Fig. 1. System setup of DOPC. A1−2, neutral-density attenuator; BE, collimated beam expander; BS1−5, beam splitter cube; C1−2, optical fiber collimator; Cam1, scientific complementary metal–oxide semiconductor (sCMOS) camera; Cam2, CMOS camera; FS1−4, fast shutter; I, Isolator; L1−3,5,6, best-form lens; L4, camera lens; Laser, CW laser, λ=532  nm; M1−4, mirror; O, object, a 1951 USAF resolution test chart; P1−3, linear polarizer; S, scattering medium; SLM, phase-only spatial light modulator; SMF, single-mode optical fiber; CB/RB/PB, calibration/reference/playback beam; ProbB, Probe beam. Red dashed line indicates the module of digital phase conjugation mirror (PCM).

Fig. 2. Anatomy and metrics of an edge. (a) A regular unenhanced edge can be divided into three portions, including ground level (G), brink (B), and upper level (U). The lengths of G and U occupy 30 pixels in our experiment. (b) For an enhanced edge, the maximum and minimum pixel intensities of the portion B are termed as summit (S) and valley (V). To quantify the absolute edge enhancement effect, the concept of edge enhancement index EI=(S−V)/(S+V)(μU−μG)/(μU+μG) is introduced, where μU and μG are mean of intensity values of U and G, respectively. (c) The noise level of an edge influences the visual enhancement effect, and thus the concept of edge ENR=S−VσU2+σG2 is defined, where σU and σG are standard deviation of the intensity values of U and G, respectively.

Fig. 3. Intensity profile of the probe beam before and after transmitting through the scattering medium. (a) Intensity profile of the incident probe beam, a quasi-binary pattern of number “0”, shaped by the resolution test chart. Three horizontal white dashed primitive lines (1–3) with the length of 280 pixels are created. The intensity distributions along the lines 1–3 are, respectively, shown in (b)–(d). A and B denote the inner and outer rim of the pattern “0”, respectively. For edge B, the mean EI and ENR are calculated as 0.91 and 42.77, correspondingly. U, upper level; B, brink; G, ground level; S, summit; V, valley. (e) Intensity profile of the probe beam after penetrating a ground glass diffuser, which is a seemingly random speckle pattern with no obvious edge profile that can be found. Scale bar, 500 μm.

Fig. 4. DOPC-based edge enhancement through scattering media. Five images, (a), (e), (i), (m), (q) are recorded by the CMOS camera (Cam2 in Fig. 1) in the playback stage. The intensity ratio (r) between the probe and the reference beams is tuned to different values (0.02, 0.10, 1.0, 10, 50) during the hologram writing. Three 280-pixel horizontal dashed lines (1–3) are created for the figures in the first row. The intensity distributions along lines 1–3 are, respectively, shown in the figures in the second, third, and fourth row, as indicated by the green lines. For example, (b)–(d) are the intensity profiles corresponding to lines 1–3 in (a), while (f)–(h) correspond to the lines in (e). U, upper level; b, brink; G, ground level; S, summit; V, valley. Scale bar, 250 μm.

Fig. 5. Edge enhancement index (EI) and edge enhancement-to-noise ratio (ENR) of edge B for different values of r (0.02, 0.10, 1.0, 10, 50). The x axis represents the common logarithmic scale of the intensity ratio between the probe and reference beams, i.e., lg(r). EI increases from 2.18 to 18.52, and ENR increases from 2.00 to 525.94.

Fig. 6. (a) Schematic diagram illustrating how the intensity ratio between two optical beams affects the resolvability of phase by the interferogram. Different types of vectors represent the electric field of different beams, as presented by the legend. ΔØ1,2, the smallest resolvable interval of phase. (b) Retrieved phase value by the four-step phase-shift method, under the “round-off” effect of the digital camera. r is the intensity ratio between the two optical beams that are interfering, i.e., r=IaIb.