Abstract

We explore the possibilities of obtaining compression in video through modified sampling strategies using multichannel imaging systems. The redundancies in video streams are exploited through compressive sampling schemes to achieve low power and low complexity video sensors. The sampling strategies as well as the associated reconstruction algorithms are discussed. These compressive sampling schemes could be implemented in the focal plane readout hardware resulting in drastic reduction in data bandwidth and computational complexity.

© 2010 Optical Society of America

Full Article  |  PDF Article

References

  • View by:
  • |
  • |
  • |

  1. Moving Picture Experts Group, “http://www.mpeg.org..”
  2. D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
    [CrossRef]
  3. N. P. Pitsianis, D. J. Brady, and X. Sun, “Sensor-layer image compression based on the quantized cosine transform,” Proc. SPIE 5817, 250-257 (2005).
    [CrossRef]
  4. N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
    [CrossRef]
  5. D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,432,843(7 October 2008).
  6. D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,174(9 December 2008).
  7. D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,179(9 December 2008).
  8. J. Tanida, T. Kumagai, K. Yamada, S. Miyatake, K. Ishida, T. Marimoto, N. Kondou, D. Miyazaki, and Y. Ichioka, “Thin observation module by bound optics (TOMBO): concept and experimental verifiligcation,” Appl. Opt. 40, 1806-1813(2001).
    [CrossRef]
  9. L. Hong, “Superresolution video reconstruction,” Proc. SPIE 5022, 631-642 (2003).
  10. R. R. Schultz and R. L. Stevenson, “Extraction of high-resolution frames from video sequences,” IEEE Trans. Image Process. 5, 996-1011 (1996).
    [CrossRef] [PubMed]
  11. A. Tekalp, M. Ozkan, and M. Sezan, “High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration,” IEEE Trans. Acoust. Speech Signal Process. 3, 169-172 (1992).
  12. M. Shankar, R. Willett, N. Pitsianis, T. Schulz, R. Gibbons, R. T. Kolste, J. Carriere, C. Chen, D. Prather, and D. Brady, “Thin infrared imaging systems through multichannel sampling,” Appl. Opt. 47, B1-B10 (2008).
    [CrossRef] [PubMed]
  13. A. Portnoy, N. Pitsianis, X. Sun, D. Brady, R. Gibbons, A. Silver, R. Te Kolste, C. Chen, T. Dillon, and D. Prather, “Design and characterization of thin multiple aperture infrared cameras,” Appl. Opt. 48, 2115-2126 (2009).
    [CrossRef] [PubMed]
  14. A. D. Portnoy, N. P. Pitsianis, X. Sun, and D. J. Brady, “Multichannel sampling schemes for optical imaging systems,” Appl. Opt. 47, B76-B85 (2008).
    [CrossRef] [PubMed]
  15. M. Shankar, N. P. Pitsianis, and D. J. Brady, “Spatio-temporal sampling for video,” Proc. SPIE 7076, 707604 (2008).
    [CrossRef]
  16. W. H. Richardson, “Bayesian-based iterative method of image restoration,” J. Opt. Soc. Am. 62, 55-59 (1972).
    [CrossRef]
  17. L. B. Lucy, “An iterative technique for the rectifiligcation of observed distributions,” Astron. J. 79, 745-754 (1974).
    [CrossRef]

2009 (1)

2008 (3)

2006 (2)

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

2005 (1)

N. P. Pitsianis, D. J. Brady, and X. Sun, “Sensor-layer image compression based on the quantized cosine transform,” Proc. SPIE 5817, 250-257 (2005).
[CrossRef]

2003 (1)

L. Hong, “Superresolution video reconstruction,” Proc. SPIE 5022, 631-642 (2003).

2001 (1)

1996 (1)

R. R. Schultz and R. L. Stevenson, “Extraction of high-resolution frames from video sequences,” IEEE Trans. Image Process. 5, 996-1011 (1996).
[CrossRef] [PubMed]

1992 (1)

A. Tekalp, M. Ozkan, and M. Sezan, “High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration,” IEEE Trans. Acoust. Speech Signal Process. 3, 169-172 (1992).

1974 (1)

L. B. Lucy, “An iterative technique for the rectifiligcation of observed distributions,” Astron. J. 79, 745-754 (1974).
[CrossRef]

1972 (1)

Baraniuk, R. G.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Baron, D.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Brady, D.

Brady, D. J.

M. Shankar, N. P. Pitsianis, and D. J. Brady, “Spatio-temporal sampling for video,” Proc. SPIE 7076, 707604 (2008).
[CrossRef]

A. D. Portnoy, N. P. Pitsianis, X. Sun, and D. J. Brady, “Multichannel sampling schemes for optical imaging systems,” Appl. Opt. 47, B76-B85 (2008).
[CrossRef] [PubMed]

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

N. P. Pitsianis, D. J. Brady, and X. Sun, “Sensor-layer image compression based on the quantized cosine transform,” Proc. SPIE 5817, 250-257 (2005).
[CrossRef]

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,432,843(7 October 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,179(9 December 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,174(9 December 2008).

Carriere, J.

Chen, C.

Dillon, T.

Duarte, M. F.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Feldman, M. R.

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

Fiddy, M. A.

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

Gibbons, R.

Hong, L.

L. Hong, “Superresolution video reconstruction,” Proc. SPIE 5022, 631-642 (2003).

Ichioka, Y.

Ishida, K.

Kelly, K. K.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Kolste, R. T.

Kondou, N.

Kumagai, T.

Laska, J. N.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Lucy, L. B.

L. B. Lucy, “An iterative technique for the rectifiligcation of observed distributions,” Astron. J. 79, 745-754 (1974).
[CrossRef]

Marimoto, T.

Miyatake, S.

Miyazaki, D.

Ozkan, M.

A. Tekalp, M. Ozkan, and M. Sezan, “High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration,” IEEE Trans. Acoust. Speech Signal Process. 3, 169-172 (1992).

Pitsianis, N.

Pitsianis, N. P.

M. Shankar, N. P. Pitsianis, and D. J. Brady, “Spatio-temporal sampling for video,” Proc. SPIE 7076, 707604 (2008).
[CrossRef]

A. D. Portnoy, N. P. Pitsianis, X. Sun, and D. J. Brady, “Multichannel sampling schemes for optical imaging systems,” Appl. Opt. 47, B76-B85 (2008).
[CrossRef] [PubMed]

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

N. P. Pitsianis, D. J. Brady, and X. Sun, “Sensor-layer image compression based on the quantized cosine transform,” Proc. SPIE 5817, 250-257 (2005).
[CrossRef]

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,432,843(7 October 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,179(9 December 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,174(9 December 2008).

Portnoy, A.

A. Portnoy, N. Pitsianis, X. Sun, D. Brady, R. Gibbons, A. Silver, R. Te Kolste, C. Chen, T. Dillon, and D. Prather, “Design and characterization of thin multiple aperture infrared cameras,” Appl. Opt. 48, 2115-2126 (2009).
[CrossRef] [PubMed]

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

Portnoy, A. D.

Potuluri, P.

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,174(9 December 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,179(9 December 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,432,843(7 October 2008).

Prather, D.

Richardson, W. H.

Sarvotham, S.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Schultz, R. R.

R. R. Schultz and R. L. Stevenson, “Extraction of high-resolution frames from video sequences,” IEEE Trans. Image Process. 5, 996-1011 (1996).
[CrossRef] [PubMed]

Schulz, T.

Sezan, M.

A. Tekalp, M. Ozkan, and M. Sezan, “High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration,” IEEE Trans. Acoust. Speech Signal Process. 3, 169-172 (1992).

Shankar, M.

Silver, A.

Stevenson, R. L.

R. R. Schultz and R. L. Stevenson, “Extraction of high-resolution frames from video sequences,” IEEE Trans. Image Process. 5, 996-1011 (1996).
[CrossRef] [PubMed]

Suleski, T.

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

Sun, X.

A. Portnoy, N. Pitsianis, X. Sun, D. Brady, R. Gibbons, A. Silver, R. Te Kolste, C. Chen, T. Dillon, and D. Prather, “Design and characterization of thin multiple aperture infrared cameras,” Appl. Opt. 48, 2115-2126 (2009).
[CrossRef] [PubMed]

A. D. Portnoy, N. P. Pitsianis, X. Sun, and D. J. Brady, “Multichannel sampling schemes for optical imaging systems,” Appl. Opt. 47, B76-B85 (2008).
[CrossRef] [PubMed]

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

N. P. Pitsianis, D. J. Brady, and X. Sun, “Sensor-layer image compression based on the quantized cosine transform,” Proc. SPIE 5817, 250-257 (2005).
[CrossRef]

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,432,843(7 October 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,179(9 December 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,174(9 December 2008).

Takhar, D.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Tanida, J.

Te Kolste, R.

Tekalp, A.

A. Tekalp, M. Ozkan, and M. Sezan, “High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration,” IEEE Trans. Acoust. Speech Signal Process. 3, 169-172 (1992).

TeKolste, R. D.

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

Wakin, M. B.

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

Willett, R.

Yamada, K.

Appl. Opt. (4)

Astron. J. (1)

L. B. Lucy, “An iterative technique for the rectifiligcation of observed distributions,” Astron. J. 79, 745-754 (1974).
[CrossRef]

IEEE Trans. Acoust. Speech Signal Process. (1)

A. Tekalp, M. Ozkan, and M. Sezan, “High-resolution image reconstruction from lower-resolution image sequences and space-varying image restoration,” IEEE Trans. Acoust. Speech Signal Process. 3, 169-172 (1992).

IEEE Trans. Image Process. (1)

R. R. Schultz and R. L. Stevenson, “Extraction of high-resolution frames from video sequences,” IEEE Trans. Image Process. 5, 996-1011 (1996).
[CrossRef] [PubMed]

J. Opt. Soc. Am. (1)

Proc. SPIE (4)

M. Shankar, N. P. Pitsianis, and D. J. Brady, “Spatio-temporal sampling for video,” Proc. SPIE 7076, 707604 (2008).
[CrossRef]

D. Takhar, J. N. Laska, M. B. Wakin, M. F. Duarte, D. Baron, S. Sarvotham, K. K. Kelly, and R. G. Baraniuk, “A new camera architecture based on optical-domain compression,” Proc. SPIE 6065, 606509 (2006).
[CrossRef]

N. P. Pitsianis, D. J. Brady, and X. Sun, “Sensor-layer image compression based on the quantized cosine transform,” Proc. SPIE 5817, 250-257 (2005).
[CrossRef]

N. P. Pitsianis, D. J. Brady, A. Portnoy, X. Sun, T. Suleski, M. A. Fiddy, M. R. Feldman, and R. D. TeKolste, “Compressive imaging sensors,” Proc. SPIE 6232, 62320A (2006).
[CrossRef]

Other (5)

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,432,843(7 October 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,174(9 December 2008).

D. J. Brady, N. P. Pitsianis, X. Sun, and P. Potuluri, “Compressive sampling and signal inference,” U.S. patent 7,463,179(9 December 2008).

L. Hong, “Superresolution video reconstruction,” Proc. SPIE 5022, 631-642 (2003).

Moving Picture Experts Group, “http://www.mpeg.org..”

Supplementary Material (4)

» Media 1: AVI (4153 KB)     
» Media 2: AVI (3829 KB)     
» Media 3: AVI (3425 KB)     
» Media 4: AVI (3953 KB)     

Cited By

OSA participates in CrossRef's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (19)

Fig. 1
Fig. 1

Video frames from (a) a multichannel camera in which each frame consists of multiple nonredundant low-resolution representations of the scene and (b) a conventional camera.

Fig. 2
Fig. 2

Spatial-compressive measurements using a multichannel camera. A small subset of subapertures was selected in each time frame, thereby decreasing the data load.

Fig. 3
Fig. 3

Multichannel infrared camera (left) used to demonstrate compressive measurements for video. It contains a 3 × 3 lenslet array to create nine nonredundant low-resolution images of the scene. Video from a conventional single-aperture infrared camera (right) is used to compare the quality of the video reconstructions.

Fig. 4
Fig. 4

(Media 1) Comparison of corresponding frames from the least-gradient video reconstruction at a certain time instant when considering (a) one subaperture image, (b) four subaperture images, (c) nine subaperture images for reconstruction, and (d) the corresponding frame obtained with a conventional infrared camera.

Fig. 5
Fig. 5

Zoom into regions of the least-gradient reconstructed images shown in Fig. 4 when considering (a) one subaperture image, (b) four subaperture images, (c) nine subaperture images for reconstruction, and (d) the conventional camera image.

Fig. 6
Fig. 6

Plots across the cross section of a row in the image from the least-gradient reconstructed images shown in Fig. 4. The circles indicate regions in which the image reconstructed by using fewer subapertures is unable to resolve details that are resolved when more subapertures are considered.

Fig. 7
Fig. 7

(Media 2) Comparison of corresponding frames from the RL video reconstruction at a certain time instant when considering (a) one subaperture image, (b) four subaperture images, (c) nine subaperture images for reconstruction, and (d) the corresponding frame obtained with a conventional infrared camera.

Fig. 8
Fig. 8

Zoom into regions of the RL reconstructed images shown in Fig. 7 when considering (a) one subaperture image, (b) four subaperture images, (c) nine subaperture images for reconstruction, and (d) a conventional camera image.

Fig. 9
Fig. 9

Plots across the cross section of a row in the image from the RL reconstructed images shown in Fig. 7. The circles indicate regions in which the image reconstructed by using fewer subapertures is unable to resolve details that are resolved when more subapertures are considered.

Fig. 10
Fig. 10

Comparison of corresponding frames from the least- gradient video reconstruction at a certain time instant using a bar target when considering (a) one subaperture image, (b) four subaperture images, (c) nine subaperture images for reconstruction, and (d) a conventional camera image.

Fig. 11
Fig. 11

Plots across the cross section of a row in the least-gradient reconstructed bar-target images shown in Fig. 10.

Fig. 12
Fig. 12

Comparison of corresponding frames from the RL video reconstruction at a certain time instant using a bar target when considering (a) one subaperture image, (b) four subaperture images, (c) nine subaperture images for reconstruction and (d) the conventional camera image.

Fig. 13
Fig. 13

Plots across the cross section of a row in the RL reconstructed bar-target images shown in Fig. 12.

Fig. 14
Fig. 14

Timing diagram illustrating the spatiotemporal sampling scheme using a multichannel camera. The circles in (a) and their associated colors indicate different subapertures. In (b), the vertical lines indicate sampling instants for conventional video and the color strips represent the sampling instants for the various subapertures with the modified sampling scheme.

Fig. 15
Fig. 15

Description of the sampling strategy involving spatiotemporal compression, shown for a 2 × 2 multichannel system. (a) The original video stream; (b) compression performed by considering different sums of subapertures [ S 1 q , S 2 q , S 3 q , and S 4 q as defined by Eqs. (4, 5, 6, 7)] at different time instants; (c) the four sums are shown for different time instants; these are used to estimate the high- resolution frames.

Fig. 16
Fig. 16

Visible-band multiaperture camera used to demonstrate spatiotemporal compressive sampling for video. The 4 × 4 lenslet array replaces the conventional single-aperture optics.

Fig. 17
Fig. 17

Images (a) obtained with the visible 4 × 4 multichannel camera and (b)  reconstructed from the 16 low-resolution images in (a).

Fig. 18
Fig. 18

(Media 3) A temporally aliased frame at a certain time instant in the video (averaged from four previous time frames) obtained from (a) subaperture 1 and (b) subaperture 2.

Fig. 19
Fig. 19

(Media 4) (a) Reconstructed frame from the spatiotemporal compressed video and (b)  frame from the multichannel camera with no temporal aliasing.

Equations (12)

Equations on this page are rendered with MathJax. Learn more.

x k ( t ) = H k f ( t ) + n k , k = 1 , 2 , , M ,
x ( t ) = H f ( t ) + n .
x k ( t ) = H k f ( t ) + n k k = 1 , 2 , , P ,
f ( t ) LG = arg min f ( t ) γ ( f ( t ) ) = f ( t ) 2 s . t . H f ( t ) = x ( t )
f LG ( t ) = arg min c ( N c f p ( t ) 2 .
f LG ( t ) = f p ( t ) N ( N T T N ) 1 ( N ) T f p ( t ) ,
f ^ k + 1 ( t ) = f ^ k ( t ) H T ( x ( t ) H ( f ^ k ( t ) ) ) ,
S 1 q = p = q 6 q 3 x 1 p ,
S 2 q = p = q 5 q 2 x 2 p ,
S 3 q = p = q 4 q 1 x 3 p ,
S 4 q = p = q 3 q x 4 p ,
S = T H F ,

Metrics