hyperspectral imaging camera using wavefront

5
Hyperspectral Imaging Camera using Wavefront Division Interference (WDI) ERAN BAHALUL, 1 ASAF BRONFELD, 1 SHLOMI EPSHTEIN, 2 YORAM SABAN, 2 AVI KARSENTY, 1 YOEL ARIELI 1,* 1 Jerusalem College of Technology, Havaad Haleumi 21, Jerusalem 9116001, Israel 2 Ben-Gurion University of the Negev, Beer Sheva 8410501, Israel *Corresponding author: [email protected] Received XX Month XXXX; revised XX Month, XXXX; accepted XX Month XXXX; posted XX Month XXXX (Doc. ID XXXXX); published XX Month XXXX A new approach for performing hyperspectral imaging is introduced. The hyperspectral imaging is based on Fourier transform spectroscopy, where the interference is performed by wavefront division interference rather than amplitude division interference. A variable phase delay between two parts of the wavefront emanating from each point of an object is created by a spatial light modulator (SLM) to obtain variable interference patterns. The SLM is placed in the exit pupil of an imaging system, thus enabling conversion of a general imaging optical system into an imaging hyperspectral optical system. The physical basis of the new approach is introduced, and an optical apparatus is built. OCIS codes: (140.3490) Lasers, distributed-feedback; (060.2420) Fibers, polarization-maintaining; (060.3735) Fiber Bragg gratings; (060.2370) Fiber optics sensors. http://dx.doi.org/10.1364/OL.99.099999 A multispectral or hyperspectral imaging system images an object and provides the emanating light spectrum from each pixel of the object. The different wavelengths emitted or reflected from the object characterize the substance of the examined object, and hence, the different parts that compose the object can be identified. Multispectral or hyperspectral imaging provides a 3D data cube of information regarding the object by combining the two-dimensional image of the object with its spectral data. Hyperspectral imaging is used in several domains of applications, such as medical science [1] agriculture [2], defense [3], astronomy and space surveillance [4], detection [5], geology [6], and earth observation [7]. Moreover, due to the continuously growing importance of hyperspectral imaging, complete reference books have already been published [8], as well as comparative international reports [9]. Because hyperspectral imaging provides a 3D data cube of information, despite the imaging sensor being inherently two- dimensional, a scanning operation is usually adopted to construct the 3D spectral image data cube. Many methods have been developed to fulfill this requirement. The early methods include spectral scanning (λ scan) using an optical filter wheel [10], push broom scanning along one spatial axis (y axis) of a line spectral image (x–λ image), and interferometer spectral imaging (such as mechanical scanning Fourier transform (FT) spectrometry [11] and static FT). There are some other methods using dynamic tunable filters, such as liquid crystal tunable filter [12] and acousto-optic tunable filter (AOTF) [13], which do not involve mechanical scanning motion. Those scanning-based approaches have a relatively long 3D image acquisition time, making them only suitable for imaging relatively stable stationary objects. For applications on time-varying objects or moving scenes, approaches for acquiring spectral image data in a single image snapshot have been reported, such as computed-tomography imaging spectrometer (CTIS) [14] and holographic spectral imaging system (HSIS) [15]. In this article, we present a new approach for performing hyperspectral imaging [16]. The hyperspectral imaging is based on Fourier transform spectroscopy, where the interference is performed by wavefront division interference (WDI) rather than amplitude division interference (ADI). This approach enables the conversion of a general imaging optical system into an imaging hyperspectral optical system by introducing a thin SLM in any plane in the imaging system. This addition of the SLM only slightly diminishes the optical performance of the imaging system and thus does not require a special optical design. The basic hyperspectral imaging optical system is shown in Fig. 1. The light from each point of the object is imaged to the image plane to form the image. On its path, the light's wavefront originated from each point is divided by the SLM into two parts, with one part of the wavefront being delayed relative to the other part. The relative phase delay θ between the two parts is given by: 2 π θ λ = Δ (1) where λ is the wavelength and Δ is the optical path difference (OPD) between the two parts of the wavefront. When the two parts of the wavefront intersect to obtain the imaged object’s point, they interfere according to the relative phase delay between them. As the OPD between the two wavefront's parts is increased progressively, each wavelength in the object’s light oscillates between destructive and constructive interference states. The integral intensity of all wavelengths detected by the detector as a function of the OPD performs the interferogram at a certain object's point.

Upload: -

Post on 23-Jan-2017

153 views

Category:

Documents


5 download

TRANSCRIPT

Page 1: Hyperspectral Imaging Camera using Wavefront

Hyperspectral Imaging Camera using Wavefront Division Interference (WDI) ERAN BAHALUL,1 ASAF BRONFELD,1 SHLOMI EPSHTEIN,2 YORAM SABAN,2 AVI KARSENTY,1 YOEL ARIELI 1,* 1Jerusalem College of Technology, Havaad Haleumi 21, Jerusalem 9116001, Israel 2Ben-Gurion University of the Negev, Beer Sheva 8410501, Israel *Corresponding author: [email protected]

Received XX Month XXXX; revised XX Month, XXXX; accepted XX Month XXXX; posted XX Month XXXX (Doc. ID XXXXX); published XX Month XXXX

A new approach for performing hyperspectral imaging is introduced. The hyperspectral imaging is based on Fourier transform spectroscopy, where the interference is performed by wavefront division interference rather than amplitude division interference. A variable phase delay between two parts of the wavefront emanating from each point of an object is created by a spatial light modulator (SLM) to obtain variable interference patterns. The SLM is placed in the exit pupil of an imaging system, thus enabling conversion of a general imaging optical system into an imaging hyperspectral optical system. The physical basis of the new approach is introduced, and an optical apparatus is built.

OCIS codes: (140.3490) Lasers, distributed-feedback; (060.2420) Fibers, polarization-maintaining; (060.3735) Fiber Bragg gratings; (060.2370) Fiber optics sensors.

http://dx.doi.org/10.1364/OL.99.099999

A multispectral or hyperspectral imaging system images an object and provides the emanating light spectrum from each pixel of the object. The different wavelengths emitted or reflected from the object characterize the substance of the examined object, and hence, the different parts that compose the object can be identified. Multispectral or hyperspectral imaging provides a 3D data cube of information regarding the object by combining the two-dimensional image of the object with its spectral data. Hyperspectral imaging is used in several domains of applications, such as medical science [1] agriculture [2], defense [3], astronomy and space surveillance [4], detection [5], geology [6], and earth observation [7]. Moreover, due to the continuously growing importance of hyperspectral imaging, complete reference books have already been published [8], as well as comparative international reports [9]. Because hyperspectral imaging provides a 3D data cube of information, despite the imaging sensor being inherently two-dimensional, a scanning operation is usually adopted to construct the 3D spectral image data cube. Many methods have been developed to fulfill this requirement. The early methods include spectral scanning (λ scan) using an optical filter wheel [10], push broom scanning along one spatial axis (y axis) of a line spectral image (x–λ image), and

interferometer spectral imaging (such as mechanical scanning Fourier transform (FT) spectrometry [11] and static FT). There are some other methods using dynamic tunable filters, such as liquid crystal tunable filter [12] and acousto-optic tunable filter (AOTF) [13], which do not involve mechanical scanning motion. Those scanning-based approaches have a relatively long 3D image acquisition time, making them only suitable for imaging relatively stable stationary objects. For applications on time-varying objects or moving scenes, approaches for acquiring spectral image data in a single image snapshot have been reported, such as computed-tomography imaging spectrometer (CTIS) [14] and holographic spectral imaging system (HSIS) [15]. In this article, we present a new approach for performing hyperspectral imaging [16]. The hyperspectral imaging is based on Fourier transform spectroscopy, where the interference is performed by wavefront division interference (WDI) rather than amplitude division interference (ADI). This approach enables the conversion of a general imaging optical system into an imaging hyperspectral optical system by introducing a thin SLM in any plane in the imaging system. This addition of the SLM only slightly diminishes the optical performance of the imaging system and thus does not require a special optical design. The basic hyperspectral imaging optical system is shown in Fig. 1. The light from each point of the object is imaged to the image plane to form the image. On its path, the light's wavefront originated from each point is divided by the SLM into two parts, with one part of the wavefront being delayed relative to the other part. The relative phase delay θ between the two parts is given by:

2πθλ

= Δ (1)

where λ is the wavelength and Δ is the optical path difference (OPD) between the two parts of the wavefront. When the two parts of the wavefront intersect to obtain the imaged object’s point, they interfere according to the relative phase delay between them. As the OPD between the two wavefront's parts is increased progressively, each wavelength in the object’s light oscillates between destructive and constructive interference states. The integral intensity of all wavelengths detected by the detector as a function of the OPD performs the interferogram at a certain object's point.

Page 2: Hyperspectral Imaging Camera using Wavefront

BeseschOPcoFo

FigTwauncoSLsura

Fia. Wanfunp

=

wpafunfun

ecause the sameparately, each oheme, the camerPDs to obtain thontent of each imaourier transform t

g. 1. The basic hyTo maximize thavefront’s parts niform intensity, omparable. The oLM divided into turrounded by a radiuses are:

ig. 2. An SLM wiIn front. b. From tWhen the circulan unaberrated onction of the opti( )

1

,p r circ

rcircR

λ =

= +

where θ is the pharticular wavelennction of the opnction:

me argument hoobject’s point hara grabs many ine interferogram aged point of theto the interferogr

yperspectral imaghe interference should be compthe area of the twobvious choice oftwo parts, as shoring, where, for 2R =

th two parts havithe side. ar SLM with the twoptical system, ical system is:

(1

2

rc circR

rcirc eR

hase delay betwength and is givenptical system is t

olds for all poias its own intentensity images at each object’s e object is calculatram.

ging optical systecontrast, the eparable. Thereforwo parts of the Sf the SLM configuown in Fig. 2: a ccomparable area12R

ing two comparawo parts is placedthe wavelength

)2

1

i

i

rc e cR

e

θ

θ

+

+

een two parts of n by eq. (1). Thethe Fourier tran

nts of the objerferogram. In thfor many differepixel. The spectrted by applying t

m nergy of the twre, at a plane wiSLM should also uration is a circulcircular part thatas S1 and S2, the (

able areas. d in the exit pupilh-dependent pu

2

rircR

(

f the wavefront ae impulse responnsform of the pu

ect his ent ral the

wo ith be lar t is eir (2)

l of upil 3)

at a nse upil

( )

(

h

R

A

ρ

π

ρ

=

=wher(u,v) arwavelen

andThe cto attaindestrucwhic

in agsystem responswavelen( )2

h

A

ρ

=The inconvoluwavelen(

(({ 2

,

g

I u v

I

A ρ

=

The a(

2

1

2

1

2

1

,I u v

I

I

λ

λ

λ

λ

λ

λ

=

=

(12)

) ((

) ( )

1 121

1

,

2

1i

F p r

J RR

R

e θ

λ

π ρπ ρ

ρ

=

+ −re F denotes a Fore the spatial coongth, and Z is the

condition for the n a maximum cctive and instructich requires that:greement with this proportionalse function of tngth and the OPD)( ) (

2

22Bρ + ntensity in the imution of the intensngth and OPD de) (

( )) (2

, ,

, , *

2

gv I u

u v

B

λ

λ

ρ ρ

=

+ actual image is giv)

( ) {( ) {

( ) {2

, , *

, , *

, , *

g

g

g

I u v A

I u v A

I u vλ

λ

λ

λ

)) ( )

) ( )

1ie

B

θ

λ

ρπ

ρ

+ −

ourier transform. 2uz

ρλ

+=

ordinates of a poie distance to the im( )A Rρ π=

( )B Rρ π=

impulse responscontrast betweenive interference i( ) 2A Bρ =

1 22R R=he above discusl to the squarethe optical systeD: ( ) ( )2Aρ ρ−mage in a certain wsity of the geomeependent PSF:

) ( )

) ( )

, , *

2

v h

A B

λ ρ

ρ ρ−ven by integrating( ) (

( ) (

( )

2 2

2 2

2 2

2

2

2

B

B

A B

ρ ρ

ρ ρ

ρ

+

+

+

(122

2

2J RR

ππ ρ

2v+

int in the image mage plane, (1 12

11

2J RR

Rπ ρ

π ρ

(1 222

2

2J RR

Rπ ρ

π ρse at a geometricn the two extremis: ( )B ρ

2 ssion. The PSF oe of the absolutem and is depen) ( ) (1 cB ρ −wavelength and tetrical image Ig (u,)

( ) (

2

1 coB ρ −g for all waveleng) ( ) (

) ( ) (

( ) ( )

2

2

2

A B

A B

A B

ρ ρ

ρ ρ

ρ ρ

)2R ρρ

(4)

(5)

plane, λ is the )ρ

(6)

)ρ (7)

cal image point me values of a (8)

(9) of the optical te of impulse ndent on the )cosθ

(10)

the OPD is the , v, λ) with the

)}osθ

(11)

gths: ( ) (

( ) }( ) }

1 cos

cos

d

B d

ρ θ

ρ λ

ρ θ

)}d

d

θ λ

λ

Page 3: Hyperspectral Imaging Camera using Wavefront

ussegIwhceof ActhPShabegrop MAeqintintThat wacoprintsimof thspth

Figof wath

This equation issing Fourier spececond term repres( ) {, , *g u v Aλhich represents ertain wavelengthf the interferogramccording to the ee object is convoSF. However, thisas some effect of ecause the PSF ireater than that ptical system. The first step peATLAB simulatioquivalent to the termediate focaltroducing an OPDhe object was deftwo different was executed by onvolving the imrocedure was oterferograms, i.emulated in three f the object, and ie Fourier tranpectrograms of eae system for λ=50

g. 3. The PSFs of Fig. 4 shows thef the three pointsavelength 750 nme border region b

s similar to the ctroscopy, where sents the interfer( )2 22A Bρ + a certain monoch, can be calculatm, similar to the cequation, each caolved with the ws convolution alsothe neighboring includes the termof an aberratio

erformed to asseon of a 4F imagingLCD componenl plane of the oD between thesefined as two partwavelengths (750calculating the mage with the PSobtained repeate., the intensity points: at the imain the border regnsform of theach point. Figure 00 nm, in two dif

optical system foe interferograms s of the object; thm, the part emanbetween them (u

equation of the the first term is rence effect. The ( ) ( )2Aρ ρ−chromatic imageted by the inversconventional FTSalculated monocwavelength and tho states that eachpixels due to thems 2B2(ρ)-2A(ρ)on-free conventioss this approach g optical system. nt was simulateoptical system ine two parts, as dets attached to eac0 nm and 380 nmsystem’s PSF atSF and integratintedly for increas a function age plane, at the mgion between thee interferogram3 presents examfferent OPDs.

or λ=500nm for tw(right) and the she part of the obnating the wavelep to down).

intensity obtaina constant and tterm: ( ) }B ρ (13e of the object atse cosine transforS. chromatic image he OPD-dependeh pixel in the imae spread of the PB(ρ). This effectonal hyperspect

was a ZEMAX anAn optical elemeed by dividing tnto two parts anescribed in figurech other, emanatim). The simulatit each wavelengng the results. Theasing OPDs. Tof the OPD, wmiddle of each pae two parts. Takims provided tmples of the PSFs

wo different OPDspectrograms (lebject emanating tength 380 nm, an

ned the 3) t a rm of ent age PSF t is tral nd ent the nd e 2. ing on gth, his The was art ng the of

Ds. eft) the nd

Fig. 4. (right) fpart emparts (uThe simagingsetup wconfigurinterferoptical ssuch ththe waobtainetwo intepoint prcalculatinterfertwice, o

Fig. 5. SFigs. imagingthe obje

Simulated interfor three points omanating λ=380nup to down).. second step wasg system. Becauswas based on ration of the exrometer was locsystem where thhat each mirror ravefront. While ed at each detectoerferometer’s mirovided the interted by applyingrogram. The objeonce by a He-Ne la

Schematic descrip6 and 7 show tg setup. Fig. 6 shoect illuminated by

rferogram (left) of the object; the nm, and the bordas to build a 4Fse we did not hava Michelson inxperimental stagcated at the intehe mirrors of the returns differentmoving the adjor varied as a funirrors. The intensrferogram. The sg an inverse Fect measured waaser and once by

ption of the experthe results of thows the interferoy the He-Ne laser

and normalizedpart emanating λder between theF setup for the ve access to a suitnterferometer. Thge is illustrated iermediate focal interferometer wt and complemenustable mirror, nction of the OPDsity measured atspectrogram of eaFourier transforas a white screewhite light.

rimental stage. e experimental ogram and the spr.

d spectrogram λ=750nm, the e two object's hyperspectral table LCD, the he schematic in Fig. 5. The plane of the were covered, ntary parts of the intensity D between the t each object's ach point was rm over the en illuminated

hyperspectral pectrogram of

Page 4: Hyperspectral Imaging Camera using Wavefront

FigimAsp

FigimAwabeobanmobyieobbesuacdeRe1.

2.

3. 4.

Sf

5.

6.

g. 6. The measurmage plane, whenAs a complemenpectrogram of the

g. 7. The measurmage plane, whenA new approachavefront divisionetween two partsbject obtains the nd tested in expeethod can be usbject is spectrallyelds uniform resbtaining the speecause the intensurrounding pointccurate, and a moeveloped. eferences D. T. Dicker, J. LernElder, and W. S. El-

C. C. Lelong, environment 66.2,X. Hong, and X. J. WE. K. Hege, D. O’CScience and Technfor Optics and PhoT. Gerhart, J. SunBertozziy, SPIE DeOptics and Photon(January 7, 2004). F. D. van der MeeHecker, W. H. Bak

red interferogramn the object is illumtary analysis, Fige object illuminate

red interferogramn the object is illumh for performingn interference is s of the wavefroninterferograms. rimental setup. Tsed for hyperspey uniform, convolsults; as a resulctrum of an objsity at each point ts, the reconstrucore sophisticated ner, P. Van Belle, S.-Deiry, Cancer biolP. C. Pinet, an

, 179-191 (1998). Wang, Infrared andConnell, W. Johnsonology, SPIE's 48thotonics (2004). nu, L. Lieu, E. Merkefense, Security, anics (2013), Proc. SP

er, H. M. A. van deker, M. F. Noomen

m and spectrograminated by HeNeg. 7 shows the inteed by the white li

m and spectrograminated by a whig hyperspectral introduced. A vant emanating froThe new approaThe results showectral imaging. Hlving a uniform ot, this approachject. Another dishas a contributiocted spectrum atreconstruction a. F. Barth, D. 4th Guogy & therapy 5.8,nd H. Poilvé, R

d Laser Engineeringon, S. Basty, and E. Annual Meeting.

kurjev, J. M. Chanand Sensing. InterPIE 5159, Imaging S

er Werff, F. J. A. vn, M. van der Meij

am of a point in te laser. erferogram and tight.

am of a point in tite light source.. imaging based ariable phase delm each point of ach was simulatw that the proposHowever, when tobject with the Ph has difficulties sadvantage is thon of light from tt each point is nalgorithm should uerry, M. Herlyn, D, 1033-1038 (2006)Remote sensing

g 36.1, 13 (2007).. L. Dereniak, OptiInternational Socie

ng, J. Gilles, and A.rnational Society fSpectrometry IX, 3

van Ruitenbeek, C. de, E. J. M. Carran

the the

the on lay an ted sed the PSF in hat the not be . E. ). of

cal ety

. L. for

380

A. za,

J. B. dObser

7. W. L. H. HuaRemoIntern

8. G. Hananalys

9. NorthSET-06

10. M. GCongr

11. A. R. 12. S. C.

10, pp13. N. Gu14. J. Ha

of Hig15. Y. Lu

and K.16. Y. Ar

(2008)

de Smeth, and Trvation and GeoinfSmith, D. K. Zhouaang, and B. Huang

ote Sensing of national Society forns and P. Geladi, Tsis, John Wiley & Soh Atlantic Treaty Or65-P3 (May 2007).Gunn, D. P. Langsress (EPSC), Vol. 6,

Harvey and D. W. Gebhart, R. C. Tho

p. 1896-1910 (2007upta and V. Volosh

artke, N. Hagan, B. h Speed Electronic

uo, P. J. Gelsinger-A. Kostuk, Opt. Lett.rieli and Y.Weitzm).

T. Woldai, Internaformation 14.1, 112a, F.W. Harrison, Hg, Second Internat

the Atmospherr Optics and PhotoTechniques and apons, Chichester, chrganization (NATO. staff, and D. BarnEPSC-DPS2011-39Fletcher-Holmes,

ompson, and A. M7). hinov, Appl. Opt. 43 A. Kinder, and E.

cs and Systems 18, Austin, J. M. Watso. 33, 18, pp. 2098-2an, United States

ational Journal of 2-128 (2012). H. E. Revercombb, tional Asia Pacific re, Environment, nics (2001).

pplications of hyperh.1, pp.7-8 (2007).

O), RTO Technical R

nes, European Pla7-1 (2011). Opt. Exp. 5368, 12

Mahadevan-Jansen

3, 13, pp. 2752-275L. Dereniak, Intern 505 (2008).

on, G. Barbastathis2100, (2008). Patent Application

Applied Earth

A. M. Larar, A. Symposium on and Space.

rspectral image

Report, RTO-TR-

anetray Science

2, 22 (2004). , Appl. Opt. 46,

59 (2004). national Journal

s,J. K. Barton, R.

n 20080158550

Page 5: Hyperspectral Imaging Camera using Wavefront

References 1. D. T. Dicker, J. Lerner, P. Van Belle, S. F. Barth, D. 4th Guerry, M. Herlyn, D. E.

Elder, and W. S. El-Deiry. “Differentiation of normal skin and melanoma using high resolution hyperspectral imaging." Cancer biology & therapy 5.8, 1033-1038 (2006).

2. C. C. Lelong, P. C. Pinet, and H. Poilvé. "Hyperspectral imaging and stress mapping in agriculture: a case study on wheat in Beauce (France)." Remote sensing of environment 66.2, 179-191 (1998).

3. X. Hong, and X. J. Wang. "Applications of multispectral/hyperspectral imaging technologies in military." Infrared and Laser Engineering 36.1, 13 (2007).

4. E. K. Hege, D. O’Connell, W. Johnson, S. Basty, and E. L. Dereniak. "Hyperspectral imaging for astronomy and space surveillance." Optical Science and Technology, SPIE's 48th Annual Meeting. International Society for Optics and Photonics (2004).

5. T. Gerhart, J. Sunu, L. Lieu, E. Merkurjev, J. M. Chang, J. Gilles, and A. L. Bertozziy,. "Detection and tracking of gas plumes in LWIR hyperspectral video sequence data." SPIE Defense, Security, and Sensing. International Society for Optics and Photonics (2013), Proc. SPIE 5159, Imaging Spectrometry IX, 380 (January 7, 2004).

6. F. D. van der Meer, H. M. A. van der Werff, F. J. A. van Ruitenbeek, C. A. Hecker, W. H. Bakker, M. F. Noomen, M. van der Meijde, E. J. M. Carranza, J. B. de Smeth, and T. Woldai.. "Multi-and hyperspectral geologic remote sensing: A review." International Journal of Applied Earth Observation and Geoinformation 14.1, 112-128 (2012).

7. W. L. Smith, D. K. Zhoua, F.W. Harrison, H. E. Revercombb, A. M. Larar, A. H. Huang, and B. Huang,. "Hyperspectral remote sensing of atmospheric profiles from satellites and aircraft." Second International Asia Pacific Symposium on Remote Sensing of the Atmosphere, Environment, and Space. International Society for Optics and Photonics (2001).

8. G. Hans and P. Geladi, Techniques and applications of hyperspectral image analysis, John Wiley & Sons, Chichester, ch.1, pp.7-8 (2007).

9. North Atlantic Treaty Organization (NATO), “Survey of Hyperspectral and Multispectral Imaging Technologies”, RTO Technical Report, RTO-TR-SET-065-P3 (May 2007).

10. M. Gunn, D. P. Langstaff, and D. Barnes, “Developments towards a filter wheel hyperspectral camera for plantary exploration”, European Planetray Science Congress (EPSC), Vol. 6, EPSC-DPS2011-397-1 (2011).

11. A. R. Harvey and D. W. Fletcher-Holmes, “Birefringent Fourier-transform imaging spectrometer”, Opt. Exp. 5368, 12, 22 (2004).

12. S. C. Gebhart, R. C. Thompson, and A. Mahadevan-Jansen, “Liquid-crystal tunable filter spectral imaging for brain tumor demarcation”, Appl. Opt. 46, 10, pp. 1896-1910 (2007).

13. N. Gupta and V. Voloshinov, “Hyperspectral imager, from ultraviolet to visible, with a KDP acousto-optic tunable filter”, Appl. Opt. 43, 13, pp. 2752-2759 (2004).

14. J. Hartke, N. Hagan, B. A. Kinder, and E. L. Dereniak, “Computed Tomographic Imaging Spectrometer (CTIS) and A Snapshot Hyperspectral Imager and Polarimeter” International Journal of High Speed Electronics and Systems 18, 505 (2008).

15. Y. Luo, P. J. Gelsinger-Austin, J. M. Watson, G. Barbastathis,J. K. Barton, R. and K. Kostuk, “Laser-induced fluorescence imaging of subsurface with a volume holographic spatial-spectral imaging system”, Opt. Lett. 33, 18, pp. 2098-2100, (2008).

16. Y. Arieli and Y.Weitzman, "Spectral Imaging Camera and Applications", United States Patent Application 20080158550 (2008).