WO2009053422A1

WO2009053422A1 - Imaging device with sub-wavelength pupil coding

Info

Publication number: WO2009053422A1
Application number: PCT/EP2008/064341
Authority: WO
Inventors: Brigitte Loiseaux; Jean-Pierre Huignard; Mane-Si Laure Lee-Bouhours; Frédéric DIAZ
Original assignee: Thales
Priority date: 2007-10-26
Filing date: 2008-10-23
Publication date: 2009-04-30
Also published as: FR2923028A1; FR2923028B1

Abstract

The invention relates to an imaging device comprising at least: - a first optical subsystem (L1) making it possible to form the image of an object; - a multi-spectral detector sensitive at least in a first range of wavelengths (?-??1, ?1+??1) comprising a first so-called reference wavelength (?i) and in a second range of wavelengths (?2-??2, ?2+??2) comprising a second so-called reference wavelength (?2); - means of post-processing of the image making it possible to restore a quality image, characterized in that it comprises a phase mask comprising patterns of lower dimensions than the smaller of the reference wavelengths, said binary blazed phase mask (Mp) exhibiting a phase function at least invariant with respect to a given parameter (such as wavelength, bi-spectral optics, bi-field optics, wide-field optics, zoom function, etc.) and being combined with the post-processing means, making it possible to correct aberrations introduced by the first optical subsystem.

Description

Subwavelength pupil coded imaging device

The field of the invention is that of optical imaging devices and in particular imaging in the infra-red for the development of low cost thermal cameras for applications such as: zone protection, starting surveillance of In general, an imaging device comprises at least one lens-type optics making it possible to form the image of an object on a detector. The simplest optics of the lens type introduces a set of important defects that are not generally revolutionary. It is thus necessary to use complementary optical elements generating significant phase differences. The combination with optical elements generating phase differences that are not revolution make it possible to compensate for the defects introduced by said lens, by minimizing the number of elements.

Currently imaging devices are defined from an architecture of optical elements that implements the minimum number of dioptres, called "ideal configuration". However, when we associate the problem posed (object and image field aperture, resolution, ...) with the environmental constraints of performance and spectral range of use, we generally end up with the implementation of a combination that puts a higher number of elements than the so-called "ideal" combination. There are coded pupil techniques which make it possible to simplify combinations but which require the realization of complex phase function.

This increase in components leads to increase not only the cost but also the weight and bulk of the systems. This constraint may prove to be very problematic in certain cases, in particular for embedded systems, even prohibitive to access to new technologies, for example in the case of worn equipment such as binoculars or helmet visuals. as much as one seeks to operate multi-spectral systems. It should be noted that for the realization of multi-spectral systems, it may be envisaged to use pupillary coding principles, making it possible to reduce the wavelength sensitivity of the optics used. It has already been proposed, in order to meet all of the aforementioned constraints, to use diffractive optics as illustrated in FIG. 1, on which a diffractive lens L ₀ coupled to a second optic L ₂ makes it possible to form the image of an object O on a detector C. Such diffractive optics generate by nature a phase function and are generally machined by a diamond tip to achieve all the patterns within a material chosen for its transparency properties in a range of given wavelengths of use. Nevertheless, a number of disadvantages remain with this type of component. Diamond machining techniques are only compatible with revolution phase functions. Moreover, there exist multi-level binary techniques that require many masking steps to be effective (typically greater than four to obtain an efficiency of more than 90%), which induces both a cost and rates of The deviation from the calculated profile is rather important, related to the limitations of both the equipment used and the positioning tolerances of the different masks.

A major disadvantage lies in the fact that such structures are by nature chromatic, providing chromatic phase functions, which does not allow them to be used in active or passive imaging applications integrating monochromatic or multispectral channels (operating typically at several wavelengths).

In this context, the present invention proposes a new type of imaging device using on the one hand an optical system capable of providing a phase function rendering the image invariant with respect to a given parameter (of the type for example: length d wave, small-field optics, wide-field optics, zoom function, etc.), thereby enabling part of the optical corrections necessary for the formation of an image and, secondly, image processing means. coupled to the detector, in order to lighten the entire imaging device while ensuring the presentation of a high quality image.

More precisely, the subject of the present invention is an imaging device comprising at least:

a first optical subsystem (Li) for forming the image of an object; a multi-spectral detector sensitive at least in a first range of wavelengths (λi-Δλ-i, λ-i + Δλ-i) comprising a first so-called reference wavelength (λi) and in a second range of wavelengths (X ₂ -AX ₂ , λ ₂ + Δλ ₂ ) comprising a second so-called reference wavelength (λ ₂ );

means for post-processing the image making it possible to restore a quality image, characterized in that it comprises a phase mask comprising patterns of dimensions smaller than the smallest of the reference wavelengths, said mask binary phase binary phase (M _p ) having a phase function at least invariant with respect to a given parameter (wavelength type, bi-spectral optical, bi-field optical, wide field optics, zoom function, ... ) and being combined with the post-processing means for correcting aberrations introduced by the first optical subsystem.

According to a variant of the invention, the phase mask has an invariant phase function by defocusing.

According to a variant of the invention, the function may be of the poiynomial type or of the exponential type. According to a variant of the invention, the phase mask has a polynomial function corresponding to the following equation: Φ (x, y) = a χ (x * + y ³ )

With x and y the coordinates perpendicular to the optical axis Z of the device. According to a variant of the invention, the image post-processing means provide a filter function that reverses the percusion response (PSF) delivered by the binary coding operation introduced by the phase mask.

According to a variant of the invention, the phase mask comprises a blazed network.

According to a variant of the invention, the blazed network is binary at two levels and comprises rungs, each rung comprising periodic patterns of variable size and distributed over a given period of time less than the reference wavelength. According to a variant of the invention, the reference wavelengths are in the infrared range.

According to a variant of the invention, the reference wavelengths are in the X-band. According to a variant of the invention, the reference wavelengths are in the ultraviolet domain.

According to a variant of the invention, the reference wavelengths are of the order of a few hundred microns.

According to a variant of the invention, the first range of wavelengths is defined in the vicinity of 3-5 microns, the second wavelength range is defined in the vicinity of 8-12 microns.

According to a variant of the invention, the phase mask is made of a material of GaAs type for example, or Zns, ZnSe or GaSNR @ from Umicore. According to one variant of the invention, the first optical subsystem is reflective or refractive lens type.

According to a variant of the invention, the first optical subsystem is of the Fresnel lens type.

According to a variant of the invention, the first optical subsystem and the phase mask are on both sides of the same optical component, one face providing a diffractive lens function, the other side providing the mask function of phase.

The invention also relates to an embedded system comprising an imaging device according to the invention.

The invention will be better understood and other advantages will become apparent on reading the description which follows, given by way of non-limiting example and thanks to the appended figures among which:

FIG. 1 illustrates an imaging device comprising diffractive optics according to the known art;

FIG. 2 illustrates a first example of an imaging device according to the invention involving phase coding;

FIG. 3 illustrates an exemplary phase function of an exemplary phase mask used in an imaging device of the invention; FIG. 4 illustrates the diffraction efficiency for a conventional ladder diffractive optical element and of the ladder grating type respectively with a binary synthesis of this element, by means of pillar-type microstructures of a diffractive optical element. conventional, as a function of the ratio of the illumination wavelength to the nominal wavelength;

FIGS. 5a and 5b respectively illustrate a conventional diffractive optic element with a ladle grating and with a binary synthesis of this element, by means of microstructures of the pillar type.

In general, the present invention provides an imaging device that combines a pupil-coded imaging approach and an approach for using linear blazed linear sub-wavelength optical surfaces.

FIG. 2 illustrates a first example of an imaging device according to the invention. This device comprises a first optical Li, a phase mask M _p , a second optical L ₂ and a sensitive detector C in a range of wavelengths (λ _o -Δλ, λ _o + Δλ), said detector being coupled in output to image processing means Tj to ensure the quality of the images in support of said phase mask. Indeed, the phase mask provides a phase law to correct some of the optical aberrations, the other part being provided by the digital image processing. More precisely, the phase mask provides an invariant image of an object O, for example by defocusing (the image at the output of the component M _p appears fuzzy) and is coupled to image processing means that make it possible to restore the image. image by applying for example a filter which reverses the coding (effect induced by the coding on the percusionneïle response).

It is thus possible to obtain a clear image for a large focusing interval. An example of a phase law can be: ΦtX y) = cc ^χ (x ³ + y ¹ )

With x and y the coordinates perpendicular to the optical axis Z of the imaging device. FIG. 3 illustrates an example of phase function evolution with a coefficient α = 90.41 which shows the significant amplitude of the phase law.

The advantage of such a solution is to simplify the complexity of the imaging chain leading to a solution closer to that of an imaging system implementing a minimal number of optical elements, relying on the advantages specific to the two following techniques:

an optical coding which implements the use of a phase law making it possible to make the optical combination insensitive to certain types of aberration and a frequency filtering in the detection plane to restore the quality of the image formed on the detection plan;

performing part of the optical function with sub-length binary blazed type diffractive elements, so as to efficiently obtain a phase function for the pupil coding, or even a part of the optical combination.

Such sub-wavelength bias-like bifunctional diffractive elements have in particular been described in the publication "Extended depth of field wave-front coding": Edward R. Dowski, Jr., and W. Thomas

Cathey (April 1995 @ Vol.34, No. 11 © Applied Optics) and will be explained in more detail below.

In general, with conventional diffractive optical elements such as Fresnel lenses and ladder or multi-level gratings, the desired blaze effect is obtained by a gradual change in the depth of a cladding material. constant. The surface profile of these elements thus consists of continuous reliefs separated by discontinuities. These elements are designed for a certain wavelength, called the design wavelength, denoted λ ₀ .

These elements have a significant limitation to their use, because they are in practice biabled only at the design wavelength. At the design wavelength, the diffraction efficiency is 100% at Fresnel losses, but as soon as the wavelength of the incident light deviates from this nominal value, efficiency in the order of blaze (order 1) drops considerably (effect amplified by the difficulty of technological achievement).

This is shown on the curves 4a and 4b of FIG. 4, which show the efficiency curve for a conventional diffractive optic element with echelette and of the echelette grating type with a binary synthesis of this element, on FIG. means of pillar-type microstructures of a conventional diffractive optical element, as a function of the ratio of the illumination wavelength to the nominal wavelength. In the scalar domain, it can indeed be shown that the diffraction efficiency as a function of the wavelength of any diffractive optical element blazed in order 1 is given by the following equation:

where sinc (z) = - -, whose curve 1 is the graphical representation

X Light lost in blaze order (order 1) is transmitted in higher order. If we take the example of a hybrid lens, this phenomenon results in the transmission of a parasitic light, which is detrimental to the quality of imaging. This is conveniently reflected by the appearance of diffracted light throughout the image plane. For example, the output image is not clear, or little contrast because of the noise level provided by the parasitic orders especially if there are hot spots at the scene that we want to image.

It is shown that this drop in efficiency is due to the low dispersion of the material, which means that for a small difference in wavelength, the phase difference at φ (λ) induced in the structure deviates from 2π, whereas it is 2π at design wavelength λ ₀ or blaze.

For example, if we consider the echelette array RE represented in FIG. 5a, Aφ (λ) represents the phase difference between the low b and the high t of a step e of the network RE. This difference is equal to 2π for an incident light at λ ₀ .

In the scalar approximation, neglecting the Fresnel losses, and placing itself at the normal incidence of the incident beam, the difference in δ

phase as a function of the wavelength Aφ (λ) and the efficiency as a function of the wavelength η (λ) are in fact given by:

η {Â) = smc ² [π (l-φ (λ) / 2π)], Eq. (3) where Δn (λ) = (n (λ) - Pair is (n (λ) - 1), for a diffractive optical element etched in a refractive index material n.

For these diffractive optical elements, we can consider that Δn (λ) = Δn (λo), because the dispersion of the material is negligible: the refractive index varies little around λ ₀ . Eq (2) becomes: Δφ {λ) = 2π- ± Eq. (4)

A and we obtain the equation Eq {1) seen above and represented in FIG. 4, replacing Aφ (λ) by this expression in Eq (2).

Thus, the low dispersion of the optical material in the conventional diffractive optical elements results in a drop in the efficiency of the diffraction with the wavelength λ ≠ λo expressed by the equation Eq (1). These conventional diffractive optical elements are therefore not broadband spectral. They can not be used in optical systems dedicated to broadband spectral applications, such as hybrid optical systems, composed of refractive and diffractive optics. Other elements of diffractive optics known as binary microstructure, also called binary blazed networks, or sub-wavelength diffractive optical elements (SWDOE: "SubWavelength Diffractive Optic Element") are known. binary synthesis of a conventional diffractive optical element: we start from the classical diffractive optical element that we want to synthesize and we sample this network to obtain points, to which we can associate an index value The sampling must be done at a period less than the design wavelength, to obtain a sub-wavelength grating.The various calculation techniques used are known to the man of the These techniques make it possible, for example, for a blazed echelette network such as the RE network shown in FIG. 5a, to define a blazed network bin. area as shown in Figure 5b. 5a, two echelon echelon gratings of period Λ {or not of the network) are represented. These steps are etched in an optical material of index n.

A blazed binary network corresponding to the RE network of FIG. 5a is shown in FIG. 5b. The network RE is sampled at the period Λ _s chosen to be smaller than the design wavelength λ ₀ . A certain number of points are obtained for each period Λ of the network. At each point a given fill factor is matched for a given microstructure type (hole, pillar): this fill factor is equal to the d dimension of the microstructure relative to the sampling period of the network: f = d / Λ _s. The filling factor of each microstructure is defined, by known calculations, to locally give a phase shift value Φ (x) similar to that of the echelette grating at the sampled point, and equal in known manner to Φ (x) ≈2π (n). -I) h (x) -, where x

A is the coordinate of the sampled point on the Ox axis of the network.

In the example of FIG. 5b, the binary microstructures are of the pillar type. A set of binary microstructures is obtained which encodes the grating pattern. This set of microstructures is repeated at the period Λ of the echelette grating of FIG. 5a. The use of this type of structure makes it possible to produce materials with an effective index whose index value makes it possible to have a wider range of materials (as a first approximation varying between the index of the substrate and the air , so typically can vary between 3 and less than 1.5, considering the materials used in particular in the infrared). Moreover, such materials offer a chromatic dispersion that depends on the geometry of the structures, which has the advantage of producing achromatic phase functions and thus "ladder" type sampling without loss of efficiency as a function of the length of the structure. wave. It is furthermore possible in this type of coding of the diffraction function to use only one level of masking during an etching operation, which may be of the RIE etching type, leading to the development of type of phase mask blazed binary. It is also possible to use replication methods for mass production, such as vitreous material molding technology, resin embossing deposited on the optical component or CD-type optical replication, which have proved their worth in terms of resolution. The advantages of this new class of optical components makes it possible to envisage the realization of more complex optical functions than those used in the design of traditional imaging systems, in particular it makes it possible to envisage performing complex and high-performance aspheric functions. amplitude (typically a few tens of wavelengths) necessary for the implementation of pupil-coded imaging principles. In particular, it makes it possible to benefit from the achromatism properties that allow the wavelength scale reduction, the amplitude of the difference in gait difference.

The present invention may also be of interest in the context of new technologies, for example in the case of worn equipment such as binoculars or helmet visuals, in which multi-spectral systems are used.

Thus and according to the invention, the imaging device comprises for example a multi-spectral detector, that is to say both sensitive in a first range of wavelengths (λ _r Δλi, λ-i + Δλ-i) having a first so-called reference wavelength (λ-i) and a sensitivity in a second wavelength range (λ2-Δλ ₂ , λ ₂ + Δλ ₂ ) comprising a second so-called second wavelength reference (A2). A single phase mask M _p having patterns smaller than the smallest of the reference wavelengths is positioned upstream of the multi-spectral detector.

In conclusion, the ability to implement pupil-coded imaging techniques, where part of the correction is provided by signal processing, also opens the way for the simplification of optical systems on the basis of a test criterion. different optimization, that of an invariance transfer function in spatial frequency, which offers additional degrees of freedom, thus making it possible to overcome certain optical elements called "athermalization", or to avoid the switching of an "optical bioc" in the case of multi-spectral imaging.

Claims

1. Imaging device comprising at least:

a first optical subsystem (Li) for forming the image of an object;

a multi-spectral detector sensitive at least in a first range of wavelengths (λi-Δλ-i, λi + Δλi) comprising a first so-called reference wavelength (λ-i) and in a second range of wavelengths (λ ₂ -Δλ ₂ , λ ₂ + Δλ ₂ ) comprising a second so-called reference wavelength (λ ₂ );

means for post-processing the image making it possible to restore a quality image, characterized in that it comprises a phase mask comprising patterns of dimensions smaller than the smallest of the reference wavelengths, said mask phase blazed binary (M _p ) having a phase function at least invariant with respect to a given parameter (wavelength type, bi-spectral optical, bi-field optical, wide field optics, zoom function, ... ) and being combined with the post-processing means for correcting aberrations introduced by the first optical subsystem.

2. Imaging device according to claim 1, characterized in that the phase mask has an invariant phase function by defocusing.

3. Imaging device according to claim 2, characterized in that the phase mask has a function of the type poiynomial or exponential.

4. Imaging device according to claim 3, characterized in that the polynomial function satisfies the following equation:

Φ (x, y) = ax (x ³ + y ³ )

With x and y the coordinates perpendicular to the optical axis Z of said imaging device.

5. Imaging device according to one of claims 1 to 4, characterized in that the image post-processing means provide a filter function that reverses the result from the binary coding operation introduced by the mask phase.

6. Device according to one of claims 1 to 5, characterized in that the phase mask comprises a blazed network.

7. Device according to claim 6, characterized in that the binary blazed network comprises echelettes, each scale having periodic patterns of period (Λs) less than the reference wavelength.

8. Imaging device according to one of claims 1 to 7, characterized in that the reference wavelength is in the field of infra-red.

Imaging device according to one of claims 1 to 7, characterized in that the reference wavelength is in the X band.

10. Imaging device according to one of claims 1 to 7, characterized in that the reference wavelength is in the field of ultraviolet.

11. Imaging device according to one of claims 1 to 7, characterized in that the reference wavelength is of the order of a few hundred microns.

An imaging device according to claim 11, characterized in that the first wavelength range is defined in the vicinity of 3-5 microns, the second wavelength range is defined in the vicinity of 8-12 microns. .

13. Imaging device according to one of the preceding claims, characterized in that the phase mask is made of a GaAs ₁ ZnS, ZnSe type material.

14. Imaging device according to one of claims 1 to 13, characterized in that the first optical subsystem is of reflective or refractive lens type.

15. Imaging device according to one of claims 1 to 14 characterized in that the first optical subsystem is of the type of lens of

Fresnel.

16. Imaging device according to one of claims 1 to 15 characterized in that the first optical subsystem and the phase mask are on both sides of the same optical component, a face providing a diffractive lens function the other side providing the phase mask function.

17. Embedded system characterized in that it comprises an imaging device according to one of claims 1 to 16.