FR3041458A1 - IMAGING DEVICE, MOTOR VEHICLE EQUIPPED WITH SUCH IMAGING DEVICE, AND ASSOCIATED METHOD - Google Patents

Abstract

An imaging device (100) includes: - an image sensor (130), - a lens, optically associated with the image sensor (130), - a control module (120) designed to trigger a plurality of images taken by the image sensor (130), at successive times, to acquire a respective plurality of images having a given spatial resolution, and - an analysis module (126) designed to produce, by digital combination of acquired images, a super-resolved image having a spatial resolution greater than said given spatial resolution. According to the invention: - said lens is a liquid lens (110) having a deformable diopter (111) whose average orientation is controllable, and - the control module (120) is further designed to control this deformable diopter (111). ) at different average orientations at each of said successive instants. An automobile vehicle (200) equipped with such an imaging device (100), and a related method are also disclosed.

Description

Imaging device, motor vehicle equipped with such an imaging device, and associated method

Technical field to which the invention relates

The present invention generally relates to the field of imaging devices.

It relates more particularly to an imaging device comprising: - an image sensor, - a lens, optically associated with the image sensor, - a control module designed to trigger a plurality of shots by the image sensor at successive instants, so as to acquire a respective plurality of images having a given spatial resolution, and - an analysis module designed to produce a super-resolved image by digitally combining the acquired images, the super-resolved image presenting a spatial resolution greater than said given spatial resolution.

It also relates to a vehicle equipped with such an imaging device, and an associated method.

Technological background

Imaging devices with a digital image sensor are becoming increasingly popular: in addition to the majority of cameras and cameras, these devices are found on most mobile phones, computers and tablets. Many motor vehicles are also provided with such an imaging device to view and analyze the road environment in which the vehicle operates, for example for obstacle detection purposes.

The digital image sensor of such a device has a sensitive, generally rectangular, surface at which an optical system of the imaging device forms an image of an environment facing it. This sensitive surface comprises a plurality of discrete photosensitive elements arranged side-by-side to form a grid covering said surface.

The digital images acquired by this image sensor are thus also formed of a grid composed of several discrete image elements or pixels (according to the Anglo-Saxon denomination). The larger the number of pixels in such an image, the better the resolution. In other words, we can distinguish in this image details even more fine that the number of pixels it has is large. Other factors may also limit the quality of the images obtained with such an imaging system, in particular: measurement noise, electronic, superimposed on the signal acquired by the image sensor, possible defects in the optical system of the imaging device, or else - turbulence of the air located between the imaging device and the area imaged by it.

The resolution of the images obtained with this imaging system can be improved by increasing the number of pixels included in its image sensor. But this solution is expensive, generally increases the acquisition time of an image, and may result, for the images obtained, a drop in the signal-to-noise ratio.

To improve the resolution of these images, it is known moreover, rather than increasing the number of pixels that includes the image sensor: - to acquire successively several images of an area to be viewed, presenting a resolution given spatial configuration, and preferentially presenting, from one image to another, a lateral shift, and then - digitally combining these acquired images so as to obtain a so-called super-resolved image of this zone to be viewed, the spatial resolution of this image super-resolved being greater than said given spatial resolution images acquired initially.

More particularly, document US 2014/0193032 discloses a method for applying this resolution improvement technique to a plurality of images of an environment located at the rear of a vehicle, acquired by a camera embedded in a vehicle. this vehicle. The lateral shift between these images is then caused in particular by vibrations of the vehicle.

Object of the invention

In this context, the present invention provides an imaging device as defined in the introduction, wherein said lens is a liquid lens, having a deformable diopter, the average orientation of the deformable diopter being controllable, and wherein the control is further provided for controlling the deformable diopter of said liquid lens at different average orientations at each of said successive instants.

By thus controlling the average orientation of the deformable diopter of the liquid lens at different orientations at said successive instants, the images formed by this liquid lens are shifted laterally relative to each other on the image sensor audits successive instants.

Thus, thanks to the invention, a lateral shift between said images acquired successively is obtained, in a deterministic and particularly controlled manner, which then facilitates obtaining a super-resolved image from said acquired images, and allows obtain optimal spatial resolution for this super-resolved image.

This imaging device also makes it possible to obtain this superresolved image in a time advantageously short. Indeed, the reaction time of this liquid lens to a change control of the average orientation of its deformable diopter is particularly short, which allows to acquire said plurality of images at a high rate, and thus reduce the time needed to finally get this super-resolved image.

This characteristic is particularly interesting when this imaging device is embedded in a motor vehicle to visualize a road environment in which the vehicle evolves, since the environment viewed is likely in this case to evolve quickly.

In addition, the invention makes it possible to obtain this difference between images acquired successively without moving parts, which thus avoids the phenomenon of wear or seizure and makes this imaging device particularly robust. Other nonlimiting and advantageous features of an imaging device according to the invention are the following: the analysis module is designed to detect an element of an environment facing the imaging device, by analysis of said super-resolved image; said element is an obstacle present on a traffic lane, remote from the imaging device; the image sensor comprises a plurality of pixels separated in pairs by a given inter-pixel distance, and the control module is designed so that two of said distinct mean orientations to which it controls said deformable diopter are sufficiently close to each other; the other so that two images, formed respectively by the liquid lens on the image sensor for each of these two mean orientations, are offset laterally relative to each other by a distance less than said distance between given pixel; the control module is designed to control the deformable diopter of the liquid lens at said distinct average orientations as a function of said given inter-pixel distance; the control module is designed to control the deformable diopter of the liquid lens at said distinct average orientations as a function of a focal length of the liquid lens; the control module is designed so that three of said distinct mean orientations to which it controls the deformable diopter of the liquid lens are non-coplanar; the analysis module is designed to perform said digital combination of the images acquired as a function of data representative of said average orientations distinct from the deformable diopter of the liquid lens; the imaging device further comprises a motion sensor, the control module being designed to control the deformable liquid diopter of said lens at said distinct average orientations as a function, moreover, of data transmitted by said motion sensor, so as to compensate for optically mechanical vibrations experienced by the imaging device; and - the focal length of the liquid lens is controllable, and the control module is further designed to control this focal length to obtain a zoom or focus function. The invention also proposes a motor vehicle comprising an imaging device as described above. The invention also proposes a method implemented in an imaging device, during which: a) a control module triggers a plurality of shots, by an image sensor optically associated with a lens, at times successive, so as to acquire a respective plurality of images having a given spatial resolution, and b) an analysis module produces a super-resolved image by digitally combining the acquired images, the super-resolved image having a higher spatial resolution at said given spatial resolution, the control module further controlling a deformable diopter of said lens, which is a liquid lens, at different average orientations at each of said successive instants.

The optional features presented above for the device can also be applied to such a method.

Detailed description of an example of realization

The following description with reference to the accompanying drawings, given as non-limiting examples, will make it clear what the invention consists of and how it can be achieved.

In the accompanying drawings: FIG. 1 schematically represents a vehicle equipped with an imaging device according to the teachings of the invention, seen from the side; FIG. 2 is a diagrammatic sectional side view of a FIG. 3 is a diagrammatic sectional side view of the liquid lens of FIG. 2, in a second configuration, FIG. 2 is a diagrammatic sectional side view of the liquid lens of FIG. 2, in a third configuration; FIG. 5 is a schematic representation of the image sensor of the vehicle imaging device of FIG. and FIG. 6 schematically represents a portion of an image obtained from an image acquired by the image sensor of FIG. 5.

In Figure 1, we can see the main elements of an imaging device 100, according to the teachings of the invention. It comprises in particular: an image sensor 130, and a liquid lens 110, that is to say a deformable lens containing a fluid, optically associated with the image sensor 130.

More particularly, the imaging device 100 here comprises a video camera integrating this image sensor 130 and this liquid lens 110.

This liquid lens 110, used alone as is the case here, or integrated into an optical system such as an imaging lens comprising several lenses, makes it possible to form an image of an environment E facing it on a sensitive surface of the image sensor 130.

The imaging device 100 is embedded here in a vehicle 200 automobile, to allow viewing of a road environment in which the vehicle 200 evolves.

More specifically, it is disposed here inside and in front of this vehicle 200, behind a protective window 211. The environment E facing said liquid lens 110 thus substantially corresponds to the road environment in which the vehicle is moving. vehicle 200 and facing the latter.

The imaging device 100 can thus be used for obstacle detection purposes, in particular obstacles facing the vehicle 200, as detailed below.

Alternatively, the imaging device is installed in the rear part, or on one side of the vehicle, so as to fulfill the role of an observation camera of a rear or side environment of the motor vehicle.

According to another variant, the imaging device is fixed in a passenger compartment of the motor vehicle.

According to yet another variant, the imaging device is installed on or in a building, or in a public place such as a street, instead of being embedded in a motor vehicle.

The image sensor 130 of the imaging device 100 is here a two-dimensional digital image sensor, for example a CCD image sensor (English acronym for "Charge-Coupled Device"), ie device charge transfer) or CMOS (acronym for "Complementary Metal-Oxide Semiconductor", that is to say semiconductor complementary metal oxide).

The image sensor 130 here has a sensitive, rectangular surface at which the liquid lens 110 of the imaging device 100 forms an image of the environment E facing it. This sensitive surface comprises a plurality of discrete photosensitive elements arranged side-by-side to form a grid covering said surface (FIG. 5).

Each of these discrete photosensitive elements is referred to below as the PC pixel, as is a discrete element of a (digital) image acquired by this image sensor.

For an image in digital form, for example for the image acquired by the image sensor 130, the expression "spatial resolution" here designates the number of pixels that it comprises.

This image here comprises an Npix number of pixels PI equal to or less than the number of PC pixels of the image sensor 130, and thus has a given spatial resolution. The term "spatial resolution" could also refer to the inverse of the smallest angular distance between two details of the environment facing the device that can be distinguished from each other in an image acquired by the sensor. images of the imaging device.

Two neighboring PC pixels of the image sensor 130 are here separated by a given inter-pixel distance d. This inter-pixel distance d corresponds to the distance between the respective centers of two neighboring PC pixels (FIG. 5).

The position of a point on the sensitive surface of the image sensor 130 is located in a system of two orthogonal axes X and Y extending parallel to this sensitive surface. The X and Y axes are shown in each of Figures 1 to 5 for ease of reading.

The liquid lens 110 is schematically shown in section and from the side in FIGS. 2 to 4.

The liquid lens 110 has a deformable diopter 111. Here it separates two liquids with different optical indices, in this case an aqueous solution 113 and an oil 114.

The shape of this deformable diopter 111 is characterized in particular by: a radius of curvature, and a mean orientation of the deformable diopter 111.

This average orientation is indicated here by a mean axis 112 (FIGS. 2 to 4), which extends perpendicularly to an average plane defined by the deformable diopter 111.

The shape of the deformable diopter 111 is controllable, in this case by means of a control signal SC (FIG. 1), here comprising a plurality of electrical control voltages applied between electrodes 115, 116, 117 of the liquid lens 110.

More specifically, the radius of curvature of the deformable diopter 111 and the direction of the mean axis 112 that it defines are controllable by means of said electrical voltages.

The liquid lens 110, when it is not electrically powered, has a Z axis of symmetry per revolution, constituting its optical Z axis at rest. The mean axis 112 defined by the deformable diopter 111 when the liquid lens 110 is not powered then coincides with this axis Z.

Here, the liquid lens 110 is arranged so that its optical axis Z at rest is substantially orthogonal to the sensitive surface of the image sensor 130, and thus to the X and Y axes.

Controlling (via said electrical control voltages) the radius of curvature of the deformable diopter 111 makes it possible to adjust the focal length f of the liquid lens 110, for example for focusing or zooming purposes.

Controlling the direction of the mean axis 112 defined by this deformable diopter 111 adjusts the average lateral position at which an image is formed by the liquid lens 110 on the image sensor 130, as detailed below.

In FIG. 3, the liquid lens 110 is controlled so that the mean axis 112 defined by its deformable diopter 111 coincides with the optical axis Z of the liquid lens at rest. This is achieved here by applying the same electrical control voltage between, on the one hand, the electrode 116 and the electrode 115, and on the other hand, between the electrode 117 and the electrode 115.

The liquid lens 110 then plays the role of a convergent lens, symmetrical by revolution about the optical Z axis of the liquid lens at rest.

In FIGS. 2 and 4, the liquid lens 110 is controlled on the contrary so that the mean axis 112 defined by its deformable diopter 111 is angularly offset with respect to the optical Z axis of the liquid lens at rest. This is achieved for example by applying different control voltages between, on the one hand, the electrode 116 and the electrode 115, and on the other hand, between the electrode 117 and the electrode 115.

The liquid lens 110 then fulfills both the function of a converging lens, because of the curvature of this deformable diopter 111, and the function of a prism deviating on average the light rays that pass through it, because of the inclination of this deformable diopter 111 relative to other dioptres 118, 119 of the liquid lens 110.

The emerging light rays Re of the liquid lens 110 are then more convergent, and have on average an angular deviation with respect to the incident light rays Ri thereon.

This angular deflection introduced by the liquid lens 110 is here even greater than the angle between the mean axis 112 defined by its deformable diopter 111 and the optical Z axis of the liquid lens at rest is large.

Due to this angular deviation, the average position of the image formed by the liquid lens 110 on the image sensor 130 is shifted laterally, that is to say shifted in the plane (X, Y) defined by the sensitive surface of the image sensor 130, with respect to the image obtained on this sensitive surface when the deformable diopter 111 is centered on the optical Z axis of the liquid lens at rest (case of FIG. 3). The average position of this image, with respect to its position when the deformable diopter 111 is centered on the optical Z axis of the liquid lens at rest, is indicated here, in the system of axes X, Y, by a vector ( Xm, Ym) (see Figures 2, 4 and 5). By way of illustration, FIG. 5 shows: - in dotted lines, an IMD image formed on the image sensor 130 when the mean axis 112 defined by the deformable diopter 111 is angularly offset with respect to the axis Z optics of the liquid lens at rest, and - in solid lines, the image ΙΜ0 which would have formed on the image sensor 130 if this mean axis 112 defined by the deformable diopter 111 coincided with the optical axis Z of the liquid lens at rest (case of Figure 3 for example).

As can be seen, the orientation at which said mean axis 112 is controlled is different in the case of Figure 4, and in the case of Figure 2. For illustrative purposes, an image formed by the liquid lens 110 on the image sensor 130 is shifted here parallel to the axis X when the liquid lens 110 passes from the configuration of FIG. 2 to that of FIG. 3 and then that of FIG.

Optionally, the liquid lens 110 is designed so that its deformable diopter 111 is controllable to at least three distinct non-coplanar average orientations. The three images formed respectively on the image sensor 130 for these three mean orientations are then offset laterally relative to each other in two different directions, for example in the X direction for two of these images, and in the Y direction for a third of these images.

In this case, the orientation of said middle axis 112 could for example be controlled so as to angularly shift it: on the one hand, with respect to the plane (Y, Z), of a first angle a, as on FIGS. 2 to 4, for example, in order to shift an image formed on the image sensor 130 parallel to the axis X, and - secondly, with respect to the plane (Z, X) (this case not is not shown in the figures), a second angle β, so as to shift the image formed on the image sensor 130 parallel to the Y axis.

Here, the orientation of said mean axis 112 is controlled only so as to angularly shift it with respect to the plane (Y, Z) of the first angle a, so as to shift an image formed on the image sensor 130 parallel to the X axis.

The average position of an image formed by the liquid lens 110 on the image sensor 130, represented here by the vector (Xm, Ym), depends in particular on: - the average orientation of its deformable diopter 111, identified for example by the first angle α (and possibly the second angle β) which marks (nt) the direction of the mean axis 112 perpendicular to this diopter, and - the distance between the liquid lens 110 and the image sensor 130, and therefore the focal length f of the liquid lens 110.

The liquid lens 110 used here is for example a liquid lens of the "Baltic 167" model of the company Varioptic.

The liquid lens 110 comprises here a deformable diopter 111. In a variant, the liquid lens 110 may comprise several deformable diopters, the shape of each of these diopters being controllable so as to obtain in particular the effects described above.

The imaging device 100 also comprises: a control module 120 adapted in particular to control the image sensor 130 and the liquid lens 110, and an analysis module 126 designed in particular for processing the images acquired by the image sensor. 130 (Figure 1).

The control module 120 here comprises: a processor 121 performing logical operations, for example a microprocessor, a storage module 122, with which the processor 121 can exchange data, the aforementioned analysis module 126, comprising here the processor 121 and the storage module 122, and a conditioning module 123, for example a digital-analog converter followed by one or more voltage amplifiers, which makes it possible to convert one or more CMD commands given by the processor 121. in digital form, so as to obtain the control signal SC adapted to control the liquid lens 110 mentioned above.

The processor 121 is adapted to control the image sensor 130, so that the latter acquires data representative of an image formed at its sensitive surface, and is adapted to receive from the image sensor 130 said data.

The processor 121 is also adapted to control the focal length f of the liquid lens 110 and the average orientation of the deformable diopter 111 thereof, via the conditioning module 123.

Optionally, the imaging device 100 also includes a motion sensor 124, such as an accelerometer and / or a gyrometer, which transmits to the processor 121 data representative of the signals it measures.

The imaging device 100 is designed to obtain a super-resolved image, having a spatial resolution greater than said given spatial resolution of one of the images acquired by the image sensor 130, and this through the implementation of in this imaging device 100, of the method described below.

During this process, a1) the control module 120 controls the deformable diopter 111 of said liquid lens 110 at different average orientations at successive times, a2) the control module 120 triggers a plurality of shots by the sensor 130 at each of said instants, so as to acquire a respective plurality of images IM-ι, IM2, ... IM1, ... IMn having a given spatial resolution, and b) the analysis module 126 produced a super-resolved IMS image by digital combination of IM-ι, IM2, ... IM1, ... IMn images acquired, the super-resolved IMS image having a spatial resolution greater than said given spatial resolution.

More specifically, during step a1), the control module 120 controls here the deformable diopter 111 of said liquid lens 110 so that any two of said successive distinct average orientations are sufficiently close to each other for the two images respectively formed by the liquid lens 110 on the image sensor 130 for each of these two mean orientations of the deformable diopter 111 are offset laterally relative to each other by a distance less than said distance between given pixel d, which for example results in the conditions (F1) and (F2) below: | Xmi + 1 - Xrrii | <d (F1) | Ymi + 1-Ymi | <d (F2) where the vector (Xm ^ Ymi) locates the average position, on the image sensor 130, of the image IM, acquired by the latter, by relative to the average position at which this image would form if the deformable diopter 111 was centered on the optical Z axis of the liquid lens 110 at rest.

Since this average position (XmpYmj) on the image sensor 130 depends on the focal length f of the liquid lens 110, thereby obtaining a subpixel shift between the average positions of two of said IM1 images; IM2, ... ΙΜ |, ... ΙΜν successively acquired, the control module 120 controls here the average orientation of the deformable diopter 111 taking into account both: - the inter-pixel distance d given between two pixels of the 130, and the focal length f of the liquid lens 110,

More specifically, the control module 120 here controls the first angle mentioned above at successive values at each of said instants, these successive values being determined as a function, inter alia, of said inter-pixel distance d and of the focal length f of the lens. liquid 110, so as to ensure that the condition (F1) is satisfied (the condition (F2) is also verified, since here the coordinate Ym, does not vary in the example described here of an image acquired to the next) .

Optionally, the control module 120 could control the deformable diopter 111 at said different mean orientations so as to obtain at the same time: for at least part of the images IM 1, IM 2,... IM 1, IM 1 acquired, an offset Xmi + 1 - Xm, along the non-zero direction X, and - for at least a portion of the images IM-ι, IM2, ... IM ,, ... IMn acquired, an offset Ymi + 1 - Ym, in the nonzero Y direction.

Optionally, the control module 120 controls the deformable diopter 111 at said distinct average orientations, furthermore according to data transmitted by the motion sensor 124 and representative of mechanical vibrations experienced by the imaging device 100, so as to compensate optically for the effect of these vibrations.

This arrangement is particularly advantageous when the imaging device 100 is embedded in a motor vehicle 200, as is the case here. Indeed, in such a case, the imaging device 100 can undergo significant vibrations that may degrade the quality of the images obtained or cause lateral shifts of these uncontrolled images.

Here, to produce said super-resolved image IMS, the analysis module 126 combines these IM-ι, IM2, ... IMj ..., IMn images acquired by the image sensor 130 according to the following method: 1) for each image IM1, IM2, ... IMj ..., acquired IMN, a recentered image IMR-ι, IMR2, ... IMR1 ... IMRn is obtained numerically, obtained by digitally shifting (laterally) this image acquired IM 1, IM2, ... IM, ..., IMn so that each of the recentered images IMR-ι, IMR2, ... IMRj ..., IMRn is centered on the same given common position; 2) the imaged image set IMRi, IMR2, ... IMRj ..., IMRn is associated with an oversampled IMSE image comprising a number of pixels SP greater than the number of pixels PI of any of the images. recentered IMR-ι, IMR2, ... IMRj ..., IMRn; each pixel PI of any of these recentered images IMR-ι, IMR2, ... IMRj ..., IMRn is interpreted as the result obtained by applying to several pixels SP of the IMSE oversampled image a function of observation F1, F2, FN (unknown at this stage but determined by optimization in the next step), associated with this refocused image; and 3) said super-resolved IMS image is determined by inverting the observation functions F1, F2, ... F | ..., FN by an optimization method.

More precisely, in step 1), each image IM 1, IM 2,... IM,..., IMn acquired by an amount equal to the opposite of the average position (Xmi, Ymi) of FIG. this image during its acquisition by the image sensor 130. All of these images are thus refocused on the same average position, in this case the position (0,0) of the plane (X, Y).

The average position (Xmi, Ymi) at which an image acquired by the image sensor 130 has been formed is accessible to the analysis module 126 since this position has been controlled by the control module 120, via the orientation of the deformable diopter 111, during step a1) described above.

The digital offset of this acquired image is achieved with a sub-pixel accuracy, that is to say with a better accuracy than that corresponding to the difference between two pixels PI of this image. To obtain this precision, this digital shift is made for example on the basis of an interpolation of the acquired image.

In a particularly interesting variant in practice of this step 1), the acquired N images ΙΜι, IM2, ... ΙΜ, ..., IMN are recentered by taking as position reference the average position (Xitin, Yitin) of the last of these images (instead of taking the position (0,0) as a position reference).

Here, in step 3), the super-resolved image IMS is obtained by interpolation of the data representative of the recentered images IMR1; IMR2, ... IMR, ..., IMRN.

In the example illustrated in FIG. 6, the IMSE oversampled image, like the IMS super-resolved image, comprises 4 (sub) SP pixels for each given PI pixel of the recentered images IMRi, IMR2, ... IMR, ..., IMRn. This oversampled IMSE image, and this IMS super-resolved image, thus each comprise, in this example, four times more pixels than an image acquired by the image sensor 130 (more precisely, they each comprise 4xNpix pixels). The super-resolved image IMS obtained in step 3) has a resolution four times greater than that of such an image IMi, IM2, ... IM, ..., IMn.

With respect to one of the IM-ι, IM2, ... IMj..IMN images acquired by the image sensor 130, the super-resolved IMS image thus obtained makes it possible to distinguish finer details from the environment. E facing the imaging device 100.

Alternatively, instead of being obtained by the method described above, the IMS super-resolved image is obtained by any resolution increase method known to those skilled in the art, for example by modifying the phase and numerical combination of Fourier Transforms acquired images.

The imaging device 100 is furthermore designed to detect an element of the environment E facing it, by analyzing said super-solute IMS image.

Using this IMS super-resolved image, rather than directly using one of the images acquired by the image sensor, advantageously makes it possible to improve the detection efficiency of this element.

Here, the imaging device 100 is designed more specifically to detect an element of this environment E corresponding to a remote obstacle present on a roadway. Detecting such an obstacle makes it possible in particular to: - monitor the evolution of a road traffic, - detect an obstacle, for example a motor vehicle, present at an intersection between traffic lanes, or approaching such an intersection or to detect a remote obstacle that the automobile vehicle 200 equipped with the imaging device 100 is likely to cross.

For these different applications, it is preferable that the imaging device 100 has a very wide field of view angularly.

A very wide angular field of view is indeed necessary to be able to detect obstacles in one of the crossing situations mentioned above. This obstacle can indeed be strongly offset laterally with respect to the mean axis (Z) of observation by the imaging device 100, for example because it is a vehicle traveling on a road perpendicular to that taken by the vehicle 200 equipped with the imaging device, these two roads intersecting at a great distance in front of the vehicle 200, or because this obstacle is located on a portion of road in front of the vehicle 200, but after a turn.

A very wide field of view is obtained for the imaging device 100, for example because the liquid lens 110 has small focal lengths, thus constituting a "wide-angle" objective.

Because of the angular width of the field of view of the imaging device 100, an obstacle remote from it then occupies a particularly small area within an image that it acquires. It is therefore particularly advantageous, as is the case here, to improve the resolution of the images of said environment E obtained by means of the imaging device 100. The detection of this obstacle, and of its possible movements, can then be ensured efficiently. by analyzing one or more of these super-resolved images IMS, despite the small size that this obstacle may present within an image IM-ι, IM2, ... IM, ..., IMN acquired by the sensor 130 of the device.

The detection of this obstacle by analyzing one or more super-resolved images IMS is for example carried out by recognizing, in this image, a shape corresponding to a predefined obstacle shape (this predefined form being for example pre-recorded in FIG. storage module 122 of the analysis module 126), or by motion detection between two successive IMS super-resolved images by means, for example, of an optical flow calculation between these two images, or by a method based on a network of artificial neurons configured during a preliminary learning phase.

Claims (12)

An imaging device (100) comprising: - an image sensor (130); - a lens optically associated with the image sensor (130); - a control module (120) adapted to trigger a plurality of images; images taken by the image sensor (130) at successive instants, so as to acquire a respective plurality of images (ΙΜ-ι, IM2, ... ΙΜ, ..., IMn) having a given spatial resolution , and - an analysis module (126) designed to produce a super-solved image (IMS) by digital combination of the acquired images (ΙΜ-ι, IM2, ... IM, ..., IMn), the superimposed image solution (IMS) having a spatial resolution greater than said given spatial resolution, characterized in that: said lens is a liquid lens (110) having a deformable diopter (111), the average orientation of said deformable diopter being controllable, and in that - the control module (120) is furthermore designed to control the deformable diopter (1 11) of said liquid lens (110) at distinct average orientations at each of said successive instants. An imaging device (100) according to claim 1, wherein the analysis module (126) is adapted to detect an element of an environment (E) facing the imaging device (100) by analysis. of said superresolute image (IMS). An imaging device (100) according to claim 2, wherein said element is an obstacle present on a traffic lane, remote from the imaging device (100). 4. An imaging device (100) according to one of claims 1 to 3, wherein: - the image sensor (130) comprises a plurality of pixels (PC) separated in pairs by an inter-pixel distance ( d) given, and wherein - the control module (120) is designed so that two of said distinct mean orientations to which it controls said deformable diopter (111) are sufficiently close to each other so that two images, formed respectively by the liquid lens (110) on the image sensor (130) for each of these two mean orientations, are offset laterally with respect to each other by a distance (| Xmi + 1 - Xrrijl, lYm ^ -Ymil) less than said given inter-pixel distance (d). An imaging device (100) according to claim 4, wherein the control module (120) is adapted to control the deformable diopter (111) of the liquid lens (110) at said distinct average orientations according to said inter-distance. -pixel (d) given. An imaging device (100) according to one of claims 1 to 5, wherein the control module (120) is adapted to control the deformable diopter (111) of the liquid lens (110) at said distinct average orientations in function of a focal length (f) of the liquid lens (110). The imaging device (100) according to one of claims 1 to 6, wherein the control module (120) is adapted to have three of said distinct average orientations to which it controls the deformable diopter (111) of the liquid lens. (110) are non-coplanar. An imaging device (100) according to one of claims 1 to 7, wherein the analysis module (126) is adapted to perform said digital combination of images (IM1, IM2, ... ΙΜ, .. ., IMN) acquired according to data representative of said distinct average orientations of the deformable diopter (111) of the liquid lens (110). The imaging device (100) according to one of claims 1 to 8, further comprising a motion sensor (124), and wherein the control module (120) is adapted to control the deformable diopter (111). said liquid lens (110) at said distinct average orientations further depending on data transmitted by said motion sensor (124) so as to optically compensate for mechanical vibrations experienced by the imaging device (100). An imaging device (100) according to one of claims 1 to 9, wherein the focal length (f) of the liquid lens (110) is controllable, and wherein the control module (120) is adapted to control this focal length (f) to obtain a zoom or focus function. A motor vehicle (200) comprising an imaging device (100) according to one of claims 1 to 10. A method implemented in an imaging device (100), wherein: a) a control module (120) triggers a plurality of shots, by an image sensor (130) optically associated with a lens, at successive instants, so as to acquire a respective plurality of images (IM-i, IM2, ... IM, ..., IMn) having a given spatial resolution, and b) an analysis module ( 126) produces a super-resolved image (IMS) by digital combination of the acquired images (IMi, IM2, ... IM, ..., IMn), the super-resolved image (IMS) having a spatial resolution greater than said given spatial resolution, characterized in that: - the control module (120) controls a deformable diopter (111) of said lens, which is a liquid lens (110), at different average orientations at each of said successive instants.