Fig. 1. Flowchart of phase contrast OCT range expansion based on data-driven. (a) Network training; (b) oversampling phase prediction

Fig. 2. UNet++ network architecture

Fig. 3. Partial phase data in the training set. (a) Oversampling phase data; (b)‒(d) aliasing phase data at different undersampling rates

Fig. 4. Phase-contrast OCT system. (a) System schematic diagram; (b) system physical diagram

Fig. 5. Loading experimental. (a) Mechanical loading experiment; (b) thermal loading experiment

Fig. 6. Experimental results of loading experiment. (a) Phases at different undersampling rate; (b) results of vector method; (c) results of CNN prediction; (d) vector method calculation results after CNN prediction

Fig. 7. MSE varying with undersampling rate. (a) Phase; (b) strain

Fig. 8. Experimental results of curing experiment. (a) Phases at different undersampling rate; (b) results of vector method; (c) results of CNN prediction; (d) vector method calculation results after CNN prediction

Fig. 9. MSE varying with undersampling rate. (a) Phase; (b) strain

Fig. 10. Experimental results of continuous samples. (a) Phases at different undersampling rate; (b) results of CNN prediction

Fig. 11. Experimental results of signal to noise ratio. (a) Undersampling phases with different noise levels; (b) results of CNN prediction