Real-time rendering : computer graphics with control engineering

Consumers today expect extremely realistic imagery generated in real time for interactive applications such as computer games, virtual prototyping, and scientific visualisation. However, the increasing demands for fidelity coupled with rapid advances in hardware architecture pose a challenge: how do you find optimal, sustainable solutions to accommodate both speed of rendering and quality? Real-Time Rendering: Computer Graphics with Control Engineering presents a novel framework for solving the perennial challenge of resource allocation and the trade-off between quality and speed in interactive computer graphics rendering.

Conventional approaches are mainly based on heuristics and algorithms, are largely application specific, and offer fluctuating performance, particularly as applications become more complex. The solution proposed by the authors draws on powerful concepts from control engineering to address these shortcomings. Expanding the horizon of real-time rendering techniques, this book:

Explains how control systems work with real-time computer graphics Proposes a data-driven modelling approach that more accurately represents the system behaviour of the rendering process Develops a control system strategy for linear and non-linear models using proportional, integral, derivative (PID) and fuzzy control techniques Uses real-world data from rendering applications in proof-of-concept experiments Compares the proposed solution to existing techniques Provides practical details on implementation, including references to tools and source code

This pioneering work takes a major step forward by applying control theory in the context of a computer graphics system. Promoting cross-disciplinary research, it offers guidance for anyone who wants to develop more advanced solutions for real-time computer graphics rendering.

List of Figures	p. xiii
List of Tables	p. xvii
List of Abbreviations	p. xix
Preface	p. xxi
Acknowledgements	p. xxiii
Summary	p. xxv
Authors	p. xxvii
Chapter 1 Introduction	p. 1
1.1 Background and Motivation	p. 1
1.2 Objectives and Contributions	p. 2
1.3 Scope of Work	p. 3
1.4 Book Outline	p. 3
Chapter 2 Preliminaries	p. 5
2.1 Fundamentals of Real-Time 3D Rendering	p. 5
2.1.1 Polygon-Based Rendering	p. 5
2.1.2 Volumetric Rendering	p. 8
2.1.3 Image-Based Rendering	p. 9
2.2 System Identification	p. 10
2.2.1 Data Collection	p. 11
2.2.2 Model Selection	p. 12
2.2.3 Computing Model Parameters	p. 13
2.2.4 Evaluating Quality of Derived Model	p. 13
2.3 Literature Review	p. 14
2.3.1 Comparative Study on Existing Research	p. 14
2.3.2 Control-Theoretic Approaches to Computer Systems	p. 16
2.3.3 Control Principles in Computer Graphics Software	p. 17
Chapter 3 Linear Model Analysis of Real-Time Rendering	p. 19
3.1 Introduction	p. 19
3.2 Background	p. 19
3.2.1 Control-Centric Definition for Rendering Time Control	p. 21
3.2.2 Challenges in Using Heuristics	p. 21
3.2.3 Purpose of Workload Characterisation and Analysis	p. 22
3.3 Case for Data-Driven Modelling	p. 23
3.3.1 Basis for Selection of System Variables	p. 23
3.4 Linear System Model Representation for Real-Time Rendering	p. 25
3.5 Experiments	p. 27
3.5.1 Experiment 1: Single-Input-Single-Output (SISO) System	p. 27
3.5.2 Experiment 2: Multiple-Input-Single-Output (MISO) System	p. 28
3.5.3 Experiment 3: Control Framework Using System Model	p. 29
3.6 Results	p. 30
3.6.1 Experiment 1	p. 30
3.6.2 Experiment 2	p. 33
3.6.3 Experiment 3	p. 38
3.7 Discussion	p. 40
3.7.1 Comparison with Other Estimation Techniques	p. 41
3.8 Superposition in 3D Rendering System Model	p. 43
3.8.1 Principle of Superposition	p. 43
3.8.2 Experiment	p. 44
3.8.3 Simulation	p. 46
3.8.4 Summary	p. 48
3.8.5 Additional Notes	p. 49
3.9 Conclusion	p. 49
Chapter 4 Modelling Non-Linear Rendering Processes	p. 51
4.1 Introduction	p. 51
4.2 Background	p. 51
4.2.1 System Modelling with Neural Networks	p. 51
4.2.2 Systems Modelling with Fuzzy Logic	p. 53
4.3 Experiments	p. 56
4.3.1 Time Delay Neural Network	p. 56
4.3.2 Adaptive Neuro-Fuzzy Inference System	p. 56
4.4 Experiment Results	p. 60
4.4.1 Time Delay Neural Networks	p. 60
4.4.2 ANFIS Model	p. 61
4.5 Discussion	p. 63
4.6 Linearised Approximation from Non-Linear Models	p. 64
4.7 Conclusion	p. 66
Chapter 5 Model-Based Control	p. 67
5.1 Introduction	p. 67
5.2 Control System Perspective of Computer Graphics Rendering Process	p. 67
5.2.1 Control System Architectures for Real-Time Rendering	p. 68
5.2.2 Control System Performance Concepts Applicable to Real-Time Rendering	p. 70
5.3 PID Control and Tuning	p. 71
5.3.1 Implementing PID Control for Rendering Process	p. 72
5.3.2 Data Preprocessing in PID Control System	p. 75
5.3.3 Gain Scheduling for Non-Linear Rendering Process Models	p. 76
5.3.4 Neural PID Control	p. 79
5.4 Experiments	p. 81
5.5 Results	p. 83
5.5.1 Simulation Environment	p. 83
5.5.2 Control System with Actual Rendering Process	p. 83
5.5.3 Gain Scheduling Control System	p. 85
5.6 Conclusion	p. 86
Chapter 6 Model-Less Control	p. 89
6.1 Introduction	p. 89
6.2 Fuzzy Control System	p. 89
6.3 Adaptive Neural Fuzzy Control	p. 90
6.4 Experiment	p. 92
6.5 Results	p. 95
6.5.1 Simulation	p. 95
6.5.2 Fuzzy Control System with Rendering Process	p. 95
6.6 Discussion	p. 97
6.7 Conclusion	p. 98
Chapter 7 Applications, Challenges, and Possibilities	p. 99
7.1 System Architectures	p. 99
7.1.1 Software Design	p. 101
7.2 Software and Hardware Performance Considerations	p. 103
7.2.1 Data Integrity	p. 103
7.2.2 Plant-Controller Communication Latency	p. 103
7.2.3 Data Structures and Handling	p. 103
7.2.4 Complexity of Control Algorithm	p. 104
7.3 Applications of Rendering Control Systems	p. 104
7.3.1 Extension of Control System Framework	p. 105
7.4 Convergence with Future Technology	p. 105
7.4.1 Greater Computing Parallelism	p. 105
7.4.2 Increased Use of Mobile Devices	p. 105
7.4.3 Vast Improvements in Internet Infrastructure	p. 106
7.5 Economic and Productivity Impacts	p. 106
7.5.1 Enhanced Product Lifespan	p. 106
7.5.2 Increased Productivity	p. 106
7.5.3 New Products and Markets	p. 107
Chapter 8 Conclusion	p. 109
8.1 Performance Analysis	p. 109
8.1.1 Frame Rate Stability	p. 109
8.1.2 Transient Response	p. 110
8.1.3 Adaptive Tracking Capability	p. 112
8.2 Summary	p. 117
8.3 Future Work	p. 118
Annex A Sample Applications	p. 121
A.1 Overview	p. 121
A.2 ProgressiveMesh Sample	p. 121
A.3 How Sample Works	p. 121
A.4 Tessellation Sample	p. 122
A.5 How Sample Works	p. 122
A.6 Samples	p. 122
Annex B Patent for Application Method and System for Adaptive Control of Real-Time Computer Graphics Rendering	p. 153
Title of Invention	p. 153
Field of Invention	p. 153
Background of Invention	p. 153
Summary of Invention	p. 154
Brief Descriptions of Figures	p. 155
Detailed Descriptions of Figures	p. 155
Control Design and Mechanism	p. 156
I PID Gain Scheduling	p. 156
II Fuzzy Control (Model-Less Control)	p. 159
Claims (Preliminary)	p. 161
Annex C Neural PID Control System Code	p. 167
References	p. 171
Publications and Achievements	p. 177
Patent Application	p. 177
Book	p. 177
Book Chapters	p. 177
Conference Papers	p. 177
Achievements	p. 178
Index	p. 179

Available:*

On Order

Summary

Summary

Table of Contents