Table of Contents
Preface v
Introduction xi
1 Feature Extraction Based on Parallel Image Shift Technologies 1
1.1 Fundamentals of Understanding and of Constructing Images 1
1.2 Image Description 2
1.3 Characteristic Features of Images 5
1.4 Theoretical Description of the Parallel Shift Technology 7
2 Function of the Area of Intersection Formation Based on Parallel Shift Technology 14
2.1 Mathematical Model of the Image Describing the Function of the Area of Intersection 14
2.2 Analytical Determination of the Function of the Area of Intersection of Simple Shapes on the Basic Parameters 18
2.3 Additional Parameters, which Accelerate Plane Image Recognition Process 25
2.4 Algorithm of Recognition Images Based on PST 29
2.5 Processing of 3D-Objects through PST 33
3 Methods of Pre-Processing Images Based on PST 35
3.1 Edge Detection Method of Plane Figures 37
3.2 Use of Parallel Shift Technology for Noise Removal in Images 40
3.3 Technical Methods for FAI Preparation 47
3.4 Model for Obtaining FAI of No Bitmap 49
4 Methods of Dynamic Images Processing Based on PST 50
4.1 Determination of Spatial Orientation of Planes 52
4.2 Use of Different Types of Shifts of Image Copies in Video Processing Devices Based on PST 55
4.3 Methods for Determining Movement Parameters Using Cyclic Shifts 57
4.4 Methods for Determination of Motion Parameters Using Non-cyclic Shifts 59
4.5 Methods for Determining Distance of a Moving Object 61
4.6 Comparison of Methods for Determination of Motion Parameters by Using Different Types of Shifts of the Image Copy 63
5 Image Processing System Based on PST 65
5.1 System of Perception of Video Information 66
5.2 Saving Template Functions of the Area of Intersection 72
5.3 Application of Image-processing Methods Based on PST for Prosthetic Visual Organs 77
5.4 Perspective Directions of Image-processing Research Using PST 80
6 Assessment of Productivity of the Image Processing and Recognition Systems Based on FAI Analysis 84
6.1 Evaluating the Efficiency of the Chosen Method of Removing Noise 85
6.2 Determination of the Permissible Error Range of Values Used to Limit the Maximum Displacement Values 90
6.3 Determination of Values Range of Acceptable Error Used in Comparison of Integral Coefficients 91
6.4 Determination of Range Values of Acceptable Error in Comparison of FAI on a Detailed Stage of the Analysis of Image Recognition 93
6.5 Evaluating the Effectiveness of Image Recognition Process in Various Parameters of Input Information Using Calculated Values of Permissible Errors 94
7 Hardware Implementation of the Image Recognition System Based on Parallel Shift Technology 101
7.1 Generalised Model of Image Recognition System Based on Parallel Shift Technology 101
7.2 Selecting a Cover for Representing the Image in the Recognition System 102
7.3 Hardware Implementation of Parallel Shift and Computation of the Area of Intersection of Images 104
7.4 Modelling of the Main Hardware Modules for Determining FAI in Modern CAD Systems 113
7.5 Selection and Storage of Patterns in UMFP 118
7.6 Optical Image Recognition System Based on Parallel Shift Technology 120
8 Methodology of Experimental Studies in Recognition Process Based on Parallel Shift Technology 126
8.1 Selection and Analysis of Characteristic Features of Images Based on Parallel Shift Technology 126
8.2 Analysis of the Accuracy of Image Recognition, Based on the Obtained FAI 131
8.3 Description of Image with the Help of Quantitative Characteristics of the Basic Properties of FAI 148
8.4 Experimental Analysis of FAI of Patterns Image 157
8.5 The Sequence of Steps for Using Parallel Shift Technology to Describe and Recognise Images 163
8.6 Advantages of Text Recognition Based on Parallel Shift Technology 166
References 187
Index 192
