Fig. 1. Basic concept of natural-light full-color motion-picture holography.

Fig. 2. Developed portable natural-light motion-picture holographic camera. (a) Schematic and (b) photograph.

Fig. 3. Image reconstruction procedure using the recorded single hologram.

Fig. 4. Results of experiments with sunlight illumination. Reconstructed images of (a) and (b) shoji and miniature models of (c) a clothes peg (snowman) and (d) another clothes peg (Santa Claus). (e)–(h) Reconstructed images in (a)–(d) with an NLM filter. (i) Recorded hologram. (j) Experimental setup using the developed camera without the black body. Panels (b) and (f) are brightness-improved images of panels (a) and (e).

Fig. 5. Comparison of denoising algorithms for single-shot full-color natural-light holography. (a) Full-color reconstructed image of a clothes peg (Santa Claus) after applying a BM3D filter and its (b) red-, (c) green-, and (d) blue-color components. (e) Red-, (f) green-, and (g) blue-color components of Fig. 4(f).

Fig. 6. Experimental results for an outdoor scene. Reconstructed images focused on (a)–(c) a house, trees, and a net and (d)–(f) a metallic bar. (g) Experimental setup using the developed camera without the black body. (a) and (d) Without numerical filter, (b) and (e) with BM3D, and (c) and (f) with BM4D.

Fig. 7. Comparison of denoising algorithms for natural-light full-color motion-picture holography. (a)–(c) Magnified images of Figs. 6(a)–6(c) whose area is indicated by a red rectangle seen in Fig. 6(a). (d)–(f) Magnified images of Figs. 6(a)–6(c) whose area is indicated by a purple rectangle seen in Fig. 6(a). (g) Plots of lines of (a)–(c) whose place is indicated by a red line seen in (a). (h) Plots of lines of (d)–(f) whose place is indicated by a purple line seen in (d).

Fig. 8. Experimental setup and results for objects illuminated by LED light. (a) Schematic of the experiment. The temporal differences from panels (b) and (c) are 7.8 s [(d) and (e)] and 9.5 s [(f) and (g)], respectively. Reconstructed images focused on the finch [(b), (d), and (f)] and an author (T.T.) [(c), (e), and (g)]. Denoising based on MIRNet was applied to the reconstructed images. Videos 1 and 2 are the reconstructed motion-picture images focused on the finch and human subjects, respectively, and contain panels (b), (d), and (f) and panels (c), (e), and (g), respectively (Video 1, MP4, 1.66 MB [URL: https://doi.org/10.1117/1.APN.4.3.036006.s1]; Video 2, MP4, 1.67 MB [URL: https://doi.org/10.1117/1.APN.4.3.036006.s2]).

Fig. 9. Comparison of denoising algorithms using a recorded hologram of static and moving objects. Reconstructed images without numerical filter [(a) and (b)], with denoising based on MIRNet [(c) and (d)], and with denoising based on MIRNet and then an NLM filter [(e) and (f)]. Reconstructed images focused on the finch [(a), (c), and (e)] and an author (T.T.) [(b), (d), and (f)].