Fig. 1. Analytical model of reflected light composition on surface of object^[34]

Fig. 2. s and p components of reflected and refracted light formed by unpolarized light incident from air to medium surface and diffuse polarized light formed by internal scattered light transmitted back into air by particles

Fig. 3. Variation of light intensity of detector with rotation angle of polarizer^[32]

Fig. 4. Polarization coordinate representation of the normal of object surface microelement^[37]

Fig. 5. Variation of polarization degree of diffuse light with zenith angle^[38]

Fig. 6. Experimental setup for azimuth deblurring based on photometric stereo vision technology^[40]

Fig. 7. Location of sphere object and three light sources observed from camera’s perspective^[40]

Fig. 8. Intensity images of real objects and surface reconstruction results of objects based on photometric stereo vision technology^[40]. (a) Intensity images of real objects; (b) surface reconstruction results of objects

Fig. 9. Reconstruction results combined with photometric stereo vision^[42]

Fig. 10. Reconstructed shapes from polarimetric multi-view stereo vision^[43]

Fig. 11. Reconstructed results from polarization and shading recovery method^[44]. (a) Polarization phase angles; (b) degree of diffuse polarization; (c) intensity; (d) reconstructed surfaces

Fig. 12. Height recovery results of objects of different material types for known light source^[45]

Fig. 13. Height recovery results of objects of different material types for outdoor environment^[45]

Fig. 14. 3D reconstructed results from polarization + RGB camera stereo vision^[48]. (a) Intensity images of real objects; (b) estimated albedo maps; (c) estimated depth maps

Fig. 15. Experimental setup of method combined with coarse depth map captured by Kinect^[49]

Fig. 16. 3D imaging results of coarse depth maps captured by Kinect^[49]

Fig. 17. 3D reconstruction maps of weak-texture objects using Astra 3D camera^[52]

Fig. 18. NIR 3D polarization imaging model^[53]. (a) NIR 3D polarization imaging system; (b) 3D model

Fig. 19. 3D results for target before correcting reflectance and after correcting reflectance^[53]. (a) Input intensity information; (b) intensity image after correcting reflectance; (c) (d) 3D shapes of Figs. 19(a) and 19(b); (e) (f) relative height values; (g) contours of 350th column shown in Figs. 19(e) and 19(f)

Fig. 20. Three-dimensional results for colored cartoon plaster target^[53]. (a) 3D-recovered result without correcting for reflectance; (b) 3D-recovered result using 3D NIR polarization method; (c) (d) relative height values of Figs. 20(a) and 20(b); (a1) (b1) approximately 10× magnification of 3D shape in region of target arm; (e) height variations in pixels of Figs. 20(a1) and 20(b1)