Summary 3D cinemas are becoming real thanks to digital image processing technology. The most feasible and stable technology based on the binocular dis

Size: px

Start display at page:

Download "Summary 3D cinemas are becoming real thanks to digital image processing technology. The most feasible and stable technology based on the binocular dis"

なおみいまいだ
4 years ago
Views:

1 Summary 3D cinemas are becoming real thanks to digital image processing technology. The most feasible and stable technology based on the binocular disparity requires stereoglasses. However, there is still a strong desire for glasses-free method in order to ease the stress to viewers. In this paper, Space-Sampling method is investigated for glasses-free 3D movies. With a layered image sensor device, Space-Sampling method captures 3D image projected through a lens. Layered display panels reproduce 3D images. They can be projected on a refractive screen to reconstruct the original 3D image space. Transparent image sensors, layered display panels, depth resolution, blur elimination and occlusion problems are discussed. Key words: 3D-image, space-sampling, transparent imaging-device, refractive screen, blur elimination, 1. 3DCG JPEG-2000 MPEG-4 AVC A Study of Space-Sampling Method for 3D Movies by Takanori SENOH (Member), Terumasa AOKI (Member) (Research Center for Advanced Science & Technology, The University of Tokyo), Hiroshi YASUDA (Member), Takuyo KOGURE (Center for Collaborative Research, The University of Tokyo). 3DCG 1) 2) 317

2 ) 4) 100 5) ISO MPEG MVC 6) 7) 8) 10) 8) 1 9), 10) Fig. 1 1 Mapping from real space to image space D f d d [,D] [f,d] d = Df (1) D f f f =50mm D =50cm f =50mm d = mm f/(d f) D f d f d f D d D 318

3 2 Fig. 2 Observation of space-sampled images through a lens Fig. 3 3 Projection method of space-sampled images [ d,f ] [D, ] d = Df D + f (2) f d (3) D (D + f)/f (D f)/(d + f) =f/(2d f) d = df (3) 2d f d = mm 50 mm cm D (D + f)/f D 11) 3 1 f d 4 12) n =2.0 r f = n n 1 r (4) 2 4 Fig. 4 Refraction screen (a) (b) 5 Fig. 5 Refraction screen material f = r 1 n =2 13) n =1.5 5 (a) 5(b) CCD CMOS 6 CMOS 14) 319

4 n n p P MOS ON MOS CMOS ZnO ITO SnO 2 GaN 15),16) 17) 3 RGB3 3.2 LCD EL 7 LCD LCD CMOS LCD RGB3 RGB3 LCD 6 CMOS Fig. 6 Transparent CMOS sensor 100% LCD LCD LCD LCD 18),19) LE EL CMOS RGB L f F = f/l 8 D a d b y δ δ = L (y f)d yf = (5) d FD f =50mm L =14.3mm y mm pel 1.5 pel 1m 7 20) 2m 0.33m ),19) 7 Fig. 7 Structure of LCD 8 Fig. 8 Blur of images 320

5 9 Fig. 9 Blur curve Fig Laplacian filter (b) y = mm (b) y =50.425mm (c) y =50.85mm (d) y =52.125mm (c) y =50.85 mm (d) y =52.125mm 10 Fig. 10 Space-sampled images mm m Laplacian 12 Fig. 12 Laplacian output 11 7x7 Laplacian Median Morphology LPF LPF

6 (b) y =50.425mm (b) y = mm (c) y =50.85 mm (d) y =52.125mm (c) y =50.85mm (d) y = mm 13 Fig. 13 Peak detection results 15 Fig. 15 Smoothing with LPF Fig tap LPF 18) z = exp(x/α β) (6) x 8 α β α 4.3 (5) y δ d (7) 16 δ = L d = L y d d (7) 16 Fig. 16 Blur vs. image location (6) α β (6) α β 17 α =16 β =8.0 z 18 3D CG 322

17 Brightness offset (b) β =8.0 (c) β =6.

125mm LPF 42 19 (b) (c) (6) α =16 β =8 128

7 (b) y =50.425mm (a) LPF (c) y =50.85 mm (d) y =52.125mm 17 Fig. 17 Brightness offset (b) β =8.0 (c) β = Fig. 19 Whitening threshold (b) y =50.425mm (a) (c) y =50.85 mm (d) y =52.125mm 18 Fig. 18 Blur-eliminated images 21) (6) β 19 (a) y =50.0mm y =52.125mm LPF (b) (c) (6) α =16 β = (b) α =16 β = (c) (b) 20 Fig. 20 Occlusion (a) A B a b D 1 D 2 a/b d 1 d 2 a b = d 1 = D 1 (D 2 f) (8) d 2 D 2 (D 1 f) 323

35 4 2006 1 f =50mm D 1 =6m D 2 =3m a/b =0.992 a b 1% 20 (a) 1% 20 (b) b gap (D 1 D 2 ) L gap = f (D ( 1 D 2 ) Bf D 2 (D 1 f) D 2 f L ) (9) 2 f =50m L =14.3mm D 1 = D 2 = 1m B = 350 mm 1/2 gap = 0.

8 f =50mm D 1 =6m D 2 =3m a/b =0.992 a b 1% 20 (a) 1% 20 (b) b gap (D 1 D 2 ) L gap = f (D ( 1 D 2 ) Bf D 2 (D 1 f) D 2 f L ) (9) 2 f =50m L =14.3mm D 1 = D 2 = 1m B = 350 mm 1/2 gap = 0.56 mm 1.6% gap L a/b TV 1) 3D Vol.19, No.4, pp.29-35, Dec. (2005). 2) MVE99-75, Feb. (2002). 3) 16 / 16-8 Mar. (2005). 4) D Vol.33, No.6, pp , Nov. (2004). 5) 2 3 Vo.57, No.2, pp , Feb. (2003). 6) Video & Test group: Call for Proposal on Multiview Video Coding, ISO/IEC JTC1/SC29/WG11, N7327, Jul. (2005). 7) Alan C. Traub: Three-Dimensional Display, US Pat , Feb. (1970). 8) 3 98 pp (1998). 9) B. Hendriks and S. Kuiper: Through A Lens Sharply, IEEE Spectrum, pp.20 24, Dec. (2004). 10) B. Berge:, , pp (2005). 11) pp , Mar. (2006). 12) Feb. (1999). 13) Quality, ixyl. html Japan (2005). 14) Jun. (2005). 15) Nature (2004). 16) T. Minami: Transparent conducting oxide semiconductors for transparent electrodes, 2005 Semicond. Sci. Tech. 20, S35-S44 (2005). 17) pp.36 37, Jan. (2006). 18) H. Takada S. Suyama, and K. Nakazawa: A new 3D display method using visual illusion produced by overlapping two luminance division displays, IEICE Trans on Electron E88-C (3), p.445 (2005). 19) A. Sullivan: 3 Deep, IEEE Spectrum, pp.22 27, Apr. (2005). 20) Vol.31, No.8, pp (1977). 21) T. Senoh, T. Aoki, H. Yasuda, and T. Kogure: A Study of Space-Sampling Method for 3D Movies, IEEE WS on 3DCINE 06, June (2006) MIT 2005 AV 324

& Vol.5 No (Oct. 2015) TV 1,2,a) , Augmented TV TV AR Augmented Reality 3DCG TV Estimation of TV Screen Position and Ro

& Vol.5 No (Oct. 2015) TV 1,2,a) , Augmented TV TV AR Augmented Reality 3DCG TV Estimation of TV Screen Position and Ro TV 1,2,a) 1 2 2015 1 26, 2015 5 21 Augmented TV TV AR Augmented Reality 3DCG TV Estimation of TV Screen Position and Rotation Using Mobile Device Hiroyuki Kawakita 1,2,a) Toshio Nakagawa 1 Makoto Sato