Fig. 1. Schematic diagram of computational imaging link

Fig. 2. Computational imaging link and classification

Fig. 3. Principle of wavefront shaping based on feedback^［16］

Fig. 4. Image result with incoherent light^［18］

Fig. 5. Scattering imaging based on optical TM^［19］

Fig. 6. Experimental results in complex scattering medium^［24］

Fig. 7. Scattering imaging base on OPC^［25］

Fig. 8. The imaging of 2.5 cm isolated chicken breast tissue^［28］

Fig. 9. Schematic diagram of SBSC imaging principle^［32］

Fig. 10. Rotation tracking results of different objects^［39］

Fig. 11. Active NLOS imaging based on streak tube camera^［41］

Fig. 12. NLOS imaging principle and reconstruction result with obstruction^［47］

Fig. 13. NLOS imaging of spatially coherent^［48］

Fig. 14. Shape recovery from coherence measurements^［48］

Fig. 15. NLOS imaging of intensity-coherent^［49］

Fig. 16. Reconstruction results of different hidden scenes^［49］

Fig. 17. Reconstruction results of character^［51］

Fig. 18. The dehazing results in the real scene^［58］

Fig. 19. Imaging result and evaluation curve^［61］

Fig. 20. Comparison of imaging results^［66］

Fig. 21. Restoration results of different targets^［69］

Fig. 22. Comparison of restoration results in water with different turbidity^［70］

Fig. 23. Normal vector of microfacet^［71］

Fig. 24. The principle of polarization 3D imaging^［72］

Fig. 25. Reconstruction results of different target objects^［75］

Fig. 26. Reconstruction results of different target objects^［79］

Fig. 27. 3D reconstruction results in different environments^［80］

Fig. 28. Reconstruction results of colored cartoon plaster targe^［81］

Fig. 29. Principle of multi-aperture imaging^［82］

Fig. 30. Multi-aperture system prototype and imaging results^［94］

Fig. 31. Optical imaging system and imaging effect of AWARW-40^{［102-103］}

Fig. 32. Multi-scale computational optical imaging system and its imaging effect^［105］

Fig. 33. Comparison of traditional design and global design^［107］

Fig. 34. Comparison of restoration image of traditional design and joint design^［107］

Fig. 35. The image quality between the joint design of three lenses and the traditional design of six lens^［116］

Fig. 36. Correction of system chromatic by optical-algorithm design method^［117］

Fig. 37. Infrared reconstruction results of targets in different scenes^［119］

Fig. 38. Convolutional neural network model^［120］

Fig. 39. Comparison of reconstruction results^［120］

Fig. 40. EDSR restoration results^［122］

Fig. 41. Decomposition principle of low-rank and sparse matrix^［123］

Fig. 42. Recovery result of LRSD-TNN^［124］

Fig. 43. Face restoration results of LRSD-TNN^［124］

Fig. 44. Detection results of infrared dim and small targets^［125］

Fig. 45. Reconstruction results of the original image under different concentration^［126］

Fig. 46. The defogging results in different scenes^［127］