SCANNING HYPERSPECTRAL CAMERA AND METHOD FIELD OF THE INVENTION The present invention relates to a scanning spectral imaging system. BACKGROUND OF THE INVENTION The spectral behavior of light reflected from substrates has been long used for characterizing the substrate’s characteristics in scientific, chemical, industrial and forensic applications. A spectral imaging camera and imaging spectrometers must be utilized when the polychromatic light in the 2-D field of view is measured simultaneously. There are numerous optical designs for realizing a spectral imaging camera or imaging spectrometers, such as Fourier Transform Spectrometers, dispersive spectrometers and others. Fig. 1a shows schematically part of a conventional spectral imaging camera 100 based on a polarization interferometer 120 comprising two birefringent wedges 130 and 140 positioned between a polarizer 150 and an analyzer 160. One of the wedges is movable and the other may be static or movable. The orientations of the optical axes of the birefringent materials of which the two wedges are made are perpendicular to each other. The axes of the polarizer and the analyzer may be mutually parallel or orthogonal. The direction of the polarizer is preferably 45° to the respective optical axes of the two wedges. When a light ray 170170 propagates through the polarization interferometer 120, after propagating through the polarizer 150, the light is polarized and its polarization is oriented 45° relative to respective mutually orthogonal birefringent optical axes of the two wedges 130 and 140. Throughout the description we will refer to the terms “input” and “output” of the interferometer such that light enters through the input and exits from the output. Likewise, optical components inserted between the light source and the input will he referred to as “upstream” of the interferometer while optical components inserted between the output and tire analyzer will he referred to as “downstream” of the interferometer. The polarized light is divided into two orthogonal polarizations that propagate through the two wedges and acquire a phase difference Df whose value is a function of the location in which it propagates through the two wedges. After passing through the two wedges, the light propagates through the analyzer 380 and the two polarizations interfere with constructive or destructive interference according to the phase difference that they acquired. Approximately, at a certain point of propagation, the absolute phase difference Df between the two orthogonal polarizations depends on the birefringent refractive indices and the thicknesses of the two wedges at that point: where l is the wavelength. n _e1, n _e2 are the extraordinary refractive indices of the wedges. no1 _, ⁿ o2 are the ordinary refractive indices of the wedges. d ₁ , d ₂ are tire thicknesses of tire wedges at a certain point.

Since the phase difference depends on the wavelength, each component of tire light characterized by a different wavelength interferes constructively or destructively according to the respective phase difference between the two polarizations of that component. The intensities of the interference of all wavelengths at each point may be directed to a certain detector element 180 in a detector array 190 which integrates all intensities to obtain the integrated intensity at that location. When the phase shift varies linearly the value of the integrated intensity as a function of the phase shift represents an interferogram. The phase shift may be varied linearly by moving at least one of the wedges at constant velocity relative to the other thus changing the ratio of their thicknesses d ₁ and d ₂ where intersected by the light ray as shown in Fig. 1b. Te spectrum of the light at each point is obtained by Fourier transforming; the interferogram obtained at that point. US20160178503A1 describes an optical device that includes a first polarizer arranged to receive light emanating from an object moving along a trajectory. The first polarizer polarizes the light emanating from the object along a first polarization direction. A waveplate having an optical retardance that varies as a function of position along the trajectory receives light from the first polarizer. The slow axis of the waveplate is at a first angle with respect to the first polarization direction. A second polarizer is arranged to receive light from the waveplate. The second polarizer polarizes light along a second polarization direction. At least one detector receives light from the second polarizer and provides an electrical output signal that varies with time according to intensity of the light received from the second polarizer. in OPTICS EXPRESS, 25, 15, p. 17402 a hyperspeetral imaging; system based on a static polarization interferometer is described in which the relative movement between a scene and the imaging system is exploited. Accordingly, the relative motion encodes the spectral information in the detector's temporal signal via a ‘ push-broom” approach. The polarization interferometer introduced in the previous section is used as the interferometer within the static Fourier-transform imaging spectrometer by placing it in the conjugate focal plane of an imaging array. Each point’s trajectory across a focal plane array is tracked to record and extract the signal from which the hyperspeetral data- cube of the full scene is calculated.

Reference is made to ‘OPTICAL FABRICATION : Thin films provide wide- angle correction for waveplate components ¹ ” by Kim Tan, Karen Hendrix and Paul McKenzie appearing in LaserFocusWorld, March 1, 2.007. '.This article discloses an imaging system, wherein as the light rays of each point of the scene are focused by the imaging; lens towards the detectors, they propagate at different angles and thus enter the static polarization interferometer at different angles. According to this article, there is a change in retardance with azimuthal angle (plane of incidence relative to optical axis) and polar angle of incidence. Consequently, for a non-collimated beam of light, retardance will change as a function of the illumination-cone ray angle, quantified as an f/number (beams with smaller f/numbers will show greater variation across the cone). Extreme rays tend to be iess or more retarded than the rays at normal incidence, depending upon azimuthal angle.

Downloaded from https://www.laserfocusworid.com/optics/article/16553005/opti cal-fabrication-thin- films-provide-wideangle-correction-for-waveplate-components As a result, this effect tends to decrease the spectral resolution of the hyper- spectrai imaging system especially in imaging systems with small f/numhers i.e. whose cone of rays subtend large angles. Moreover, since the spectrometer is required to he located before the detector and there is a cone of rays that converges, the diameter at which the rays enter the spectrometer is significantly larger than the sample pixel size, which will reduce the spectral resolution. Also, since it is close to the focal plane, extremely high quality components and level requirements ate required.

US Patent No 10,302,494 (Alex Hegyi et al.) discloses a system for obtaining spectral information from a moving object. In some embodiments, light that has been polarized along a first polarization direction is received by a waveplate sandwiched between a first polarizer and a second polarizer. The waveplate has an optical retardance that varies as a function of position along the trajectory direction of the object. In some configurations, the waveplate may he a Wollaston prism or other optical retardance device.

It is therefore an object of the present invention to provide new methods to increase the spectral resolution of the hyperspectral imaging system.

SUMMARY OF THE INVENTION

In accordance with the present invention there is provided an inexpensive high resolution spectral imaging camera or a spectral imager for accurate spectral analysis of an image.

The present invention provides a spectral imaging camera having increased spectral resolution, while employing optical components of reduced tolerance, thus improving production capacity and reducing costs.

In accordance with a broad aspect of the invention there is provided a spectral imaging optical system and method having the features of the respective independent claims.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to understand the invention and to see how it may be carried out in practice embodiments will now be described, by way of non-limiting example only with reference to the accompanying drawings, in which: Fig, la shows schematically a detail of a spectral imaging system that includes a conventional polarization interferometer;

Fig, lb shows a detail of birefringent wedges used in the polarization interferometer of Fig. la;

Fig. 2 shows schematically a detail of a spectral imaging optical system based on a conventional static polarization interferometer;

Fig, 3 shows schematically a detail of a static polarization interferometer according to a first embodiment of the invention;

Fig, 4 shows schematically a detail of a static polarization interferometer according to a second embodiment of the invention;

Figs, 5a and 5b show _' elevations of wedge-shaped bundle of optical fibers;

Fig. 6 shows schematically a detail of a static polarization interferometer using the wedge-shaped optical fiber bundle shown in Figs. 5a and 5b;

Fig, 7 show's schematically a detail of a hyper -spectral imaging system according to another embodiment of the invention;

Fig, 8 show's schematically a non-polarized ray propagating through a thick Wollaston prism;

Fig, 9 show's schematically a detail of a spectral imaging optical system based on a static polarization interferometer according to a farther embodiment of the invention;

Figs. 10a and 10b show schematically cross-sections of the index ellipsoids of positive and negative birefringent materials;

Fig, 11 show's schematically a detail of a spectral imaging optical system based on a static polarization interferometer according to an embodiment of the invention; and

Fig, 12 show/s schematically a detail of a spectral imaging optical system based on a static polarization interferometer according to another embodiment of the invention.

DETAILED DESCRIPTION OF EMBODIMENTS in the following description of some embodiments, identical components that appear in more than one figure or that share similar functionality will be referenced by identical reference symbols. in many cases there is a relative motion between an object and a measuring optical system measuring the object. This movement can be exploited to simplify the system and to reduce its cost. In OPTICS EXPRESS, 25, 15, p. 17402 a hyperspectral imaging system based on a static polarization interferometer is described in which the relative movement between a scene and an imaging system is exploited (see https://doi.org/10.1364/QE.25.017402).

Fig. 2 shows schematically a detail of a similar spectral imaging optical system 200, wherein a static polarization interferometer 120 is located at a plane m which an intermediate real image is obtained or directed on the detector may. The static polarization interferometer 120 is similar to the polarization interferometer shown in Fig. 1, comprising two birefringent wedges, which are located between a polarizer and an analyzer. The orientations of the optical axes of the birefringent materials of which the two wedges are made are perpendicular to each other. After propagating; through the first polarizer and the two wedges, the two orthogonal polarizations acquire a phase difference Af whose value is a function of the location in winch they propagate through the two wedges.

When there is relative motion between the object and the measuring optical system this movement is exploited to simplify the system by obviating the need to move one of the wedges relative to other. Since the image of tire scene moves from line to line successively on the detector array of the imaging system in a direction opposite to the direction of the relative motion the phase shift between the two polarizations of light varies linearly from line to line at the successive lines of the detector array in all embodiments, relative motion can be the result of three different events: the object moves upward and the detector array remains stationary or moves upward but at a slower speed than the object; the detector array moves downward and the object remains stationary or moves downward but at a slower speed than the detector array; or both move simultaneously i.e. the object moves upward and the detector array moves downward in all of these cases successive images of the object will move clown on the detector array. Conversely, in the three reverse cases, the object will move up on tire detector array. Likewise, the object may have a sideways component of motion relative to detector array although in all cases only the vertical component of relative motion is relevant, which can be computed based on tire locus of image points on tire detector array. Using a “push-broom” approach, each point of the scene is tracked to record and extract the signal from which the hyperspectral data-cube of the full scene is calculated. Throughout the description of all embodiments we use the term ‘ point” to mean an area of the object that is focused as a point in the image. So the actual size of the point depends on the resolution of the imaging optics.

When the static polarization interferometer is formed from two birefringent wedges , then due to the symmetry double-sided interferograms are obtained. However it should he noted that also only a single birefringent wedge may be used for creating the static polarization interferometer in which case only one-sided interferograms are obtained.

However, in the imaging system 200 when the light rays such as rays 170, 170' and 170" of respective points of the scene are focused by the imaging lens 210 towards a specific detector element .180 in a detector array 400, they propagate at different angles and thus they enter the static polarization interferometer 120 at different angles. As a result, there is a change in the retardance of the different rays that may decrease the spectral resolution of the hyperspectral imaging system especially in imaging systems whose f/rmmbers are not high.

Moreover, since the spectrometer is required to he located upstream of the detector and the image is formed by a cone of rays that converges, the diameter at which the rays enter the spectrometer is significantly larger than the sample pixel size, which will reduce the spectral resolution. Also, since the spectrometer is close to the focal plane extremely high quality components and level requirements are required.

Fig. 3 shows a detail of a system 300 according to an embodiment of the invention in which a negative-power microlens array 310 is attached or juxtaposed to the input of the static polarization interferometer 120 described above (e.g. Fig 2 and Fig.l) or is disposed upstream thereof. In this case when the light rays such as rays 170, 170' and 170" of a certain point of the scene are focused by the imaging lens (not shown) towards a specific detector element in the detector array 190, they are focused to the (virtual) focal plane of a microlens on the negative-power microlens array 310 from where they diverge and propagate through the static polarization interferometer 120 mutually parallel to the chief ray 170' Le. the ray that passes through the center of the aperture stop of each mierolens. Note that in this and subsequent embodiments, the imaging lens shown as 210 in Fig. 1 is not shown in the figures, only its focused rays being shown. It will he understood tit at its location and operation are the same for all figures. As a result, the retardance of the different rays is similar and this increases the spectral resolution of the hyperspectral imaging system even in imaging systems whose f/numbers are not high. Thus, by way of example, the invention has been shown to achieve good results with a simplified and less expensive optical system having an f- number lower than 2. Optionally a positive-power microicns array 320 is attached or juxtaposed to the output of the static polarization interferometer 120 or is disposed downstream thereof to focus the rays to the detector element 180 in a detector array 190. Although the microlens array 320 is shown positive in the figure, it can be negative. In both eases, the object is imaged at the focal plane, the difference being that for positive microlenses, the focal plane is in front of the microlens (as shown in the figure), while for negative microlenses, the focal plane is behind the microlens. The microlens arrays 310 and 320 may also be attached or juxtaposed directly to the polarizer and/or the analyzer and/or on the wedges of the static polarization interferometer 120 and/or the detectors, respectively. The microlens arrays 310 and 320 may also he manufactured using polarizing and/or birefringent materials such that they will be formed integrally with the polarizers and/or with the wedges. The various microlens elements on the two microlens arrays 310 and 320 may or may not he parallel anti may or may not he of the same size. In other words, the respective optical axes of nominally opposing mierolenses in the two facing arrays may or may not be coaxial and their diameters may or may not be equal.

It should be noted that while the static polarization interferometer 120 is described above and in subsequent embodiments as employing at least one wedge formed of birefringent material and having an optical axis that is orthogonal to the optical axis of the optical system 100. The static polarization interferometer may alternatively be formed as a flat plate formed of birefringent material having an optical axis that is parallel to an optical axis of the optical system. Specific embodiments employing flat plates are described below with reference to Figs. 11 and 12.

Owing to the relative motion between the object and the measuring optical system the image of the scene moves from line to line successively on the detector array of the imaging system in a direction opposite to the direction of the relative motion (as explained above), the phase shift between the two polarizations of light varies linearly from line to line at the successive lines of the detector array. Using a “push-broom” approach, each point of the scene is tracked to record and extract the signal from which the hyperspectral data-cube of the full scene is calculated. Fig. 4 shows another embodiment according to the present invention in which instead of the negative-power microlens array 310 a positive-power microlens array 330 is added or juxtaposed upstream of the static polarization interferometer 120 (e.g. Fig. 2 and Fig.1). In this case when the light rays such as rays 170, 170 and 170² of a certain point of the scene are focused by the imaging lens (not shown) towards a specific detector element 180 in the detector array 190, from which they are focused to the front focal plane of a microlens on the microlens array 330 and from where they converge and propagate through the static polarization interferometer 120 mutually parallel to the chief ray 170. As a result, the retardance of the different rays is similar and this increases the spectral resolution of the hyperspectral imaging system even in imaging systems with small f/numbers. Optionally a positive-power microlens array 320 may be attached to the static polarization interferometer 120 to focus the rays to the detector. The microlens arrays 320 and 330 may also be attached or juxtaposed directly on the polarizer and/or the analyzer and/or on the wedges of the static polarization interferometer 120 and/or the detectors, respectively. The microlens arrays 320 and 330 may also be manufactured using polarizing and/or birefringent materials such that they will attached to the polarizers and/or on the wedges. The microlenses on the two microlens arrays 320 and 330 may or may not be parallel and may or may not be of the same size (as explained above). Owing to the relative motion between the object and the measuring optical system, the image of the scene moves from line to line successively on the detector array of the imaging system in a direction opposite to the direction of the relative motion, the phase shift between the two polarizations of light varies linearly from line to line at the successive lines of the detector array. Using a “push-broom” approach, each point of the scene is tracked to record and extract the signal from which the hyperspectral data-cube of the full scene is calculated. Figs 5a and 5b show respectively plan and side elevations of a wedge-shaped bundle of optical fibers 500. Each fiber in the bundle of optical fibers 500 formed of a birefringent material. When light with two orthogonal polarizations propagates through each optical fiber the two orthogonal polarizations acquire a phase difference ∆^ whose value is a function of (i) the difference between the refractive indices of the birefringent material and (ii) the length of the optical fiber. Since the bundle of optical fibers 500 has a wedge form, the phase difference acquired by the two orthogonal polarizations depends on the location in which the light propagates through the wedge. Fig. 6 shows schematically a static polarization interferometer 120 according to another embodiment wherein the wedge-shaped optical fiber bundle 500 described above is located between a polarizer 150 and an analyzer 160 whose axes may be parallel or perpendicular. The orientations of the respective optical axes of the birefringent material of which each optical fiber is made is preferably 45° to the respective optical axes of the two polarizers. After propagating through the polarizer and the two wedges, the two orthogonal polarizations acquire a phase difference ∆f whose value is a function of the location in which it propagates through the two wedges. Fig. 7 shows a hyper-spectral imaging system 700 according to an embodiment of the invention. In this embodiment, the light rays such as rays 170, 170 and 170² of a certain point of the scene are focused by the imaging lens of the imaging system on a static polarization interferometer 120 as shown in Fig. 6. However since the static polarization interferometer 120 comprises a bundle of optical fibers, the two orthogonal polarizations of all rays that propagate through a certain optical fiber acquire approximately the same phase difference regardless of the angle at which they were coupled to each fiber. As a result, the retardance of the different rays is similar and this increases the spectral resolution of the hyperspectral imaging system even in imaging systems whose f/numbers are not high and which therefore subtend high cone angles. Due to the relative motion between the moving object and the static measuring optical system (the hyper-spectral imaging system), the image of the scene moves from line to line successively on the detector array of the imaging system in a direction opposite to the direction of the relative motion (as explained above), the phase shift between the two polarizations of light varies linearly from line to line at the successive lines of the detector array. Using a “push-broom” approach, each point of the scene is tracked to record and extract the signal from which the hyperspectral data-cube of the full scene is calculated. It is understood that there can be more than only one wedge-shaped bundle of optical fibers in the system. A microlens array of any kind may also be added upstream or downstream the static polarization interferometer 120. Fig. 8 shows an optical system 800 comprising a thick Wollaston prism 810 which is struck obliquely by a light ray 170² at some angle relative to the optical axis of the optical system. The Wollaston prism 810 comprises two orthogonal prisms 820 and 830 of birefringent material that separate light into two separate linearly polarized outgoing beams 840 and 850 with orthogonal polarization that are split after propagating through the prism. The two beams 840, 850 are polarized according to the optical axis of the two right angle prisms. Other rays striking the same point but at different angles will split in different places across the prism face. This can prevent the usefulness of a thick Wollaston prism for a static polarization interferometer. However, Fig. 9 shows schematically a detail of a hyperspectral imaging system 900 according to another embodiment of the invention based on a static polarization interferometer 120 in which relative movement between a scene and the imaging system is exploited. The static polarization interferometer 120 comprises polarizers and a thick Wollaston prism 810 as described above. In addition, a negative- power microlens array 310 is attached or juxtaposed to the static polarization interferometer 120. In this case when the light rays such as rays 170, 170 and 170² of a certain point 910 of the scene are focused by the imaging lens 210 towards a certain detector 180 at the detector array 190, they are focused to the (virtual) focal plane of a microlens on the negative-power microlens array 310 and such they diverge and propagate through the static polarization interferometer 120 mutually parallel to the chief ray 170. As a result, all the rays from a certain point of the scene that are imaged on the same point on the static polarization interferometer 120 will propagate through the Wollaston prism at the same angle. All of them will have the same retardance for both polarizations. Approximately both polarizations that are split at the prism face may be focused by an additional lens 920 to the same detector element 180 on the detector array 190. Owing to the relative motion between the object and the measuring optical system, the image of the scene moves from line to line successively on the detector array of the imaging system in a direction opposite to the direction of the relative motion, the phase shift between the two polarizations of light varies from line to line at the successive lines of the detector array. Using a “push-broom” approach, each point of the scene is tracked to record and extract the signal from which the hyperspectral data- cube of the full scene is calculated. This can also be achieved by using a positive-power microlens array instead of a negative-power microlens array. In this case the light rays are focused by the imaging lens to the front focal plane of a microlens on the microlens array such that they converge and propagate at a given angle through the static polarization interferometer 120 while remaining parallel. Fig. 10 shows a cross section of the index ellipsoids of a positive and a negative birefringent materials. It can be seen that as an unpolarized ray of light advances on the surface defined by this cross-section, the refractive index difference of the two polarizations of the light changes as a function of the angle of progression of the ray. Fig. 11 shows schematically a detail of a system 1100 according to another embodiment of the invention having a static polarization interferometer 210 comprising a waveplate 1110 located between a polarizer 150 and an analyzer 160. The waveplate 1110 may be a flat plate made of birefringent material. The orientation of the optical axis 1120 of the birefringent material constituting the waveplate is parallel to the optical axis 1130 of the optical system or at least has a component that is parallel. The waveplate has an optical retardance that varies as a function of the direction of the propagation of the light rays through the waveplate as described in the previous paragraph. When a ray 170² propagates through the first polarizer the two orthogonal polarizations of the ray 242 and 244 that enter the waveplate split and acquire a phase difference ∆^ whose value is a function of the direction in which light ray 170² propagates through the waveplate 1110. The second polarizer is arranged to receive the light from the waveplate and polarizes light along a second polarization direction. Since the waveplate 1110 is a flat plate the two orthogonal polarizations of the ray 1140 and 1150 that emerge from static polarization interferometer 120 are parallel. Fig. 12 shows schematically a hyperspectral imaging system 1200 based on a static polarization interferometer 120 in which the refractive index difference of the two polarizations of the light changes as a function of the angle of the progression of the rays. Here too, the relative movement between a moving scene and the static imaging system is exploited. The rays emanating from an arbitrary point of the scene such as 1210 are collimated inside the optical system by any known method such as for an example a negative lens 1220. In some plane where the rays are collimated, there is placed a static polarization interferometer 210 (whose action was described previously). As described above, the refractive index difference of the two polarizations of the light changes as a function of the angle of the progression of the rays. Accordingly, the rays emanating from any arbitrary point of the scene propagates with a different angle through the static polarization interferometer 210 and thus the two polarizations of these rays acquire a different phase shift. Since the difference in phase shift is a function of the angular displacement of the ray from the principal optical axis of the optical system as distinct from being linearly dependent on the location where the ray is focused on the detector array as in previous cases, the frequency of the resulting interferogram increases or decreases with time i.e. has a “chirp”, which must be corrected prior to carrying out Fourier analysis on the interferogram. Since the two respective orthogonal polarizations of each ray that emerges from the static polarization interferometer 210 are each mutually parallel, another lens 1230 or any other optical system focuses the collimated rays to a detector 180 at the detector array 190. When there is a relative motion between the moving object and the static measuring optical system, this movement is exploited to simplify the system instead of moving one of the wedges relative to other. Since the image of the scene moves from line to line successively on the detector array of the imaging system in a direction opposite to the direction of the relative motion, the phase shift between the two polarizations of light varies linearly from line to line at the successive lines of the detector array. Using a “push-broom” approach, each point of the scene is tracked to record and extract the signal from which the hyperspectral data-cube of the full scene is calculated. It will be understood that while typically the required relative movement is achieved using a fixed optical system and a moving object, either or both can move and therefore statements relating to an object that moves relative to an optical system are to be construed as meaning that (i) the optical system is fixed and the object moves; or (ii) the object is fixed and the optical system moves; or (iii) both move but at different speeds or in different directions such that there is relative movement between the two. This principle may extend to any optical system in which the spectral behavior of the system varies with the angle and/or with the location a detector in the detector array and this dependency is known. In this scheme, when there is a relative motion between the optical system and the scene, the scene is scanned and every point in the scene is tracked as described above. The intensity function of each point can be analyzed as a function of its position across the detector and/or as a function of the angle of progression of the rays through the optical system. The intensity function of each point can be expressed by a Fredholm integral equation of the first kind with a known Kernel. Solving this Fredholm integral equation results in the spectrum of each point. In any case described above, instead of using dichroic polarizers wherein one polarizer blocks light of a first polarization but passes light of a second polarization, a PBS (Polarizing Beam Splitter) can be used. The two polarizations of light will then pass through the PBS and may be split and any of the procedures that described above may be realized for each polarization separately. If the imaging system according to any of the embodiments is used to analyze an object that is completely uniform, the system does not need to be scanned because every detector has the same spectrum from every point of the object (all points of the object are spectrally identical). Thus, by detecting a series of detectors with a different phase difference between the polarizations we get the interferogram and we can perform a Fourier transform to calculate the spectrum. If the object is not completely uniform but has a certain area which is known to be uniform, the optical system can zoom in on this region and the resulting detector rays will cover the detector array with the same spectrum (because all points in the region are spectrally identical) thus avoiding the need for scanning of at least this region. It should also be noted that features that are described with reference to one or more embodiments are described by way of example rather than by way of limitation to those embodiments. Thus, unless stated otherwise or unless particular combinations are clearly inadmissible, optional features that are described with reference to only some embodiments are assumed to be likewise applicable to all other embodiments also.