Fig. 1. Diagram of the code-pair encoding method.

Fig. 2. Four standard pairs are simplified to one standard group when we use four-level phase 0, π/2, π, and 3π/2, and set the phase shift between the upper pixel and the bottom pixel as π/2.

Fig. 3. Interference between the phase-readout reference beam and the reconstructed beam. Here, the reconstructed beam is not loaded by information. (a),(b) Interference results at different times.

Fig. 4. Intensity values of two pixels in one pair with different phase differences between the phase-readout reference beam and the reconstructed beam. The horizontal axis is the phase, and the vertical axis is the intensity value. The red dot denotes the upper pixel, and the green dot denotes the bottom pixel.

Fig. 5. Phase code in one unit and intensity values in one unit after interferometry.

Fig. 6. Four-level phase pattern according to the pair code rule. Four pixels in the red frame are the standard pixels.

Fig. 7. BER curves of phase retrieval results with and without the pair code rule according to different phase differences between the phase-readout reference beam and the reconstructed beam.

Fig. 8. Diagram of the optical setup. SF, spatial filter; HWP, half-wave plate; BS, beam splitter; L1–L6, lens; SLM, spatial light modulator. The media is an Irgacure 784-doped PMMA photopolymer with a thickness of 1.5 mm. (L1=300  mm, L1–L5=150  mm, L6=200  mm, pixel pitch of the SLM is 8 μm, the pixel pitch of the CCD is 3.75 μm)

Fig. 9. Interference result in the CCD.

Fig. 10. Phase retrieval results in the experiment. (a) Original phase pattern. (b) Phase retrieval pattern without the pair code rule. (c) Phase error distribution corresponding to (b). (d) Phase retrieval pattern with the pair code rule. (e) Phase error distribution corresponding to (d).