Fig. 1. Schematic of light intensity change

Fig. 2. Light path of microscope

Fig. 3. Translation and rotation between captured images

Fig. 4. Registration flow diagram

Fig. 5. Identifying areas to be repaired

Fig. 6. Schematic of Criminisi algorithm

Fig. 7. Flow chart of proposed algorithm

Fig. 8. Phase retrieval images without over-registration restoration. (a) Initial phase; (b) over-focus image; (c) under-focus image; (d) over-focus image after translation and rotation; (e) focus intensity image; (f) phase result of the retrieval image solved directly without any processing

Fig. 9. Phase retrieval images after registration restoration. (a) Under-focus image; (b) over-focus image to be repaired; (c) image repaired by the method in Ref. [6]; (d) image repaired by the proposed method; (e) focusing intensity obtained by the method in Ref. [6]; (f) focusing intensity obtained by the proposed method; (g) phase result retrieved by the method in Ref. [6]; (h) phase result retrieved by the proposed method

Fig. 10. Experimental device of dual-camera dynamic phase imaging system

Fig. 11. Defocus images acquired simultaneously. (a) Under-focus image; (b) over-focus image

Fig. 12. Experimental results of microlens array. (a) Under-focus image to be repaired; (b) image restored by the method in Ref. [6]; (c) image restored by the proposed method; (d) phase result retrieved by the method in Ref. [6]; (e) phase result retrieved by the proposed method

Fig. 13. Experimental results of microlens array

Table 1. Comparison of correlation coefficient and root mean square error under different methods