Human and Machine Vision II

Description

Perspectives in Computing: Human and Machine Vision II compiles papers presented at the second Workshop on Human and Machine Vision held in Montreal, Canada on August 1-3, 1984.

This book discusses the perception of transparency in man and machine, human image understanding, and connectionist models and parallelism in high level vision. The theory of the perceived spatial layout of scenes, generative systems of analyzers, and codon constraints on closed 2D shapes are also elaborated. This text likewise covers the environment- and viewer-centered perception of surface orientation, autonomous scene description with range imagery, and pre-attentive processing in vision.

This publication is recommended for students and researchers interested in both fields of visual perception and computer vision.

Front Cover 2
Human and Machine Vision II 5
Copyright Page 6
Table of Contents 7
Preface 9
Contributors 11
Chapter 1. Perception of Transparency in Man and Machine 13
- 1. INTRODUCTION 13
- 2. ADDITIVE COLOR MIXTURE 14
- 3. SUBTRACTIVE COLOR MIXTURE 16
- 4. NONVERIDICAL PERCEPTION OF TRANSPARENCY 18
- 5. REFLECTANCE VS LIGHTNESS 19
- 6. PERCEPTION OF THE DEGREE OF TRANSPARENCY 20
- 7. FIGURAL CUES 22
- 8. HUMAN AND MACHINE JUDGMENTS OF TRANSPARENCY 24

Front Cover 2
Human and Machine Vision II 5
Copyright Page 6
Table of Contents 7
Preface 9
Contributors 11
Chapter 1. Perception of Transparency in Man and Machine 13
- 1. INTRODUCTION 13
- 2. ADDITIVE COLOR MIXTURE 14
- 3. SUBTRACTIVE COLOR MIXTURE 16
- 4. NONVERIDICAL PERCEPTION OF TRANSPARENCY 18
- 5. REFLECTANCE VS LIGHTNESS 19
- 6. PERCEPTION OF THE DEGREE OF TRANSPARENCY 20
- 7. FIGURAL CUES 22
- 8. HUMAN AND MACHINE JUDGMENTS OF TRANSPARENCY 24
- REFERENCES 24
Chapter 2. Human Image Understanding: Recent Research and a Theory 25
- RECOGNITION: UNITS AND CATEGORIES 26
- THEORETICAL DOMAIN: PRIMAL ACCESS 27
- BASIC PHENOMENA OF OBJECT RECOGNITION 28
- RECOGNITION-BY-COMPONENTS 28
- A PERCEPTUAL BASIS FOR A COMPONENTIAL REPRESENTATION 31
- COMPONENTS GENERATED FROM DIFFERENCES IN NONACCIDENTAL PROPERTIES AMONG GENERALIZED CONES 34
- A LIMITED NUMBER OF COMPONENTS? 41
- RELATIONS OF RBC TO PRINCIPLES OF PERCEPTUAL ORGANIZATION 41
- EXPERIMENTAL SUPPORT FOR A COMPONENTIAL REPRESENTATION 46
- SUMMARY AND IMPLICATIONS OF THE EXPERIMENTAL RESULTS 61
- CONCLUSION 66
- REFERENCES 66
Chapter 3. Describing Surfaces 70
- 1. INTRODUCTION 70
- 2. SURFACE CURVES 72
- 3. COMPUTATIONAL EXPERIMENTS 86
- APPENDIX A 94
- APPENDIX B: THE GENERALIZED CONE THEOREM 94
- ACKNOWLEDGMENTS 95
- REFERENCES 95
Chapter 4. Connectionist Models and Parallelism in High Level Vision 98
- 1. INTRODUCTION 98
- 2. CONNECTIONIST MODELS 101
- 3. PARALLEL VISUAL RECOGNITION 108
- 4. LIMITS ON PARALLELISM 114
- REFERENCES 118
Chapter 5. Toward a Theory of the Perceived Spatial Layout of Scenes 121
- 1. INTRODUCTION 121
- 2. MEASURING PERCEIVED LAYOUT FROM RESPONSES ABOUT A SCENE 131
- 3. EXPERIMENTAL PROCEDURE TO MEASURE PRECEIVED LAYOUT 132
- 4. RESULTS OF EMPIRICAL MEASUREMENTS OF PERCEIVED LAYOUT 135
- 5. HOW GOOD ARE THESE MEASURES OF PERCEIVED LAYOUT OF SPACE 144
- 6. PROPERTIES OF THE PERCEIVED LAYOUT OF SPACE 146
- 7. VARIABLES THAT AFFECT THE PERCEIVED LAYOUT OF SPACE 148
- 8. THE THEORETICAL IMPORTANCE OF THE STUDY OF PERCEIVED LAYOUT 153
- REFERENCES 156
- ACKNOWLEDGMENTS 156
Chapter 6. Generative Systems of Analyzers 161
- 2. AN EXAMPLE 161
- 1. INTRODUCTION 161
- 3. THE GENERATIVE STRUCTURE 168
- 4. THE STRUCTURE OF GROUPING 172
- 5. WIRING TOGETHER THE EXTERNAL AND INTERNAL ANALYZER SYSTEMS 178
- 6. ANALYZER SYSTEMS ENCODING THE CARTESIAN REFERENCE FRAMES 181
- 7. THE JOHANSSON MOTION PHENOMENON 191
- 8. THE ANALYSIS OF COMPLEX SHAPE 193
- REFERENCES 200
Chapter 7. Early Vision: From Computational Structure to Algorithms and Parallel Hardware 202
- 1. INTRODUCTION 202
- 2. ANALOG NETWORKS FOR SOLVING VARIATIONAL PROBLEMS 209
- 3. CONCLUSION 214
- REFERENCES 216
- ACKNOWLEDGMENTS 216
Chapter 8. Codon Constraints on Closed 2D Shapes 219
- 2. THE REPRESENTATION 219
- 1. INTRODUCTION 219
- 3. PART DESCRIPTORS: CODONS 220
- 4. CONSTRAINTS ON SMOOTH CODON STRINGS 222
- 5. CLOSED CODONS 224
- 6. MIRROR REVERSAL, HOLES, AND FIGURE-GROUND 225
- 7. INDEX DEVELOPMENT 226
- 8. MAPPING 3D ~ 2D 227
- APPENDIX I: A BINARY MAPPING FOR CODON STRINGS 229
- 9. SUMMARY 229
- APPENDIX II: NUMBERS OF POSSIBLE OPEN, SMOOTH CODON STRINGS 230
- APPENDIX III: NUMBER OF SMOOTH CLOSED CODON STRINGS 232
- ACKNOWLEDGMENTS 235
- REFERENCES 235
Chapter 9. Environment-Centered and Viewer-Centered Perception of Surface Orientation 236
- 1. INTRODUCTION 236
- 2. EXPERIMENT I 238
- 3. EXPERIMENT II 243
- 4. CONCLUSIONS 246
- REFERENCES 247
- ACKNOWLEDGMENTS 247
Chapter 10. Perception of Organization in a Random Stimulus 249
- ACKNOWLEDGEMENT 253
- REFERENCES 254
Chapter 11. Autonomous Scene Description with Range Imagery 255
- 1. INTRODUCTION 255
- 2. TAXONOMY OF CONTOURS IN LIGHT-STRIPE IMAGES 256
- 3. CONTOUR ANALYSIS 258
- 4. SURFACE ANALYSIS 260
- 5. OBJECT ANALYSIS 264
- 6. MULTIPLE SCENE ANALYSIS 265
- REFERENCES 267
- 7. SUMMARY 267
- ACKNOWLEDGMENTS 267
Chapter 12. Intelligible Encoding of ASL Image Sequences at Extremely Low Information Rates 268
- 1. INTRODUCTION AND OVERVIEW 268
- 2. EARLY STUDIES OF COMPRESSED ASL 271
- 3. COMPRESSION TECHNIQUES 275
- 4. EXPERIMENT 1. RATINGS OF 81 IMAGE TRANSFORMATIONS 284
- 5. EXPERIMENT 2. INTELLIGIBILITY TESTS OF 12 IMAGE TRANSFORMATIONS 288
- 6. EXPERIMENT 3. POLYGONAL APPROXIMATIONS 297
- 7. COMPRESSED CODES FOR ASL IMAGE SEQUENCES 304
- 8. THE TRADEOFFS BETWEEN INTELLIGIBILITY AND BITS PER SECOND 315
- REFERENCES 321
- ACKNOWLEDGMENTS 321
Chapter 13. Preattentive Processing in Vision 325
- ACKNOWLEDGMENTS 345
- REFERENCES 345
Chapter 14. Early Orientation Selection: Tangent Fields and the Dimensionality of Their Support 347
- 1. INTRODUCTION 347
- 2. THE FRAMEWORK 349
- 4. TANGENTS, VECTOR FIELDS, AND THE RECOVERY OF CONTOURS 351
- 5. OUTLINE OF A MODEL FOR ORIENTATION SELECTION 351
- 3. STATIC AND DYNAMIC FLOWS 351
- 6. DOT PATTERNS AND THE GEOMETRY OF ORIENTATION SELECTION 352
- 7. A MODEL FOR EARLY ORIENTATION SELECTION AND GROUPING 353
- 8. RESPONSE MATCHING PROBLEMS AND FUNCTIONAL MINIMIZATION 358
- 9. RESULTS OF THE ORIENTATION SELECTION ALGORITHM 361
- 10. PREDICTIONS 361
- 11. FROM CONTOURS TO HAIR PATTERNS 365
- 12. TYPE I AND TYPE II PROCESSES 366
- 13. ESTIMATING TANGENTS IN TYPE II PATTERNS 367
- 14. THE DIFFERENCE BETWEEN TYPE I AND TYPE II PROCESSES 368
- 15. PSYCHOPHYSICAL PREDICTIONS FOR TYPE II PATTERNS 370
- 16. SIMPLE AND COMPLEX CELLS 371
- 17. CONCLUSIONS 374
- ACKNOWLEDGMENTS 375
- REFERENCES 375
PERSPECTIVES IN COMPUTING 377

Description

TOC

Reviews (0)