OPTICAL IMAGING SYSTEM WITH MULTIPLE PARALLEL PATHS AND COMMON ENTRY OPENING

Description

Domaine technique

[0001] The present description relates to an optical imaging system with several channels which are arranged in parallel and with an input aperture which is common to all these channels. It also relates to a method for capturing several images which are associated with an optical input field which is the same for all these images. Technique antérieure

[0002] Many fields require detecting the presence of a specific gas in a scene, this gas may be invisible. For example, such gas detection is useful in industry when the gas is toxic or polluting, and / or to search for the possible presence of a gas leak in facilities for transferring or using this gas. In particular, it is often useful to detect the presence of hydrocarbon gas in industrial facilities. Gas detection in a scene is also useful for military applications, in particular to search for the possible presence of combat gas in an intervention zone.

[0003] Typically, the gases of interest have characteristic absorption bands in the spectral range between 3 µm (micrometer) and 5 µm, commonly referred to as the mid-infrared or MWIR region, or between 8 µm and 14 µm, commonly referred to as the far-infrared or LWIR region. Imaging optical systems incorporating germanium (Ge) lenses are then often used to optically detect the gas, but other materials that are also transparent in the desired spectral range can be used alternatively for such systems. But these materials, including germanium, have their optical properties that vary greatly depending on the thermal evolution of the environment.It is then known to use thermal compensation devices, also called athermalization devices, so that the images which are captured remain clear whatever the operating temperature of the system within a prescribed thermal interval.

[0004] Furthermore, it is also known to equip an optical imaging system with a vignetting screen. Such a vignetting screen usually has the function of removing stray light which would otherwise enter the optical imaging system at very inclined angles of incidence relative to the optical axis of the system, without participating in the formation of images.

[0005] Furthermore, the selective detection of a gas requires implementing several optical paths in parallel in the optical imaging system that is used, these optical paths being functional in spectral windows that are different from one optical path to another. It is thus possible to identify the gas or the family of gas that is present in the scene, according to its image rendering for all the spectral windows. But a challenge is then to have multi-path optical imaging systems that are compact, lightweight, easy to manufacture, particularly with regard to the alignment of their optical components, and that are operational in sufficiently large operating temperature ranges.

[0006] An architecture that offers many advantages for such optical imaging systems is described in the article by J. Tanida et al., entitled "Thin observation module by bound optics (TOMBO): concept and experimental verification", Appl. Opt., vol. 40, pp. 1806-1813 (2001). Such a TOMBO system comprises: a matrix image sensor; and an imaging matrix, comprising several imaging optics which are arranged in parallel and adapted to simultaneously form, in useful areas of the matrix image sensor which are dedicated one by one to the imaging optics, respective images of a scene contained in an input optical field which is identical for all these imaging optics, the useful areas of the matrix image sensor being disjoint and each imaging optic with the corresponding useful area belonging to one of the optical paths of the system separately from each other optical path.

[0007] To prevent an image that is intended to be captured in one of the useful areas from overflowing beyond the peripheral limit of this useful area into a neighboring useful area, it is known to use separating walls between the optical paths. Document US 10,375,327 B2, or US 2018 / 0191967 A1, describes such TOMBO systems with separating walls between neighboring optical paths, for the gas detection application. Such separating walls, which are opaque and commonly called low walls, are arranged longitudinally between neighboring optical paths, at the level of the imaging matrix, extending to the matrix image sensor. However, they have the following drawbacks: when these separating walls extend across the entire imaging matrix, the lenses of the latter can no longer be produced in the form of single-piece lens matrices, and because of this alignments must be carried out optically by optically, which is tedious and complex, and which increases the cost price of the system to a significant extent; and such separating walls are especially effective in preventing image overlaps between neighboring optical paths when they extend to near or very near the surface of the matrix image sensor. But for the MWIR and LWIR spectral ranges, the matrix image sensor is contained in a reduced pressure enclosure, so that a window of this enclosure is necessarily present between the imaging matrix and the matrix image sensor.For this reason, the dividing walls end at the porthole, and therefore at a distance from the matrix image sensor, and because of this their effectiveness in avoiding image overlap between neighboring optical paths is insufficient.

[0008] Also to avoid such image overlaps within each useful area, such as resulting from overflows of each image beyond the peripheral limit of the useful area which is dedicated to this image, it is also known to provide guard bands ("clearance bands" in English) between useful areas which are neighbors within the matrix image sensor. These guard bands are dimensioned so that each overflow of any one of the images in one of the guard bands towards a neighboring useful area is shorter than a width of this guard band, for any value of the system temperature in the prescription interval. In general, the useful areas are determined by reading selections of the matrix image sensor during each image capture sequence, restricted to those of the photodetectors which belong to one of the useful areas.But such guard bands cause underutilization of the image sensor.

[0009] To limit or eliminate overlaps between images formed by neighboring optical paths, it is possible to separate the optical paths laterally from each other. But this again results in underutilization of the image sensor.

[0010] Finally, increasing the distance between the imaging matrix and the matrix image sensor to reduce the overlaps between images that are formed by neighboring optical paths prevents the system from being compact. However, there is a significant challenge for such optical imaging systems to be compact, for example to be easily transportable by an operator or by an autonomous vehicle, in particular a drone, or to be mounted on a swiveling turret.

[0011] US2012 / 344851 discloses an example of the prior art. Problème technique

[0012] From this situation, an object of the present invention is to provide an athermalized TOMBO system for which image overlaps are eliminated in all useful areas of the matrix image sensor.

[0013] Additional aims of the invention are that the system is compact, lightweight, easy to manufacture, particularly with regard to the alignment of its optical components, and that it can have a wide field of view with a large aperture. Indeed, the wide field of view makes it possible to detect a leak inside a scene that is laterally extended, and the large aperture makes it possible to increase a signal-to-noise ratio of the images that are captured. Résumé de l'invention

[0014] To achieve at least one of these aims or another, a first aspect of the invention proposes a new optical imaging system with several optical paths arranged in parallel, of the TOMBO system type as mentioned above. It therefore comprises a matrix image sensor and an imaging matrix according to the aforementioned arrangement. The optical imaging system of the invention further comprises: a thermal compensation device, arranged to produce a variable spacing between the imaging matrix and the matrix image sensor, and comprising, in at least one longitudinal section plane which is perpendicular to the imaging matrix and to the matrix image sensor, a meander composed of two segments each consisting at least in part of a material whose thermal expansion coefficient value is different from that for the other segment, the two segments being connected to each other by respective distal ends thereof, a proximal end of one of the segments being connected to the imaging matrix and a proximal end of the other of the segments being connected to the matrix image sensor,a difference between the values ​​of the coefficient of thermal expansion and the respective lengths of the two segments being adapted to produce the variable spacing between the imaging matrix and the matrix image sensor so that the images which are formed in the useful zones of the matrix image sensor are clear when the temperature of the system has any value within the prescription interval; and a vignetting screen ("baffle" in English), arranged upstream of the imaging matrix with respect to a direction of propagation of radiation which enters the system to form the images, the vignetting screen having an opening for the radiation which is common to the optical paths of the system.

[0015] According to a first additional characteristic of the invention, the opening of the vignetting screen is dimensioned to laterally limit illumination in each image which is formed by any one of the optical paths, called the optical path considered, in order to prevent this image from overflowing into the useful zone of another of the optical paths, adjacent to the optical path considered. In this way, any superposition of images is prevented in all the useful zones, for any value of the temperature of the system in the prescription interval.

[0016] According to a second additional characteristic of the invention, the distal ends of the segments of the thermal compensation device are located upstream of the imaging matrix relative to the direction of propagation of the radiation which enters the system to form the images. Furthermore, the vignetting screen is supported by the thermal compensation device upstream of the imaging matrix, again relative to the direction of propagation of the radiation which enters the system to form the images. Preferably, the vignetting screen can be supported by the thermal compensation device at the distal ends of the segments of this thermal compensation device.

[0017] Thus, the vignetting screen is supported by the thermal compensation device. In other words, the invention proposes to combine two functions for the thermal compensation device. The first function consists in adapting the distance between the imaging matrix and the matrix image sensor in accordance with the variations in the optical characteristics of the imaging matrix, essentially its focal length, as a function of the temperature. In this way, the images that are captured by the matrix image sensor remain sharp throughout the prescription interval, which may be large. The second function of the thermal compensation device consists in supporting the vignetting screen. Such a combination of functions is particularly effective in obtaining a system that is compact.In particular, the additional function of supporting the vignetting screen, introduced by the invention, does not necessarily require the meander of the thermal compensation device to be extended, compared to its length necessary for the athermalization function. According to the invention, the respective materials of the two segments of the thermal compensation device are chosen to produce the athermalization function and also to support the vignetting screen at a position which eliminates image overlaps in the useful areas of the matrix image sensor.

[0018] Thanks to the invention, the use of partition walls between the optical paths is no longer essential. The imaging matrix can then be made up of single-piece lens matrices, so that the optical alignment of the entire system is simple.

[0019] Also, due to the absence of separating walls, a porthole can be inserted between the imaging matrix and the matrix image sensor, in particular a reduced pressure enclosure porthole when the image sensor requires such an enclosure for its operation.

[0020] Generally, when those of the useful areas which are aligned between two opposite edges of the matrix image sensor are counted progressively from one of these edges, the opening of the vignetting screen and the distance between this vignetting screen and the imaging matrix can be dimensioned so that the overflow in the guard band which is intermediate between the first and second useful areas, of the image corresponding to the second useful area is limited by the vignetting screen so as to stop at a limit between the guard band and the first useful area.Alternatively, the opening of the vignetting screen and the distance between this vignetting screen and the imaging matrix can be dimensioned so that the overflow in the guard band which is intermediate between the first and second useful zones, of the image corresponding to the first useful zone is limited by the vignetting screen so as to stop at a limit between the guard band and the second useful zone. This is a condition of non-overlap between images of neighboring optical paths, expressed for the guard band which is the outermost laterally.

[0021] In preferred embodiments of the invention, at least one of the following additional features may be optionally reproduced, alone or in combination of several of them: a separation distance between the vignetting screen and the imaging matrix, measured parallel to a common direction of respective optical axes of the imaging optics, may be between 1.0 times and 6.0 times an average value of the focal lengths of the imaging optics; the imaging optical system may further comprise an enclosure which is arranged to contain the matrix image sensor, this enclosure being provided with a porthole which is transparent to the radiation which forms the images, and which is located between the imaging matrix and the matrix image sensor; the imaging optics may have a row and column arrangement in the imaging matrix, each row being perpendicular to each column, and the imaging matrix may then be oriented so that each row of imaging optics is parallel to a row or column direction of photodetectors of the matrix image sensor;the row and column arrangement of the imaging optics in the imaging matrix may be of one of the following dimensions: 1 x 2, 2 x 1, 2 x 2, 2 x 3, 3 x 2, 3 x 3, 2 x 4, 4 x 2, 3 x 4, 4 x 3 and 4 x 4. However, a limited number of optical paths may be preferred, so as not to require guard bands that are too wide when they are very much offset from a central optical axis of the imaging matrix; when the row and column arrangement of the imaging matrix comprises at least three imaging optics that are aligned in a row or column direction, then any two of the guard bands that are parallel to each other may have widths that are identical; each imaging optic may comprise at least one lens that is germanium-based; the imaging matrix may consist of single-piece matrices of lenses;each imaging optic may comprise a spectral filter, and the spectral filters of two of the imaging optics which are distinct in the imaging matrix may determine spectral transmission windows which are different; the spectral filter of each imaging optic may be located between two lenses of this imaging optic, in particular to avoid parasitic images which would result from multiple reflections between this spectral filter and the matrix image sensor, and to avoid increasing the distance between the imaging matrix and the matrix image sensor;the aperture number, commonly noted F#, of each optical channel of the imaging optical system can be between 0.9 and 1.5, corresponding to large pupil diameter values. Such aperture values, associated with the fact that the diameter of the optics must be less than or equal to the individual size of their image areas allocated on the detector in accordance with the TOMBO architecture, can advantageously be obtained by using materials with high refractive index values, such as germanium. But these materials generally have greater thermal sensitivity than materials with lower refractive index values, such as chalcogenide glasses. The thermal compensation device provided by the invention is then even more advantageous; the useful field of view, commonly noted FOV, of each optical channel can be between 20° (degree) and 60°, corresponding to wide or very wide fields of view;and the thermal prescription interval may have a length greater than or equal to 60°C (degree Celsius), preferably greater than or equal to 80°C. Generally, the thermal prescription interval is written on the imaging optical system, or may be indicated in a hardware or electronic notice that is provided with this system. ;

[0022] A second aspect of the invention provides a method for capturing multiple images associated with an input optical field that is identical for all of those images, this method being implemented using an imaging optical system that is in accordance with the first aspect of the invention.

[0023] Such a method may be intended to reveal a presence of a gas within the input optical field. In this case, each optical channel is provided with a spectral filter as indicated above, and the spectral filters may advantageously be selected so that the spectral transmission window of at least one of the spectral filters is in a transparency band of the gas, and the spectral transmission window of at least one other of the spectral filters is in an absorption band of the gas.

[0024] In particular, the gas sought by such a process may be natural gas, a hydrocarbon gas(es), a toxic gas, for example hydrogen sulfide (H 2 S) or carbon monoxide (CO), or a greenhouse gas, in particular carbon dioxide (CO 2 ). Brève description des figures

[0025] The characteristics and advantages of the present invention will appear more clearly in the detailed description below of non-limiting exemplary embodiments, with reference to the appended figures among which: [ Fig. 1 ] is a sectional view of a part of an optical imaging system according to the invention; [ Fig. 2 ] shows useful light rays for an optical path of the imaging optical system of [ Fig. 1 ] ; [ Fig. 3 ] is a perspective view of an optical path assembly support, which can be used in the imaging optical system of [ Fig. 1 ] ; [ Fig. 4 ] is an exterior perspective view of the imaging optical system of [ Fig. 1 ] ; [ Fig. 5a ] shows several optical paths of the imaging optical system of [ Fig. 1 ], which are juxtaposed in a longitudinal section plane; [ Fig. 5b ] shows thermal and geometric sizing parameters of the imaging optical system of [ Fig. 1 ], in the longitudinal section plane of [ Fig. 5a ] ; And [ Fig. 5c ] corresponds to [ Fig. 5a ] to show optical and geometric sizing parameters of the imaging optical system of [ Fig. 1 ]. Description détaillée de l'invention

[0026] For the sake of clarity, the dimensions of the elements shown in these figures do not correspond to actual dimensions or to actual dimensional ratios. In addition, some of these elements are represented only symbolically, and identical references indicated in different figures designate identical elements or those having identical functions.

[0027] The embodiment of the invention which is now described with reference to [ Fig. 1 ]-[ Fig. 4 ] comprises a 2 x 3 matrix of juxtaposed optical channels. Reference 2 generally designates this matrix of optical channels, which has been called an imaging matrix in the general part of this description. Fig. 1 ] shows the following elements of the imaging optical system: a matrix image sensor, which is designated by the reference 1; two of the optical paths, which are juxtaposed and designated by the references 2 1 and 2 2 , respectively; the mechanical connection between the matrix image sensor 1 and the imaging matrix 2, which is symbolically represented by broken lines and constituted by a thermal compensation device 3; and a vignetting screen 4.

[0028] The matrix image sensor 1 may be of a bolometer or microbolometer type which are arranged in a matrix of rows and columns, for example 1024 columns by 768 rows. Advantageously, it may be of an operational type without a cooling system. This matrix image sensor 1 is contained in a reduced pressure enclosure 1a, the latter being provided with a window 1b which is transparent to the radiation intended to be detected by the matrix image sensor 1. For example, and in particular for gas detection applications, the spectral sensitivity range of the matrix image sensor 1 may comprise at least one of the following two wavelength intervals: 3 µm - 5 µm and 8 µm - 14 µm.

[0029] In such an embodiment, all optical paths may have constitutions that are similar, and each optical path may comprise two lenses and a spectral filter. Thus, optical path 2 1 may comprise lenses 21 1 and 22 1 , as well as spectral filter 23 1 , and optical path 2 2 may comprise lenses 21 2 and 22 2 , as well as spectral filter 23 2 , and likewise for the other optical paths of the system. Advantageously, the lenses 21 1 and 21 2 , as well as their counterparts of the other optical paths of the system, can be formed in a single piece of transparent material, for example germanium, to constitute a first matrix of lenses 21. Similarly, the lenses 22 1 and 22 2 , as well as their counterparts of the other optical paths of the system, can be formed in a single other piece of transparent material, which can also be germanium, to constitute a second matrix of lenses 22.Such a design of the entire optical path of the imaging optical system facilitates the optical alignment of the entire system, thanks to the fact that the lenses are directly manufactured parallel to each other within each lens array. Preferably, the imaging array 2 is oriented so that its rows or columns of optical paths are parallel to the rows of bolometers or microbolometers in the matrix image sensor 1.

[0030] The optical paths are distinguished from each other by their respective spectral filters, which have transmission windows that are different from each other, all included in the spectral sensitivity range of the matrix image sensor 1. These transmission windows of the spectral filters can be selected according to a gas to be searched for to which the imaging optical system is dedicated. The person skilled in the art then knows how to select the spectral filters according to the absorption and transparency bands of the gas to be searched for.

[0031] Each optical channel combines a scene that is contained in the input optical field of the imaging optical system, and that is common to all the optical channels, with a respective portion of the matrix of the image sensor 1, called the useful area of ​​the matrix image sensor for this optical channel. Thus, at each reading sequence, the matrix image sensor 1 simultaneously captures all the images of the scene that are formed in the useful areas by all the optical channels. Such a multi-channel imaging optical system architecture is known in the prior art by the acronym TOMBO, for “Thin Observation Module by Bound Optics”.

[0032] [ Fig. 2 ] shows light rays that are transmitted by one of the optical paths, for example the optical path 2 1 . The useful area of ​​the matrix image sensor 1 for each optical path may be 254 columns by 319 lines for example, and two of the optical paths that are neighboring in the imaging matrix 2 may be separated by 130 lines or columns of the matrix of the image sensor 1. ZU 1 denotes the useful area of ​​the optical path 2 1 . For such an embodiment, all the optical paths may have a length L mod of the order of 13 mm (millimeter), measured between the front face of the lens matrix 21 and the photosensitive surface of the matrix image sensor 1.

[0033] In particularly advantageous embodiments of the imaging optical system, some of the components of its optical paths may be supported by an assembly support 20 which is common to them, as indicated in [ Fig. 1 ]. An example of such a support 20 is shown in [ Fig. 3 ]. It is provided with holes T, for example cylindrical, which are assigned one by one to the optical paths to allow the passage of light rays to the matrix image sensor 1. These holes T are separated from each other by portions of the material of the support 20, which is opaque for the spectral sensitivity range of the matrix image sensor 1. These portions of opaque material constitute longitudinal separations between neighboring optical paths, i.e. separation walls. In this way, stray light which would be likely to pass from one optical path to a neighboring optical path is eliminated. Finally, raised ribs NR can be provided on these separation walls, on one side of the support 20, to hold in place the respective spectral filters of the optical paths. [ Fig. 3 ] thus shows the support 20 by its side which is intended to be oriented towards the scene to be analyzed, that is to say by its side which is opposite the matrix image sensor 1 within the system. The rim R 21 is intended to support the lens matrix 21, and a similar rim R 22 is provided on the other side of the support 20 to support the lens matrix 22 in front of the matrix image sensor 1 (see [ Fig. 1 ]). Wedging and centering systems which are within the reach of those skilled in the art may also be provided to press the lens matrices 21 and 22 against the support 20, at the level of the edges R 21 and R 22 respectively.

[0034] The 20 bracket and the 4 vignette screen are visible in [ Fig. 1 ] And [ Fig. 4 ]. The vignetting screen 4 has a single inlet opening O 4 which is common to all the optical paths, and through which the light rays enter the system for all these optical paths, until they reach the useful zone which corresponds separately to each of them on the matrix image sensor 1. Advantageously, the optical imaging system can be protected on its upper part by a thermal screen 9, commonly called a cap, to prevent it from undergoing inhomogeneous thermal variations caused by temporary exposure to solar radiation or to cold air currents. Indeed, as will appear below, the thermal compensation device 3 which is used within the optical imaging system requires, in order to function correctly, that the temperature be substantially homogeneous and stabilized within the system.

[0035] The athermalization function of the thermal compensation device 3 is to keep the images captured by the matrix image sensor 1 sharp when the temperature of the system varies. To this end, the thermal compensation device 3 causes, as a function of the temperature of the system, variations in the distance between an optical center of the imaging matrix 2 and the matrix image sensor 1. This distance is the focal length of the imaging matrix 2, denoted f, and its variations as a function of temperature are given approximately by the formula known to those skilled in the art: Δf = α − 1 n − 1 ⋅ dn dT ⋅ ΔT ⋅ f where α is the thermal expansion coefficient of the material of the lens matrices 21 and 22, i.e. germanium in this case, n is the optical refractive index of this material, and T is the temperature of the imaging optical system. The factor in parentheses is commonly called the thermo-optical coefficient and denoted β. [ Fig. 5a ] shows the focal length f of the imaging optical system which is the subject of this detailed description. In this figure, the imaging matrix 2 is replaced by a single matrix of lenses each equivalent to the optical doublet of the corresponding optical path. The plane of this figure is perpendicular to that of [ Fig. 1 ].

[0036] The configuration of the thermal compensation device 3 which is used by the invention is shown in [ Fig. 5b ]. In the plane of this figure, which is the same longitudinal section plane of the system as that of [ Fig. 5a ], perpendicular to both the imaging matrix 2 and the matrix image sensor 1, the thermal compensation device 3 has a meander shape which extends upstream of the imaging matrix 2, that is to say on a side thereof which is opposite the matrix image sensor 1. It comprises two segments 31 and 32 which are each made of a material different from that of the other segment, with a difference between the respective values ​​of the coefficient of thermal expansion of these two materials which is non-zero and noted ΔCTE. In a simplified approach, the two segments 31 and 32 are parallel, with a common length which is noted L'. The respective proximal ends of the two segments 31 and 32 are not directly connected to each other. The proximal end 31p of the segment 31 is connected to the matrix image sensor 1 by an intermediate part 33 whose thermal variations can be neglected.The proximal end 32p of the segment 32 is connected to the support 20 of the imaging matrix 2. The two segments 31 and 32 extend upstream of the imaging matrix 2, or in front of it opposite the matrix image sensor 1, as shown in [. Fig. 5b ], and their respective distal ends, denoted 31d and 32d are connected to each other. When the thermal compensation device 3 has such a configuration, the thermal variation that it produces for the separation distance between the imaging matrix 2 and the matrix image sensor 1 is: ΔL ′ = ΔCTE ⋅ ΔT ⋅ L ′ The insert in [ Fig. 5b ] shows the thermal variation of the distance L' as the difference between the respective length variations ΔL 31 and ΔL 32 of the two segments 31 and 32: ΔL'= ΔL 31 - ΔL 32 . The thermal compensation device 3 is effective in keeping the images captured by the sensor 1 sharp when the system temperature varies, i.e. for the athermalization function, if ΔL'=Δf. Two dimensioning parameters of the device 3 are available to satisfy this condition: L' and ΔCTE. When they are chosen in this way, i.e. when the following condition is satisfied: L ′ = β ⋅ f ΔCTE the imaging optical system is said to be athermalized and the preceding condition called the athermalization condition. The thermal range of use of the system as prescribed by its manufacturer, and which has been called the prescription range of the imaging optical system in the general part of this description, is thus widened compared to a system without a thermal compensation device. The thermal compensation device 3 which has just been described is often referred to as a passive device, because it does not use an active component - such as a motor or a drive controlled by the user - to adjust the distance between the imaging matrix 2 and the matrix image sensor 1. The preceding athermalization condition leaves one parameter of the thermal compensation device 3 which is still available to produce an additional function: the difference in values ​​of the coefficient of thermal expansion ΔCTE.This additional function will be to support the vignetting screen 4, preferably at the distal ends 31d and 32d.

[0037] The distance which is thus variable, for the athermalization function, between the matrix image sensor 1 and the imaging matrix 2, as well as the presence of the porthole 1b between these two components, prevent the provision of longitudinal separation walls between the optical paths which would extend from the imaging matrix 2 to the photosensitive surface of the matrix image sensor 1. The function of such separation walls, as known before the present invention, is to eliminate image overflows between neighboring optical paths which produce image overlaps. These image overlaps, also called image superpositions, appear at the image edges which are parallel and neighboring. For example, with reference to [ Fig. 5a ], the image which is formed by the optical path 2 1 thus overflows from the useful area ZU 1 which is assigned to this optical path inside the matrix image sensor 1, into the useful area ZU 3 which is assigned to the optical path 2 3 beyond the boundary between the two useful areas. In the image overlap area, the image information is mixed so that it is no longer usable, in particular for detecting the presence of a gas sought in the portions of the optical input field of the system which are concerned by these image overlaps. It is possible to separate the useful areas of two neighboring optical paths by an intermediate band which is unused in the matrix of the matrix image sensor 1. Such a separation band is denoted BS in [ Fig. 5a ], and has been called guard band in the general part of the present description. Thus, when at least one of the two images formed by the optical paths 2 1 and 2 3 overflows into the separation band BS which is intermediate between the useful zones ZU 1 and ZU 3 , this overflow does not directly encroach into the other of these useful zones, as long as the overflow remains inside the separation band BS. Nevertheless, such separation bands are areas of non-use of the photosensitive surface of the matrix image sensor, and the invention as described below makes it possible to reduce the width of these separation bands BS, and therefore reduce the number of photodetectors (pixels) of the matrix image sensor 1 which are not used.

[0038] According to the invention, the images formed respectively by all the optical paths on the matrix image sensor 1 are limited by the input opening O 4 of the vignetting screen 4, which is common to all these optical paths. Preferably, this input opening O 4 is centered relative to the imaging matrix 2 and to the matrix image sensor 1. The effectiveness of the vignetting screen 4 in suppressing image overlaps is determined by the size of its input opening O 4 , and by the position of the edges of this opening O 4 in front of the imaging matrix 2. When each image is thus limited, the useful areas can be brought closer together, so that the width of each separation strip BS can be reduced. The number of photodetectors of the matrix image sensor 1 which are not used is thus reduced. In other words, this sensor has a utilization rate of its photosensitive surface which is increased.

[0039] Still according to the invention, the thermal compensation device 3 is further used to support the vignetting screen 4. The remaining available parameter of the thermal compensation device 3, or one of its remaining available parameters when it comprises more than two materials with different values ​​of the coefficient of thermal expansion, is then used to adjust the position of the vignetting screen 4 in front of the imaging matrix 2. Such a selection for the parameter ΔCTE, following the athermalization of the imaging optical system which was presented above with reference to [ Fig. 5b ], is now described with reference to [ Fig. 5c ].

[0040] [ Fig. 5c ] is drawn in a longitudinal section plane of the system which is parallel to the rows or columns of the imaging matrix 2, for example again the plane of [ Fig. 5a] et [Fig. 5b ]. It is assumed that the imaging matrix 2 has N optical paths which are juxtaposed in this longitudinal section plane, N being an integer greater than or equal to 2, preferably equal to 2 or 3, generally less than or equal to 4. By adopting the following notations: H det for the size of the matrix image sensor 1 in the longitudinal section plane, I BS for the width of the separation bands BS, all assumed to be equal to each other, and FOV for the useful field of view of the system for all the optical channels in the longitudinal section plane considered, the size of any useful area ZU in this longitudinal section plane is 2·f·tan(FOV / 2), where tan(·) is the tangent function, and: Hdet=N⋅2⋅f⋅tanFOV2+N−1⋅lBS The useful field of view FOV is common to all optical paths, and an image of its contents is formed separately by each optical path in the useful area ZU of the matrix image sensor 1 which is dedicated to this optical path.

[0041] The image overlap is most critical between the first two or last two useful areas of the matrix image sensor 1, counting the useful areas within the longitudinal section plane considered, progressively from one edge of the sensor 1 to its opposite edge. In other words, image overlaps occur first between the first two optical paths or between the (N-1)th< and Nth< optical paths. Denoting FOV ext the extreme field of view of the second optical path, which defines the maximum image overlap with the first optical path without exceeding the intermediate separation band in the useful area of ​​this first optical path, or by symmetry the extreme field of view of the (N-1)th< optical path which defines the maximum image overlap with the Nth< optical path, it comes from [ Fig. 5c ] : L = N − 1 Φ tan FOV ext 2 − tan FOV 2 where L is the distance in front of the imaging matrix 2 at which the vignetting screen 4 is located, and Φ is an individual pupil diameter of the optical paths, assumed to be identical for all optical paths. In an optimized embodiment of the imaging optical system, two optical paths that are neighboring are assumed to be contiguous within the imaging matrix 2, i.e. the respective pupils of the optical paths are contiguous. At the same time, by definition of the extreme field of view FOV ext , the following equation corresponds to the separation band BS between the first two or the last two useful zones ZU within the longitudinal section plane: l BS = f ⋅ tan FOV ext 2 − tan FOV 2

[0042] Eliminating tan(FOV ext / 2) between the two previous equations, and knowing that the pupil diameter Φ is equal to f / F# where F# is the aperture number of the system for each optical path, and using in addition the expression of H det which was given above, it comes: L = N − 1 ⋅ H det − N − 1 ⋅ l BS 2 F # ⋅ l BS ⋅ N 2 ⋅ 4 ⋅ tan 2 FOV 2 This equation, of a purely optical and geometric nature, provides the distance L at which the vignetting screen 4 must be located in front of the imaging matrix 2 to avoid image overlaps within the useful areas of the matrix image sensor 1. It was obtained for the overflow of the image formed by the second optical path into the useful area of ​​the first optical path.

[0043] In preferred embodiments of the invention which reduce the overall size of the system, the vignetting screen 4 may be supported by the distal ends 31d and 32d of the segments 31 and 32 of the thermal compensation device 3 as shown in [ Fig. 5b ], these distal ends 31d and 32d being connected to each other. In other words: L = L'. Then, the remaining available parameter ΔCTE is given by the equation: N − 1 ⋅ H det − N − 1 ⋅ l BS F # ⋅ l BS ⋅ N ⋅ 2 ⋅ tan FOV 2 = 1 ΔCTE ⋅ α − 1 n − 1 ⋅ dn dT Under these conditions, the length of the thermal compensation device 3 in front of the imaging matrix 2 corresponds to the position for the vignetting screen 4 in order to eliminate image overlaps in the useful areas. The imaging optical system with several optical paths then has maximum compactness.

[0044] As an example, the following numerical values ​​can be adopted: the matrix image sensor 1 may have a dimension of 1024 columns by 768 lines, with a row and column pitch which is equal to 17 µm, corresponding to a sensor size of H det =17.4 mm (millimeter) by 13.1 mm; three optical paths may be juxtaposed in the length of this matrix image sensor 1: N=3; each separation strip BS may correspond to 130 adjacent columns of the matrix image sensor 1, corresponding to I BS =2.21 mm; the useful field of view FOV may be equal to 30° (degree) for each optical path; the aperture number F# may be equal to 1.2 also for each optical path; and the two lens matrices 21 and 22 may be made of germanium. For these numerical values, the parameters of the thermal compensation device 3 are L = L' = 40 mm and ΔCTE=20.5 µm·m -1 ·°C -1 . This value for ΔCTE can be obtained using aluminum and the alloy which is commercially designated as Invar ® , separately for each of the two segments 31 and 32 of the thermal compensation device 3. The quotient L / f of the distance L by the focal length f is then equal to 5.6. Under these conditions, the thermal prescription interval can be from -40°C to +60°C. For other embodiments in which the segments 31 and 32 are one made of aluminum and the other of ABS polymer, for acrylonitrile butadiene styrene, the quotient L / f can be equal to 1.8.

[0045] It is repeated that the mathematical model which has just been presented is simplified, and is only intended to show the principle of the invention. Those skilled in the art will know how to improve it to make it more exact, taking into account concrete details of the production of the optical imaging system.

[0046] Furthermore, it is possible to place the vignetting screen 4 at levels in the length of the thermal compensation device 3 which are different from that of the distal ends 31d and 32d. The use of the invention then results in values ​​for the parameter ΔCTE which are different from that which results from the last equation above, but the deduction of these values ​​is within the reach of the person skilled in the art once he has fixed the geometry of the thermal compensation device 3 with respect to the imaging matrix 2 and the vignetting screen 4. Finally, the thermal compensation device 3 may consist of more than two segments whose variations in length combine to produce athermalization, while supporting the vignetting screen 4 at a position suitable for eliminating image overlaps within the useful areas.Additional parameters are then available, which can facilitate the use of the invention for particular applications.

[0047] Finally, it is understood that the invention can be reproduced by modifying still other secondary aspects of the simplified embodiment which has been described in detail above, while retaining at least some of the advantages cited. In particular, an optical imaging system according to the invention can be used in applications other than the search for one or more gases in an environment. Furthermore, all the numerical values ​​which have been cited have been cited only for illustration purposes, and can be changed depending on the application considered.

Claims

1. An optical imaging system comprising several optical pathways (21, 222. arranged in parallel, and for operation when a temperature of the system is included in a prescription interval, said system comprising: - a matrix image sensor (1); and - an imaging matrix (2), comprising several imaging optics which are disposed in parallel and adapted to form simultaneously, in useful zones (ZU) of the matrix image sensor (1) which are dedicated one-to-one to the imaging optics, respective images of a scene contained in an input optical field which is identical for all said imaging optics, the useful zones of the image sensor being disjoint and each imaging optic with the corresponding useful zone belonging to one of the optical pathways of the system separately from each other optical pathway; the optical imaging system being characterised in that it further comprises: - a thermal compensation device (3), disposed to produce a variable spacing between the imaging matrix (2) and the matrix image sensor (1), and comprising, in at least one longitudinal cross-sectional plane which is perpendicular to the imaging matrix and to the matrix image sensor, a meander comprised of two segments (31, 32) each comprising a material whose thermal expansion coefficient value is different from that for the other segment, the two segments being connected to each other by respective distal ends (31d, 32d) of said segments, a proximal end (32p) of one of the segments being connected to the imaging matrix and a proximal end (31p) of the other of the segments being connected to the matrix image sensor, a difference between said thermal expansion coefficient values and respective lengths of the two segments being adapted to produce the variable spacing between the imaging matrix and the matrix image sensor so that images formed in the useful zones (ZU) of said matrix image sensor are sharp when the temperature of the system has any value inside the prescription interval; and - a baffle (4), disposed upstream of the imaging matrix (2) with respect to a direction of propagation of a radiation that enters the system to form images, the baffle having an aperture (O4) for the radiation that is common to the optical pathways (21, 22) of the system, and in that the aperture (O4) of the baffle (4) is dimensioned to laterally limit lighting in each image which is formed by any of the optical pathways, referred to as the optical pathway under consideration, in order to avoid said image overflowing into the useful zone (ZU) of another of the optical pathways, next to said optical pathway under consideration, thus preventing any superimposition of images in all the useful zones, for any value of the temperature of the system in the prescription interval, and further in that the distal ends (31d, 32d) of the segments (31, 32) of the thermal compensation device (3) are located upstream of the imaging matrix (2) with respect to the direction of propagation of the radiation which enters the system to form images, and the baffle (4) is supported by the thermal compensation device (3) upstream of the imaging matrix (2), also with respect to the direction of propagation of the radiation entering the system to form images.

2. The optical imaging system according to claim 1, wherein the baffle (4) is supported by the thermal compensation device (3) at the distal ends (31d, 32d) of the segments (31, 32) of said thermal compensation device.

3. The optical imaging system according to claim 1 or 2, wherein a separation distance between the baffle (4) and the imaging matrix (2), measured in parallel to a common direction of respective optical axes of the imaging optics, is between 1.0 times and 6.0 times a mean value of the focal lengths of the imaging optics.

4. The optical imaging system according to any of the preceding claims, further comprising an enclosure (1a) which is disposed to contain the matrix image sensor (1), said enclosure being provided with a window (1b) which is transparent to the radiation which forms images, and which is located between the imaging matrix (2) and the matrix image sensor.

5. The optical imaging system according to any of the preceding claims, wherein the imaging optics have a row and column arrangement in the imaging matrix (2), each row being perpendicular to each column, and the imaging matrix is oriented such that each row of imaging optics is parallel to a row or column direction of photodetectors of the matrix image sensor (1).

6. The optical imaging system according to claim 5, wherein the row and column arrangement of the imaging optics in the imaging matrix (2) is of one of the following dimensions: 1 x 2, 2 x 1, 2 x 2, 2 x 3, 3 x 2, 3 x 3, 2 x 4, 4 x 2, 3 x 4, 4 x 3 and 4 x 4.

7. The optical imaging system according to any of the preceding claims, wherein each optical imaging system comprises at least one germanium-based lens (211, 212, 221, 222).

8. The optical imaging system according to any of the preceding claims, wherein each imaging optic comprises a spectral filter (231, 232), and the spectral filters of two of the imaging optics that are distinct in the imaging matrix (2) determine spectral transmission windows that are different.

9. A method for capturing several images associated with an input optical field which is identical for all said images, the method being implemented using an optical imaging system which is in accordance with any of the preceding claims.

10. The method according to claim 9, which is intended to reveal presence of a gas inside the input optical field, and wherein the optical imaging system is in accordance with claim 8, the spectral filters (231, 232) being selected such that the spectral transmission window of at least one of the spectral filters is in a transparency band of the gas, and the spectral transmission window of at least one other of said spectral filters is in an absorption band of said gas.

11. The method according to claim 9 or 10, wherein the gas is natural gas, or a hydrocarbon gas, or a toxic gas, or a greenhouse gas.