Title Page
Contents
국문요약 13
1. Introduction 16
1.1. Research Background And Purpose 16
1.1.1. Research Background 16
1.1.2. Research Purpose 20
1.1.3. Research significance 21
1.2. Research Scope and Method 23
1.2.1. Research scope 23
1.2.2. Research Method 26
1.2.3. Research Model 28
1.3. Research steps and thesis structure 29
2. Theoretical investigation 32
2.1. Research status of augmented reality 32
2.2. Research status of face feature point detection 43
2.3. Digital development user experience related theoretical basis 56
2.3.1. User experience theory of digital development 56
2.3.2. User satisfaction of software systems 58
2.4. Peking Opera facial makeup culture 59
2.5. Existing problems 62
3. Digital drawing and Deformation Processing of Peking Opera facial makeup 64
3.1. Feature analysis of Peking Opera facial makeup 64
3.2. Collection of Peking Opera facial makeup patterns 67
3.3. Deformation processing of Peking Opera facial makeup 70
3.3.1. Corresponding processing of Peking Opera facial makeup and facial feature points 70
3.3.2. Facial makeup pattern deformation processing 71
3.4. Experimental results and analysis 73
4. Key technologies and improvements of Peking Opera's facial makeup augmented reality 75
4.1. Face detection and feature point location based on recursive shape reconstruction 75
4.1.1. The shape increment estimation method 79
4.1.2. Recursive shape reconstruction detection model 81
4.1.3. Parameter learning of the detection model 85
4.1.4. Experiment 92
4.2. A facial feature point enhancement method for model compression 104
4.2.1. Face shape detection in local area 107
4.2.2. Enhancement of face shape in local areas 110
4.2.3. Parameter learning of reconstruction model 114
4.2.4. Experiment 115
4.3. Summary of this chapter 126
5. Peking Opera facial makeup Augmented Reality Application Design and Implementation 128
5.1. Real-time video augmented reality system framework of Peking Opera facial makeup 130
5.2. Implementation of Peking Opera facial makeup augmented reality system based on facial dynamics 138
5.2.1. Face detection 138
5.2.2. Face pose calculation 142
5.2.3. Face feature point detection under complex facial poses 145
5.3. Digital facial dynamics of Peking Opera facial makeup 151
5.4. Summary 156
6. Peking Opera facial makeup Augmented Reality Application User Experience Analysis 158
6.1. User experience analysis of Peking Opera facial makeup augmented reality application 158
6.1.1. Peking Opera facial makeup augmented reality user experience test 159
6.1.2. Analysis of user experience test results 161
6.2. Users' satisfaction with the performance of Peking Opera facial makeup augmented reality application 167
6.2.1. User satisfaction test of performance 167
6.2.2. User satisfaction results analysis of performance 169
6.3. Significance of augmented reality development of Peking Opera facial makeup 173
6.3.1. Transform the expression medium of traditional Peking Opera facial makeup 174
6.3.2. Change the experience mode of traditional Peking Opera facial makeup 176
6.3.3. Expand the expression content of Peking Opera facial makeup 177
6.4. Influence on the film and television industry 178
6.5. Summary 180
7. Conclusion 182
7.1. Conclusion 182
7.2. Future works 186
Reference 188
ABSTRACT 198
〈Table 4-1〉 Comparison result of different parameter sets on LFPW (68 face feature points)... 87
〈Table 4-2〉 The average error results of different parameters on the LFPW (68 face... 88
〈Table 4-3〉 Average error results of specific parameter sets under different iteration times 89
〈Table 4-4〉 Average error results of specific parameters under different initialization times 89
〈Table 4-5〉 Error results of multi-parameter strategy on LFPW (68 face feature points)... 90
〈Table 4-6〉 The average calibration error excludes the average error of each area of the... 100
〈Table 4-6〉 The average calibration error excludes the average error of each area of the... 101
〈Table 4-8〉 Refactoring of different strategies in LFPW database 119
〈Table 6-1〉 User Experience Test Questionnaire 161
〈Table 6-2〉 User experience test score statistics table 162
〈Table 6-3〉 Data analysis of individual user experience indicators 163
〈Table 6-4〉 Analysis of user experience index classification data 164
〈Table 6-2〉 User satisfaction with performance test questionnaire 169
〈Table 6-6〉 User satisfaction score of performance 171
〈Table 6-7〉 Statistical table of data of single index of performance 172
〈Table 6-8〉 Statistical table of indicators of Performance 172
〈Figure 1-1〉 Schematic diagram of facial feature points on human face. (a3) Schematic... 24
〈Figure 1-2〉 Face images of different poses in unconstrained natural environment 25
〈Figure 1-3〉 Research Model 28
〈Figure 1-4〉 Thesis Process 31
〈Figure 2-1〉 Workflow of Argumented Reality 33
〈Figure 2-2〉 Milgram Reality 33
〈Figure 2-3〉 classification of augmented reality based on augmented objects 35
〈Figure 2-4〉 Classification of face feature point detection technology 44
〈Figure 2-5〉 Sheng, Dan, Jing, Mo, Chou facial makeup 62
〈Figure 3-1〉 Different kinds of facial makeup 66
〈Figure 3-2〉 The structure of Peking Opera facial makeup 67
〈Figure 3-3〉 The corresponding pattern segmentation structure of Peking Opera facial makeup 69
〈Figure 3-4〉 Vector graphics of Peking Opera facial makeup 70
〈Figure 3-5〉 Marking of some pattern control points 71
〈Figure 3-6〉 The orientation of the facial feature... 71
〈Figure 3-7〉 Deformation control point 72
〈Figure 3-8〉 Deformation of some patterns 74
〈Figure 4-1〉 Schematic diagram of recursive shape reconstruction method 79
〈Figure 4-2〉 Recursive shape reconstruction detection model diagram 83
〈Figure 4-3〉 The positioning result of the shape... 92
〈Figure 4-4〉 Comparison of CED curves of four methods... 97
〈Figure 4-5〉 Comparison of the average calibration errors of the... 98
〈Figure 4-6〉 Comparison of the average calibration errors of the four methods tested... 102
〈Figure 4-7〉 CED curve comparison of the four methods in the local area of the face... 104
〈Figure 4-8〉 GSR method flow chart 107
〈Figure 4-9〉 C-GSR method diagram 112
〈Figure 4-10〉 S-GSR method diagram 114
〈Figure 4-11〉 Parameter estimation on LFPW database 115
〈Figure 4-12〉 Refactoring strategy evaluation on LFPW... 117
〈Figure 4-13〉 Reconstruct the CED curve comparison graph of different shape subsets... 120
〈Figure 4-14〉 Experimental comparison of five reconstructed structures on... 122
〈Figure 4-15〉 A comparison chart of the detection results of the... 124
〈Figure 4-16〉 Comparison of CED curves of different methods under artificial... 125
〈Figure 5-1〉 Optical perspective augmented reality system flow chart 131
〈Figure 5-2〉 Flow chart of video perspective augmented reality system 132
〈Figure 5-3〉 Peking Opera facial makeup augmented reality design structure 134
〈Figure 5-4〉 Peking Opera facial makeup augmented reality... 135
〈Figure 5-5〉 Peking Opera facial makeup augmented reality application interaction process 136
〈Figure 5-6〉 Pixel feature map 140
〈Figure 5-7〉 Multi-gesture face image 143
〈Figure 5-8〉 Facial pose calculation chart 145
〈Figure 5-9〉 Partial facial feature point models that can be trained offline for four poses 146
〈Figure 5-10〉 Virtual and real face fusion under different facial... 153
〈Figure 5-11〉 The fusion of virtual and real faces with different deflection angles. (a1)-(a4)... 154
〈Figure 5-12〉 The fusion of virtual and real faces with different pitch angles. (a1)-(a4) fusion... 154
〈Figure 5-13〉 Fusion of virtual and real faces with expressions. (a) Face fusion in the... 155
〈Figure 5-14〉 Virtual and real face fusion under different occlusions 156
〈Figure 5-15〉 Digitalized "Peking Opera Face Change" 156
〈Figure 6-1〉 Peking Opera facial makeup Augmented reality application user experience... 161
〈Figure 6-2〉 Evaluation index of performance 168
〈Figure 6-3〉 "Star Wars" CGI technology drawing actress 179
〈Figure 6-4〉 A deepfake creation model using two encoder-decoder pairs 180
〈Figure 6-5〉 Face-to-face experiment in video 180