Fig. 1. Schematic diagram of two-plane representation of four-dimensional light field

Fig. 2. Different types of light field imaging systems. (a) Micro light field camera; (b) handheld light field camera; (c) massive camera array

Fig. 3. Schematic diagram of optical structure of handheld light field imaging system

Fig. 4. Representation of light field images

Fig. 5. Structure diagram of LFCNN^[23]

Fig. 6. Structure diagram of ResLF^[32]

Fig. 7. Structure diagram of LFT^[41]. (a) General flow chart of LFT algorithm; (b) angular Transformer; (c) spatial Transformer; (d) multi-head self-attention mechanism

Fig. 8. Light field image spatial super-resolution dataset for real-world scenarios^[52]. (a) Dataset capturing system; (b) thumbnails of captured dataset

Fig. 9. Structure diagrams of disparity network and color network for light field angular super-resolution^[53]

Fig. 10. Structure diagram of spatial-angular versatile convolution^[36]

Fig. 11. Structure diagram of HDDR^[66]

Fig. 12. Diagram of light field hybrid imaging system. (a) Imaging system based on high- and low-resolution cameras; (b) imaging system based on beam splitter

Fig. 13. Structure diagram of hybrid light field denoising network^[85]

Fig. 14. Structure diagram of view adaptive light field deblurring network^[88]

Fig. 15. Structure diagram of DeOccNet^[89]. (a) DeOccNet network framework diagram; (b) residual atrous spatial pyramid pooling (ResASPP) module

Fig. 16. Comparison of ISTY and existing light field image occlusion removal methods^[94]. (a) Schematic of existing methods; (b) schematic of ISTY

Fig. 17. Structure diagram of light field snow removal network^[99]

Fig. 18. Structure diagram of L3Fnet^[100]

Fig. 19. Schematic diagram of light field image reflection removal dataset^[110]. (a) Background layer; (b) reflection layer; (c) light field image with reflection

Fig. 20. Examples of scenes in benchmark dataset for light field HDR imaging^[115]