Contents
xi
5 Optimization Algorithm and RFID System Physical
Anti-Collision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
5.1 Physical Anti-Collision Based on Particle Swarm
Optimization (PSO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
5.1.1 Design and Application of Detection System . . . . . . . . . . . . . 158
5.1.2 Establishment of 3D Tag Network Based on Template
Matching and Edge Detection . . . . . . . . . . . . . . . . . . . . . . . . . . 160
5.1.3 Prediction of Tag Distribution Based on PSO Neural
Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.2 Physical Anti-Collision Based on Support Vector Machine
(SVM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.2.1 RFID Detection System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.2.2 Position Location of Tags . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.2.3 Predict Model of RFID Tags’ Distribution Based
on SVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
5.3 Physical Anti-Collision Based on Wavelet . . . . . . . . . . . . . . . . . . . . . . 182
5.3.1 Design of Image Analysis System . . . . . . . . . . . . . . . . . . . . . . 182
5.3.2 3D Coordinate Measurement of RFID Tag . . . . . . . . . . . . . . . 184
5.3.3 RFID Multi-tag Wavelet Neural Network . . . . . . . . . . . . . . . . 193
5.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6 Deep Learning and RFID System Physical Anti-Collision . . . . . . . . . . 201
6.1 RFID Multi-tag 3D Measurement System . . . . . . . . . . . . . . . . . . . . . . 202
6.1.1 System Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
6.1.2 Image Process Module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
6.2 The Theory RFID Multi-tag Image Denoising by FDnCNN . . . . . . . 207
6.2.1 The Proposed FDnCNN Network Architecture . . . . . . . . . . . 207
6.2.2 Noise Level Map . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6.2.3 Loss Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
6.2.4 Experimental Result and Real-Time Analysis . . . . . . . . . . . . 211
6.3 Image Motion Blur Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
6.4 Multi-level Wavelet CNN for Image Restoration
in Pre-Processing Sub-System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
6.4.1 Image Restoration Based on Denoising Prior . . . . . . . . . . . . . 219
6.4.2 Half Quadratic Splitting (HQS) . . . . . . . . . . . . . . . . . . . . . . . . 221
6.4.3 Method from Multi-level WPT to MWCNN . . . . . . . . . . . . . 222
6.4.4 Network Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
6.5 Multi-tag 3D Position Measurement Based on Image Match . . . . . . 226
6.5.1 The Image Matching Method . . . . . . . . . . . . . . . . . . . . . . . . . . 226
6.5.2 RFID Tag 2D Position Measurement . . . . . . . . . . . . . . . . . . . . 227
6.5.3 RFID Tag Vertical Position Measurement . . . . . . . . . . . . . . . . 227