Fig. 1. Single-pixel imaging experimental devices^[16]. (a) Passive single-pixel imaging; (b) active single-pixel imaging

Fig. 2. Image processing algorithms related to single-pixel imaging

Fig. 3. Single-pixel imaging and neural network structure

Fig. 4. Sensing matrix + CNN structure^[59]

Fig. 5. Sensing matrix+nonlinear operator+CNN structure^[61]

Fig. 6. Motion compensation of illumination patterns^[63]. (a)(c) Object motion state; (b)(d) transform of illumination patterns

Fig. 7. Simulation and reconstruction results of single pixel imaging of dynamic objects^[63]. (a) Numbers on the disc; (b) reconstructed results using conventional methods when the disc is rotated at 4 rounds per second; (c) reconstructed results using illumination pattern motion compensation method when the disc is rotated at 4 rounds per second; (d) reconstructed results using conventional methods when the disc is rotated at 8 rounds per second; (e) reconstructed results using illumination pattern motion compensation method when the disc is rotated at 8 rounds per second

Fig. 8. EENet architecture^[73]

Fig. 9. DAttNet architecture^[74]

Fig. 10. Scheme of optical encryption device based on single-pixel imaging^[77]

Fig. 11. Labyrinths optical encryption system based on single-pixel imaging^[82]

Fig. 12. Single-pixel encryption system based on 3D sparse space^[86]

Fig. 13. Security enhancement scheme based on reversible matrix modulation^[87]

Fig. 14. Optical data container and XOR encoding^[89]. (a) Plaintext coding result; (b) key bit stream; (c) XOR encoding result

Fig. 15. Watermark image encryption^[92]

Fig. 16. Single-pixel visual cryptography system^[95]

Fig. 17. Edge detection results in the frequency domain based on single-pixel imaging^[104]. (a)(c)(e)(g) Original images; (b)(d)(g)(h) edge detection results

Fig. 18. Masks with low spatial frequency^[105]. (a) Conventional random mask; (b) masks with low spatial frequency

Fig. 19. Illumination pattern adaptive. (a) Using K-SVD algorithm^[109]; (b) using PCA algorithm^[110]

Fig. 20. Single-pixel imaging network frame based on deep neural networks^[117]

Fig. 21. Deep convolutional adaptive network structure^[122]