Plate Recognition 15. BIOMETRICS 15.1 Attrasoft Facial Recognition Classifications 15.2 How to Build an Image Recognition Solution with the ImageFinder 15.3 Fingerprint Data 15.4 Image Processing 15.5 BioFilter 15.6 Neural Filter 15.7 NeuralNet Filter 15.8 Improvement 16. SEGLOCATOR 16.1 Coordinate System 16.2 Six Examples 16.3 Segment Output 17. CUSTOMIZED POINT-LOCATOR 17.1 Locating Eye and Nose 17.2 Feret 100 17.3 Run Your Data 17.4 Analysis 18. BIOFILTER II 18.1 Input 18.2 Data 18.2.1 Good Match 18.2.2 Bad Match 18.3 Templates 18.3.1 File Input 18.3.2 Directory Input 18.4 Training 18.5 Training the BioFilter II 18.6 1:N and N:N Matching Example 1 18.7 1:N and N:N Matching Example 2 18.8 Two-Layer Neural Net Architecture 19. NEURAL FILTER II 19.1 Overview 19.2 Training 19.3 1:N and N:N Matching Example 1 19.4 1:N Example 2 19.5 Two-Layer Neural Net Architecture 20. NEURALNET FILTER II 20.1 Overview 20.2 Positive Match Examples 20.3 Negative Match Examples 21. API 21.1 ImageFinder Internal Structures 21.2 Software Structure 21.3 Standard API 21.4 Example: Unsupervised Learning 22. IMAGEFINDER FOR DOS 22.1 Introduction 22.2 Batch Files 22.3 Batch Commands 23. REFERENCE MANUAL 23.1 Input 23.2 Image Processing 23.3 Normalization 23.4 BioFilter 23.5 Neural Filter 23.6 NeuralNet Filter 23.7 Batch Commands 23.8 BioFilter II 23.9 Neural Filter II 23.10 NeuralNet II Filter 23.11 Segment-Locator 23.12 Other Objects In the Page 23.13 Examples 24. IMAGEFINDER SUPPORT SERVICE PACKAGES 24.1 What is Support Service? 24.2 What is a Feasibility Study? 24.3 ImageFinder Services CHAPTER 25. README.TXT 25.1 Software Requirements 25.2 Install the Software 25.3 Image Recognition 25.3.1 Image Processing 25.3.2 BioFilter Templates 25.3.3 BioFilter Training and Matching 25.3.4 Neural Filters 25.3.5 NeuralNet Filter 25.3.6 Batch File
About Attrasoft ImageFinder Software Requirements Install the Software Information and Support Statement of Copyright Restriction
1. INTRODUCTION
1.1 What will the ImageFinder do? 1.2 Biometrics 1.3 Software Requirements 1.4 Customized Software 1.5 Trails
2.1 ImageFinder Internal Structures 2.2 Image Preprocessing 2.3 Normalization 2.4 Feature Recognition 2.5 ABM Matching Engine 2.6 Multi-Layered ABM 2.7 Chapter Overview
3.1 Introduction 3.1.1 Expectation 3.1.2 FaceFinder Input 3.1.3 FaceFinder 3.1.4 Restrictions 3.2 Identification 3.2.1 Operation 3.2.2 Template Matching Output 3.2.3 Checking Identification Results 3.2.4 Analysis 3.2.5 Examples 3.2.6 Identify with the Template File 3.3 Verification 3.3.1 Operation 3.3.2 Example
3.1.1 Expectation 3.1.2 FaceFinder Input 3.1.3 FaceFinder 3.1.4 Restrictions
3.2.1 Operation 3.2.2 Template Matching Output 3.2.3 Checking Identification Results 3.2.4 Analysis 3.2.5 Examples 3.2.6 Identify with the Template File
3.3.1 Operation 3.3.2 Example
4.1 Why Image Matching? 4.1.1 Expectations 4.2 Run DocumentFinder 4.2.1 Input 4.2.2 Training 4.2.3 Matching 4.3 Restrictions 4.4 Examples 4.5 Simple Matching 4.5.1 Operation 4.5.2 File Input Examples 4.5.3 Dir Input Examples 4.6 Three-Step Matching 4.6.1 Matching Procedure 4.6.2 Example “BioAPI” 4.7 Parameters 4.8 Checking the Results 4.8.1 B1_matchlist.txt 4.8.2 Example “Abm54” 4.8.3 Example “All” 4.9 Query Set Vs Target Set 4.9.1 Query Set against Target Set 4.9.2 Examples
4.1.1 Expectations
4.2.1 Input 4.2.2 Training 4.2.3 Matching
4.5.1 Operation 4.5.2 File Input Examples 4.5.3 Dir Input Examples
4.6.1 Matching Procedure 4.6.2 Example “BioAPI”
4.8.1 B1_matchlist.txt 4.8.2 Example “Abm54” 4.8.3 Example “All”
4.9.1 Query Set against Target Set 4.9.2 Examples
5.1 Introduction 5.2 Operations 5.2.1 Data 5.2.2 Commands 5.2.3 Parameters 5.3 Getting Started 5.3.1 The Problem 5.3.2 Good Match 5.3.3 Bad Match 5.4 Multiple Matches 5.4.1 Multiple Good Matches 5.4.2 Multiple Bad Matches 5.4.3 Limitation of the Software
5.2.1 Data 5.2.2 Commands 5.2.3 Parameters
5.3.1 The Problem 5.3.2 Good Match 5.3.3 Bad Match
5.4.1 Multiple Good Matches 5.4.2 Multiple Bad Matches 5.4.3 Limitation of the Software
7. BIOFILTERS
7.1 Test-Driving the Label Matching Problem, Unsupervised N:N Matching 7.2 BioFilter Overview 7.3 Unsupervised 1:N Matching 7.4 Training 7.5 Supervised N:N Matching 7.6 Supervised 1:N Matching 7.7 Checking the Results 7.8 File Input 7.9 Analysis 7.10 Summary
8.1 Test-Driving the Label Matching Problem 8.2 Neural Filter Overview 8.3 Training 8.4 N:N Matching 8.5 1:N Matching 8.6 Checking the Results 8.7 Analysis 8.8 Summary
9.1 Test-Driving the Logo Matching Problem 9.2 NeuralNet Filter Overview 9.3 Training 9.4 N:N Matching 9.5 1:N Matching 9.6 Summary
10.1 Overview 10.2 Filter Parameters 10.3 Image Processing 10.3.1 Edge Filters 10.3.2 Threshold Filters 10.3.3 Clean Up Filters 10.3.4 Starting the Selection 10.4 Normalization Filter 10.4.1 Reduction Filters 10.4.2 Parameters 10.5 BioFilter 10.6 Neural Filters 10.7 NeuralNet Filter 10.7.1 Symmetry 10.7.2 Translation Type 10.7.3 Scaling Type 10.7.4 Rotation Type 10.7.5 Area of Interest (AOI) 10.7.6 Blurring 10.7.7 Sensitivity 10.7.8 Internal/External Weight Cut 10.7.9 L/S Segments 10.7.10 Image Type 10.7.11 Output 10.7.12 Segment 10.7.13 Bypass 10.7.14 Summary
10.3.1 Edge Filters 10.3.2 Threshold Filters 10.3.3 Clean Up Filters 10.3.4 Starting the Selection
10.4.1 Reduction Filters 10.4.2 Parameters
10.7 NeuralNet Filter 10.7.1 Symmetry 10.7.2 Translation Type 10.7.3 Scaling Type 10.7.4 Rotation Type 10.7.5 Area of Interest (AOI) 10.7.6 Blurring 10.7.7 Sensitivity 10.7.8 Internal/External Weight Cut 10.7.9 L/S Segments 10.7.10 Image Type 10.7.11 Output 10.7.12 Segment 10.7.13 Bypass 10.7.14 Summary
11.1 Creating Batch Code 11.2 Sample Batch 11.3 Overview 11.4 Batch Execution Code 11.5 BioFilter Examples 11.6 NeuralFilter Examples 11.7 NeuralNet Filter Examples
12.1 Overview 12.2 Parameters 12.3 Short-Search, Long-Search, and File- Search 12.4 ImageFinder Operations for Short- Search 12.5 ImageFinder Operations for Short- Search (Advanced) 12.6 TradeMark Retrieval 12.6.1 United Way - Rotation Symmetry 12.6.2 Tabasco - Rotation Symmetry 12.6.3 Mr. Potato - Scaling Symmetry 12.6.4 Monopoly - Scaling Symmetry 12.6.5 Chemical Compound 12.7 Stamp Recognition 12.7.1 Example 1 12.7.2 Example 2
12.6.1 United Way - Rotation Symmetry 12.6.2 Tabasco - Rotation Symmetry 12.6.3 Mr. Potato - Scaling Symmetry 12.6.4 Monopoly - Scaling Symmetry 12.6.5 Chemical Compound
12.7.1 Example 1 12.7.2 Example 2
13.1 Long-Search File Structure 13.2 Trademark Example 13.3 Keyword-Search vs Content-Based Search
14.1 Parameter Advisor 14.2 License Plate Recognition
15.1 Attrasoft Facial Recognition Classifications 15.2 How to Build an Image Recognition Solution with the ImageFinder 15.3 Fingerprint Data 15.4 Image Processing 15.5 BioFilter 15.6 Neural Filter 15.7 NeuralNet Filter 15.8 Improvement
16.1 Coordinate System 16.2 Six Examples 16.3 Segment Output
17.1 Locating Eye and Nose 17.2 Feret 100 17.3 Run Your Data 17.4 Analysis
18.1 Input 18.2 Data 18.2.1 Good Match 18.2.2 Bad Match 18.3 Templates 18.3.1 File Input 18.3.2 Directory Input 18.4 Training 18.5 Training the BioFilter II 18.6 1:N and N:N Matching Example 1 18.7 1:N and N:N Matching Example 2 18.8 Two-Layer Neural Net Architecture
18.2.1 Good Match 18.2.2 Bad Match
18.3.1 File Input 18.3.2 Directory Input
19.1 Overview 19.2 Training 19.3 1:N and N:N Matching Example 1 19.4 1:N Example 2 19.5 Two-Layer Neural Net Architecture
20.1 Overview 20.2 Positive Match Examples 20.3 Negative Match Examples
21.1 ImageFinder Internal Structures 21.2 Software Structure 21.3 Standard API 21.4 Example: Unsupervised Learning
22.1 Introduction 22.2 Batch Files 22.3 Batch Commands
23.1 Input 23.2 Image Processing 23.3 Normalization 23.4 BioFilter 23.5 Neural Filter 23.6 NeuralNet Filter 23.7 Batch Commands 23.8 BioFilter II 23.9 Neural Filter II 23.10 NeuralNet II Filter 23.11 Segment-Locator 23.12 Other Objects In the Page 23.13 Examples
24.1 What is Support Service? 24.2 What is a Feasibility Study? 24.3 ImageFinder Services
CHAPTER 25. README.TXT
25.1 Software Requirements 25.2 Install the Software 25.3 Image Recognition 25.3.1 Image Processing 25.3.2 BioFilter Templates 25.3.3 BioFilter Training and Matching 25.3.4 Neural Filters 25.3.5 NeuralNet Filter 25.3.6 Batch File
25.3.1 Image Processing 25.3.2 BioFilter Templates 25.3.3 BioFilter Training and Matching 25.3.4 Neural Filters 25.3.5 NeuralNet Filter 25.3.6 Batch File
Send questions or comments to: webmaster@attrasoft.com or contact us at: (912) 484-1717, (912) 897 - 1717 Attrasoft,P. O. Box 13051, Savannah, GA. 31406, USA Copyright © 2004 Attrasoft, Inc. All rights reserved.
TradeMark; Attrasoft