METHOD FOR CORRECTING OPTICAL ABERRATIONS INTRODUCED INTO AN IMAGE BY AN OPTICAL LENS, DEVICE AND SYSTEM FOR PERFORMING THIS METHOD

Description

[0001] The present invention relates to a method for correcting optical aberrations introduced by an optical lens in an image captured with said optical lens. It also relates to a system and an apparatus implementing such a method.

[0002] The field of the invention is the field of correction of images captured with an optical lens, with a view to correcting optical aberrations due to said optical lens. État de la technique

[0003] Optical lenses are used in various devices, such as cameras, camcorders, smartphones, etc., to image a scene.

[0004] Generally, an optical lens consists of a stack of optical elements, such as optical lenses, separated from each other by an air gap, also called an "air gap," or by a spacer. They are usually assembled using a device called a barrel or tube.

[0005] To create an image, the optical lens works in conjunction with a photosensitive sensor, such as a CMOS or CCD sensor. The central plane of the image sensor is called the image plane, and the sensor contains a multitude of pixels. The assembly comprising the optical lens and the image sensor is generally called the "optical module" or "camera module."

[0006] The trend toward miniaturization of camera modules used in electronic devices, such as smartphones, reduces tolerances in component manufacturing, particularly for the optical lens and its components. Manufacturing defects can therefore appear as flaws in the image captured by the image sensor. For example, optical aberrations can result in blurry images. The brightness of a pixel in the original image is distorted, altering the brightness of an adjacent pixel in the captured image.

[0007] One objective of the present invention is to remedy at least one of the aforementioned drawbacks.

[0008] Another object of the invention is to provide a solution for correcting optical aberrations introduced by an optical lens in an image captured by said lens. US patent 2011 / 242372 A1 is relevant to this application. Exposé de l'invention

[0009] The invention proposes to achieve at least one of the aforementioned goals by means of an image acquisition method with a device comprising a camera module having an optical lens associated with an image sensor, said method comprising at least one iteration of an image acquisition phase by said device comprising the following steps: capturing an image with said camera module; and correcting said captured image according to at least one correction matrix calculated, directly or indirectly, from at least one matrix, called an aberration matrix, determined for said optical lens, and representative of optical aberrations introduced by said optical lens into an image.

[0010] Thus, the invention proposes to perform image correction on an image captured with an optical lens within the device itself, which contains said optical lens. In this way, the image correction can be adapted to each device individually and can be performed on the fly at the moment the image is captured by the device. The solution proposed by the present invention is therefore more customizable and more scalable.

[0011] To achieve this, the invention proposes using a correction matrix calculated based on at least one aberration matrix previously determined for the optical objective used. The correction matrix can be determined directly from the aberration matrix, or indirectly, for example from a previously calculated aberration kernel matrix or correction kernel matrix, as will be described later.

[0012] Optical aberration refers to optical blur or optical distortion. Optical blur generally results in a spreading of a point of light. Distortion generally results in a shift of the optical point.

[0013] The optical elements that make up an optical lens are stacked along a stacking direction, also called the Z-axis hereafter, or the axis of the optical lens. The plane perpendicular to the Z-axis, that is, the plane along which each optical element extends, is called the XY plane hereafter.

[0014] By "geometric parameter of an optical interface", we mean, for example, and without loss of generality: a position of the optical interface within the lens, in the Z axis; a position of an APEX of the optical interface, in particular in the XY plane and / or a position of an APEX of the optical interface, in particular along the Z axis, an inclination (TIP and / or TILT) of said optical interface with respect to the Z axis, a decentering of an optical interface, or of an optical element, with respect to the Z axis, in the XY plane.

[0015] In this application, the term "buried optical interface" of an optical lens means an interface within the optical lens that is visible, or accessible, only through at least one other optical interface of the lens. The at least one other optical interface through which the buried interface is visible may be an optical interface of the same optical element, or an optical interface of a different optical element than the buried interface.

[0016] By "pixel" we mean an elementary pattern of an image containing for example color values ​​R, G, B, or brightness and chrominance, therefore an element that forms a 'point' of an image.

[0017] In this document, a region of the image sensor centered on the coordinates (X i ,Y i ), in the (X,Y) plane, can be designated by R(X i ,Y i ), or R i .

[0018] In this document, MA denotes the aberration matrix for the entire optical lens. MAi, or MAi(Xi, Yi), denotes the aberration matrix for the region Ri of the lens. The aberration matrix MA can be obtained by concatenating or adding the aberration matrices MAi for all regions of the image sensor.

[0019] The correction matrix, or each correction matrix, is designated by the letters MC.

[0020] The image sensor can be any type of photosensitive sensor, such as a CMOS sensor or a CCD sensor, etc.

[0021] According to particularly advantageous embodiments, the image acquisition phase may include a step of determining a distance, denoted DOC, between the optical lens and the image sensor, at least one correction matrix being a function of said DOC distance.

[0022] Indeed, the method according to the invention makes it possible to take into account the optical aberrations introduced by the optical lens in an image when the distance between the optical lens and the image plane, i.e., the image sensor plane, is changing, and to correct these aberrations with correction matrices determined as a function of this distance. In fact, the inventors observed that, for a given lens, the optical aberrations introduced by said lens can vary depending on the distance between said lens and the image plane.

[0023] The image plane can be the image sensor plane.

[0024] Thus, for each DOC distance j, with j ≥ 2, an aberration matrix MA i corresponding to the entire image sensor can be determined. Alternatively, for each DOC distance j, several aberration matrices MA ij can be determined for each region of the image sensor. The aberration matrix MA j, or each individual MA ij aberration matrix, can then be used to directly or indirectly derive a correction matrix MC j for the entire sensor, or correction matrices MC ij for each region R i of the image sensor.

[0025] Each DOC distance can, for example, be measured by a distance sensor, such as an optical sensor, a magnetic sensor, a capacitive sensor, etc.

[0026] The, each, DOC distance can for example be calculated from information provided by a focus adjustment mechanism modifying, or controlling, the distance between the optical lens and the image sensor.

[0027] In particularly advantageous embodiments, the image acquisition phase may include a step of determining a distance, DOS, between the optical lens and the scene, with at least one correction matrix being a function of said DOS distance.

[0028] Thus, the method according to the invention makes it possible to take into account the optical aberrations introduced by the optical lens in an image when the distance between the optical lens and the imaged scene changes. Indeed, the inventors observed that, for a given lens, the optical aberrations introduced by said lens can vary depending on the distance between said lens and the scene, or each part of the scene.

[0029] Thus, for each DOS distance k, with k ≥ 2, an aberration matrix MA k corresponding to the entire image sensor can be determined. Alternatively, for each DOS distance k, several aberration matrices MA ik can be determined for each region of the image sensor. The aberration matrix MA k, or each individual MA ik aberration matrix, can then be used to directly or indirectly derive a correction matrix MC k for the entire sensor, or correction matrices MC ik for each region R i of the image sensor.

[0030] The distance, or each distance, DOS can for example be measured by a LIDAR, a time-of-flight camera, an ultrasonic sensor, textured image analysis, etc.

[0031] Only one DOS distance can be measured for the optical lens. Alternatively, DOS distances can be measured for different regions Ri of the optical sensor for the same scene. This is because an imaged scene may contain several objects located at different DOS distances. In this case, a DOS distance ik is determined for each region Ri of the sensor, and the correction matrix is ​​selected individually for each region Ri based on the DOS distance ik measured for that region Ri of the sensor.

[0032] Depending on the embodiment, an image can be corrected according to several correction matrices, each selected according to: of a region R i of the image sensor, the distance DOCj between the image sensor and the lens, and the distance DOS k between the optical lens and the scene in said region R i. This correction matrix can be denoted MC ijk .

[0033] This involves determining, for each region R i of the sensor, several aberration matrices MA ijk, each corresponding to a DOC distance j and a DOS distance k in said region R i, with: i=1,...,I ​​and I≥2, j=1,...,1 and J≥2 k=1,...,I ​​and K≥2

[0034] For example, by taking ten (10) regions of the image sensor, 3 DOC distances j, and 5 DOS distances k for each region of the image sensor, then 15 aberration matrices can be determined for each region of the image sensor, and in all 150 aberration matrices for the image sensor.

[0035] In the following, for region R i of the sensor, the DOC distance j and the DOS distance k: MA ijk is the aberration matrix, and MC ijk is the correction matrix. Of course, when the DOC distance and / or the DOS distance are not variable, then the MA notations ijk and MC ijk do not necessarily imply that these distances are variable.

[0036] According to some embodiments, the method according to the invention may include a step of calculating at least one correction matrix outside the imaging device. In this case, the correction matrix may in some cases be stored in said imaging device.

[0037] Alternatively, the method according to the invention may include a step of calculating at least one correction matrix within said apparatus.

[0038] In this case, the calculation of the correction matrix used for image correction is performed by a computing unit of the imaging device.

[0039] According to some embodiments, the step of calculating at least one correction matrix can be carried out during the acquisition phase so that said at least correction matrix is ​​calculated on the fly for each image captured.

[0040] Thus, for each image acquisition, at least one correction matrix is ​​calculated, taking into account the conditions under which the image is acquired, so that the image correction is customized for each image. The image correction is therefore more precise. Furthermore, in this embodiment, it is possible to calculate only the correction matrices used for correcting the acquired image. For example, it is possible to calculate only the correction matrices for the DOC distance and / or the DOS distance(s) determined for that image.

[0041] According to one embodiment, the step of calculating at least one correction matrix can be carried out prior to the acquisition phase so that said calculation step is common to several iterations of the image acquisition phase.

[0042] In this case, image correction can be performed more quickly because it is not necessary to calculate at least one correction matrix on the fly. In this embodiment, all correction matrices corresponding to the configurations, and in particular to the DOC and DOS distances likely to be used for the images to be corrected, can be calculated and stored within the device in a database. Then, at each iteration of the acquisition phase, at least one correction matrix corresponding to the image acquisition configuration during that acquisition phase can be selected, specifically based on the DOC distance and / or at least one DOS distance, to perform the correction of the image acquired during that acquisition phase.

[0043] At least one correction matrix can be calculated from at least one aberration matrix.

[0044] In particular, each correction matrix can be calculated from an aberration matrix. For example, the correction matrix MC ijk can be calculated from the aberration matrix MA ijk.

[0045] At least one correction matrix can be calculated by inverting the corresponding aberration matrix, which can be done for example by calculations in the spatial frequency domain.

[0046] Based on implementation examples, at least one correction matrix can be obtained using the following relationship: MC ijk = TF − 1 TF G 0 TF MA ijk where G0 is a function representing the shape of the point dispersion function to be obtained after correction. This can be a function approximating a Dirac delta function in 2 dimensions, i.e., equal to 1 at the center and practically 0 around it. Denoting MTOBSijk as the matrix of a test light beam as observed for the measurement of an aberration matrix, and MT as its native shape as it actually is during emission, the correction matrix MCijk can be calculated as follows: MC ijk = TF − 1 TF G 0 . TF MT ijk TF MTOBS ijk In a simplified form, it is possible to assume that TF(G0) = 1, the constant function that returns 1 with zero phase at all frequencies, which can be written: MC ijk = TF − 1 TF MT ijk TF MTOBS ijk with : TF the Fourier transform operator (in 2D); TF-1< the inverse Fourier transform operator (in 2D); MC ijk the correction matrix for the image sensor region R i, at the lens-sensor distance DOC j, and at the lens-scene distance DOS k for said region R i; MA ijk the aberration matrix obtained for the image sensor region R i and the distances DOC j and DOS k; and MT ijk the test light beam matrix, i.e. the matrix describing the test light beam used to measure the aberration matrix MA ijk; and MTOBS ijk the illumination matrix observed in the sensor plane with respect to the test matrix MT ijk presented at the input of the optical lens.

[0047] At least one correction matrix can be calculated from at least one kernel aberration matrix, including coefficients allowing the said at least one aberration matrix to be deduced, and previously calculated.

[0048] Using a kernel aberration matrix avoids having to store all the aberration matrices used to determine the correction matrices, thus reducing the memory space used to store the aberration matrices.

[0049] In this case, several aberration matrices can be determined before the correction phase. Then, from these matrices, a common matrix can be identified. This common matrix, called the kernel aberration matrix, can be stored in place of the multitude of aberration matrices, linked to one or more relationships, to allow the calculation of each aberration matrix. This last matrix is ​​then used to determine a correction matrix as described later.

[0050] The kernel aberration matrix can, following a non-limiting example of implementation, be determined in the following way, from the aberration matrices.

[0051] Initially, a basis of FA functions is established to describe the values ​​of the coefficients of the aberration matrices as a function of the (X,Y) positions in the image sensor field, such that: MA X Y = ∑ l αA l X Y FA l X Y + E X Y , Or I sweeps the basis of the FA functions, with 1≤I≤L, and L≥1. The term E(X,Y) models the differences between the combination of the FA I and MA functions.

[0052] The basis FA can preferably be chosen to be sufficiently large so that the term E(X,Y) has a negligible, or even zero, effect. These functions may be orthogonal to each other in the sense of an inner product, but not necessarily. For example, the basis of functions FA I can be the basis of functions of Zernike polynomials. To obtain the functions αA I (X,Y), it suffices to use a known state-of-the-art method such as a projection of MA(X,Y) onto the FA I (X,Y), in the sense of an inner product, and to apply the appropriate matrix multiplication to account for their non-orthogonality, if necessary, to obtain the functions αA I (X,Y). Establishing a precursor aberration kernel (PNA) amounts to searching for a parametric model of the functions αA I (X,Y), such that for example a polynomial expression represents αA I (X,Y), for example, if I ∈ {1, 2}, such that αA 1 (X,Y) = a 01 .((Xx 01 ) 2< +(Yy 01 ) 2< ) and αA 2 (X,Y) = a 02 .((Xx 02 )*(Yy 02 )) +z 02 . The model representing the αA I (X,Y), here αA 1 (X,Y) and αA 2 (X,Y), is then the table of values ​​((a 01 , x 01 , y 01 ) ; (a 02 , x 02 , y 02 , z 02 )) in this example. We can call this set of coefficients the precursor matrix, the kernel of aberrations, or PNA (here, which can be a 1x7 matrix, or a 2x4 matrix with a zero coefficient added to the first row). Because the FA I() functions are chosen to best model the modes of the aberrations obtained, the parametric model representing the αAl(X,Y) functions as a function of the geometric parameters contains significantly fewer coefficients than the numerical representation of the FA I(X,Y) functions. Here, for example, it contains 8 coefficients instead of, say, the 16 million coefficients that would have been needed to represent the aberration functions represented by 4 x 4 coefficients on an X, Y field of 1000 x 1000 positions.One final step remains to obtain the aberration matrix: representing the evolution of the PNA coefficients as a function of the sensor-objective distance and the DOC and DOS parameters. This again involves a set of several parameters that appear in functions modeling the PNA values ​​as a function of DOC and DOS. It is this final set of coefficients that can advantageously constitute the kernel aberration matrix.

[0053] Following another implementation example, the kernel aberration matrix can be a table of matrices containing the PNA coefficients, or directly the PNA for several sets of DOC and DOS parameters. The essential point is to be able to retrieve the objective's aberration matrices from the geometric parameters, preferably while storing less data than these aberration matrices represent.

[0054] The kernel aberration matrix can be determined in the device performing the image capture, or outside the device performing the image capture.

[0055] For example, the function basis FA I can be the function basis of Zernike polynomials.

[0056] At least one correction matrix can be calculated from at least one kernel correction matrix, comprising coefficients allowing the said at least one correction matrix to be deduced, and previously calculated.

[0057] Using a kernel correction matrix avoids having to store all the correction matrices used to correct each image captured during each iteration of the correction phase, thus reducing the memory space used for storing the correction matrices.

[0058] The kernel correction matrix can be determined as follows. Several correction matrices can be determined from several aberration matrices or from a single kernel aberration matrix. Then, from these correction matrices, a common correction matrix, denoted MC, can be identified. This common correction matrix can be stored in place of the multitude of correction matrices, along with one or more relationships, to allow the calculation of each correction matrix.

[0059] The determination of the correction kernel matrix can be determined in a similar manner to that described above for the aberration kernel matrix using the same FA I function basis, or another function basis that would be appropriate.

[0060] The correction kernel matrix can be determined in the device performing the image capture or outside the device performing the image capture.

[0061] The method according to the invention may further include, prior to the first iteration of the acquisition phase, a characterization phase comprising a determination of at least one, and in particular of each, aberration matrix.

[0062] The characterization phase can generally be performed outside the image acquisition device. In this case, the characterization phase is carried out with the optical lens, or image sensor, which is not yet mounted in the device.

[0063] Of course, according to alternative embodiments, the characterization phase can also be carried out within the device, in particular with the optical lens, or the image sensor, mounted in the device.

[0064] The characterization phase may include a determination of: several aberration matrices for several regions R i of the image sensor; several aberration matrices for several DOC distance values ​​i; and / or several aberration matrices for several DOS distance values ​​j.

[0065] Indeed, an MA aberration matrix can be determined for the entire image sensor. Alternatively, MA i aberration matrices can be determined, each for a region R i of the image sensor, with i=1,...,I ​​and I≥2.

[0066] Alternatively, or in addition, MA j aberration matrices can be determined, each for a DOC lens-sensor distance i, with j=1,...,J and J≥2.

[0067] Alternatively, or in addition, MA k aberration matrices can be determined, each for a DOS objective-scene distance k, with k=1,...,K and K≥2.

[0068] Following a non-limiting combination, several MA ijk aberration matrices can be determined, each for a region R i, a DOC distance j, and a DOS objective-scene distance k in said region R i.

[0069] Following another non-limiting combination, several aberration matrices MA ik can be determined, each for a region R i, and a lens-scene distance DOS k in said region R i. In this case, the aberration matrix, and therefore the image correction, does not take into account the lens-image sensor distance DOC.

[0070] According to embodiments, at least one aberration matrix can be determined by optical measurement, on the actual optical objective, with an optical measuring device.

[0071] An optical measuring device may include a Modulation Transfer Function (MTF) or Optical Transfer Function (OTF) measuring device, a Point Spread Function (PSF) measuring device, or a wavefront measuring device. Such devices are well known to those skilled in the art and will not be described in further detail here for the sake of brevity. They generally comprise a light source, a target or pattern to be observed, an image sensor (or several), and a processing unit.

[0072] Alternatively, or in addition, at least one aberration matrix can be determined by simulation in a numerical simulator on a numerical model of the optical lens.

[0073] In this case, the optical lens and image sensor are modeled in software, such as ZEMAX ®< software, in which it is possible to simulate MTF, OTF or PSF functions, or even wavefront functions, by simulating the emission and propagation of a test light beam and measuring the illumination received on the image sensor.

[0074] Regardless of the method of implementation, whether by measurement on the optical lens or by simulation on a numerical model of the optical lens, an aberration matrix can be determined in the following way.

[0075] The optical lens associated with the image sensor is illuminated by a test light beam, and the illumination received at the image sensor is then measured. This illumination at the sensor contains information about the optical aberrations introduced by the lens in each captured image.

[0076] The aberration matrix can be obtained by region of the image sensor, each region corresponding to one or more pixels on the sensor. In this case, an illumination pattern is presented to the optical lens, such as a checkerboard pattern alternating between white and black, and the illumination received at the sensor is then measured. This illumination at the sensor contains information about the optical aberrations introduced by the lens in each captured image.

[0077] A transformation of the captured images may be necessary to obtain, for example, the point spread function (PSF), which would correspond to a test beam originating from a single point light illuminating the lens, moved across several regions of the visible field by the lens to obtain the PSFs of multiple sensor regions. However, if the measurement, or simulation, allows the optical lens to be illuminated from a single movable point, these transformations can generally be omitted.

[0078] Following non-limiting examples of implementation, at least one, in particular each, aberration matrix can be: a PSF matrix of values ​​of a point spread function (“Point Spread Function” or “PSF” in English), and / or an OTF matrix of values ​​of an optical transfer function (“Optical Transfer Function” or “OTF” in English), and / or a matrix of values ​​obtained by wavefront analysis.

[0079] Of course, other types of aberration matrices can be used and the invention is not limited to a particular type of aberration matrix.

[0080] The method according to the invention may include a step of calculating a matrix, called the kernel aberration matrix, comprising coefficients allowing at least one aberration matrix to be deduced.

[0081] The aberration kernel matrix can be determined as described above, either inside or outside the device capturing the image to be corrected.

[0082] The method according to the invention may further include a step of calculating a matrix, called the correction kernel matrix, comprising coefficients allowing at least one correction matrix to be deduced.

[0083] The correction kernel matrix can be determined as described above, either within or outside the device capturing the image to be corrected.

[0084] According to another aspect of the present invention, an image acquisition system is proposed with a device comprising a camera module having an optical lens associated with an image sensor, said system comprising: a characterization device for said optical lens to determine at least one matrix, called an aberration matrix, representative of optical aberrations introduced by said optical lens into an image captured by said camera module, said camera module to capture an image, and a processing unit in said device; configured to implement the image acquisition method according to the invention.

[0085] The system according to the invention may include, in terms of hardware and / or software means, any combination of the optional characteristics stated above for the process according to the invention and which are not repeated here for the sake of brevity.

[0086] The characterization device may include an optical measuring instrument to measure at least one aberration matrix on the actual optical lens. Alternatively, the characterization device may include a numerical simulator to determine at least one aberration matrix by simulation on a numerical model of the optical lens.

[0087] The processing unit can be a processor, a computer, or any programmable electronic chip. The processing unit can be a processor or a graphics card of the device that captures the image.

[0088] According to another aspect of the present invention, an image acquisition device is proposed comprising: a camera module comprising an optical lens associated with an image sensor, and a processing unit; configured to implement the steps of the image acquisition process according to the invention.

[0089] The device according to the invention may include, in terms of hardware and / or software means, any combination of the optional characteristics stated above for the process according to the invention and which are not repeated here for the sake of brevity.

[0090] Following non-limiting examples of embodiment, the device according to the invention can be a camera, a video camera, a smartphone, a tablet, a computer, a webcam, etc. Description des figures et modes de réalisation

[0091] Other advantages and features will become apparent upon examination of the detailed description of non-limiting embodiments and the accompanying drawings, in which: there FIGURE 1 is a schematic representation of a non-limiting example of an optical element that can be used in an optical lens as defined in the present invention; the FIGURE 2 is a schematic representation of a non-limiting example embodiment of a camera module as defined in the present invention; the FIGURE 3 is a schematic representation of a non-limiting example of a characterization phase that can be implemented in a process according to the present invention; the FIGURES 4 , 5a, 5b et 5c are schematic representations of non-limiting examples of an image acquisition method according to the present invention; the FIGURE 6 is a schematic representation of a non-limiting example of a device for characterizing an optical objective; the FIGURE 7 is a schematic representation of a non-limiting example of an embodiment of an image acquisition device according to the invention; and the FIGURE 8 is a schematic representation of a non-limiting example embodiment of another image acquisition device according to the invention.

[0092] It is understood that the embodiments described below are by no means exhaustive. In particular, variants of the invention may be conceived comprising only a selection of the features described below, isolated from the other features described, if this selection of features is sufficient to confer a technical advantage or to differentiate the invention from the prior art. This selection includes at least one preferably functional feature without structural details, or with only a portion of the structural details if this portion alone is sufficient to confer a technical advantage or to differentiate the invention from the prior art.

[0093] In particular, all the variants and embodiments described can be combined with each other if there are no technical obstacles to this combination.

[0094] In the figures and in the rest of the description, elements common to several figures retain the same reference.

[0095] There FIGURE 1 is a schematic representation of a non-limiting example of an optical element that can be used to manufacture an optical lens.

[0096] Optical element 100 of the FIGURE 1 can be used with at least one other optical element to make an optical lens. An example of an optical lens, given by way of non-limiting example, will be described with reference to the FIGURE 2 .

[0097] The optical element 100 can be a lens, a plate, etc. In the following, and without loss of generality, we consider that the optical element is a lens.

[0098] Optical lens 100, for example, can be manufactured by injection molding. An injection molding process generally follows the sequence of steps below: injection of the polymer and filling of the mold, pressurization, pressure maintenance, cooling, and demolding Injection molding processes for lenses, although common, can fluctuate and generate errors in the characteristic parameters of the lenses, particularly with regard to their geometry.

[0099] Lens 100 has a given geometric shape. It comprises two interfaces, 1021 and 1022, each also having a given geometric shape. Thus, the geometric shape of lens 100 is determined by: a geometric shape of each of the optical interfaces 102 1 and 102 2; a center of curvature, denoted CC1 and CC2, of each of the optical interfaces 102 1 and 102 2; a position of an apex, denoted A1 and A2, of each of the optical interfaces 102 1 and 102 2; at least one thickness, denoted H1 and H2, of the lens 100 along its periphery; an inner diameter, respectively D11 and D21, and / or an outer diameter, respectively D12 and D22, of each of the interfaces 102 1 and 102 2; a concentricity or eccentricity value of the interfaces 102 1 and 102 2; a surface roughness of each of the optical interfaces 102 1 and 102 2; etc.

[0100] The value of at least one of these geometric parameters can be provided by the manufacturer. Alternatively, or in addition, the value of at least one geometric parameter can be measured, for example, by optical or mechanical profilometry. Alternatively, or in addition, the value of at least one geometric parameter can be determined by simulation, based on a numerical model of lens 100. Alternatively, or in addition, the value of at least one geometric parameter can be measured, for example, by optical interferometry.

[0101] Furthermore, lens 100 has optical characteristics since it is an optical element. It is therefore characterized by at least one optical parameter such as, for example: a refractive index, denoted I1 and I2, of each of the optical interfaces 102 1 and 102 2; an Abbé number, denoted Ab, etc.

[0102] Any combination of these individual parameters, and in particular the set of these parameters, can be used to numerically model the optical element 100.

[0103] There FIGURE 2 is a schematic representation of a non-limiting example embodiment of a camera module comprising an optical lens and an image sensor within the meaning of the present invention.

[0104] The 200 camera module includes a 202 image sensor. The image sensor can be any type of photosensitive sensor such as a CMOS sensor (called "CMOS Imager System" which provides the acronym CIS), or a CCD sensor.

[0105] The camera module 200 also includes an optical lens 204. The optical lens 204 has the function of focusing an image of a scene in an image plane, namely the plane of the image sensor 202.

[0106] A 204 optical objective is generally made up of a stack of optical elements comprising any combination of optical elements such as lenses, spacers and opacifying washers, etc.

[0107] During the manufacturing of the optical lens, each optical element of the lens is individually selected and stacked with the other optical elements in an assembly barrel, according to a given order. The stack is then secured to the barrel using known techniques, for example, by gluing.

[0108] On the FIGURE 2 and by way of non-limiting example only, the optical objective 204 comprises four lenses 206 1 -206 4 stacked, in a stacking direction 210, also called the Z axis, in a barrel 212. At least two of the lenses 206 1 -206 4 may be separated from each other by an empty space, called an "air gap", or by a spacer, or washer, also called a "spacer".

[0109] At least one of the lenses 206 1 -206 4 could, for example, be lens 100 of the FIGURE 1 .

[0110] Each of the lenses 206 1 -206 4 has two interfaces, namely an interface, called upstream, and an interface, called downstream, in the direction of the stacking 210. Thus, the lens 206 1 has an upstream interface 214 1 and a downstream interface 214 2, the lens 206 2 has an upstream interface 214 3 and a downstream interface 214 4, the lens 206 3 has an upstream interface 214 5 and a downstream interface 214 6 and the lens 208 4 has an upstream interface 214 7 and a downstream interface 214 8.

[0111] Thus, for the 204 optical objective of the FIGURE 2 , and in general for any optical objective comprising a stack of optical elements, it is possible to determine a dataset, called a geometric set, denoted JG in the following, comprising data relating to at least one geometric parameter of at least one, and in particular of each, optical interface 206 i of said stack.

[0112] Such a geometric set JG may include data relating to, or values ​​of, any of the following geometric parameters: at least one position of at least one optical interface 206 1 -206 4 of the objective 204 along the Z axis; at least one decentering value of at least one optical interface 206 1 -206 4 with respect to the Z axis, or with respect to a center position of another interface, in the XY plane; or at least one tilting value of at least one optical interface 206 1 -206 4 with respect to the Z axis, or with respect to the tilt of another interface; at least one topography or shape profile value of at least one optical interface 206 1 -206 4.

[0113] In general, the geometric set JG can include, for each optical interface of the 204 optical lens, M geometric parameters with M ≥ 1 and preferably M ≥ 2. If the 204 optical lens comprises N optical elements, each optical element having two interfaces, then the geometric set JG can include 2N x M parameters and can correspond to a matrix with 2N rows and M columns. Of course, the geometric set JG can include the same number of geometric parameters for at least two optical interfaces, or different numbers of geometric parameters for at least two optical interfaces.

[0114] The JG geometric set can directly include the values ​​of the geometric parameters. These values ​​can be measured by optical interferometry or by confocal measurement(s), preferably from a side or face of the optical lens 204, so as to avoid rotating it.

[0115] The geometric set of the optical lens 204, possibly with the individual parameters of each optical element, such as any combination of the parameters described with reference to the FIGURE 1 , can be used to digitally model the camera module in a digital simulator, such as, for example, and without loss of generality, the commercially available “ZEMAX ®<” simulator.

[0116] Optionally, the distance, denoted DOC, between the optical lens 204 and the image sensor 202 can be changed, for example by a focusing or zoom mechanism. The value of this DOC distance can be measured by a sensor designed for this purpose, or can be provided by the zoom mechanism, or from a configuration of said zoom mechanism.

[0117] Furthermore, the distance, denoted DOS, between the optical lens and the imaged scene generally varies from one image to another. Moreover, within the same image, objects in the scene may be located at different distances from the optical lens. For a given image, this DOS distance, and possibly the DOS distances for different regions of the optical sensor, can be determined either by analyzing the captured image or by one or more sensors, such as LiDAR sensors.

[0118] There FIGURE 3 is a schematic representation of a non-limiting example of a characterization phase that can be implemented in the present invention.

[0119] Phase 300 of the FIGURE 3 allows the determination of one or more aberration matrices, representative of the optical aberrations introduced into an image taken by an optical lens, and in particular by the 200 optical lens of the FIGURE 2 .

[0120] The characterization phase 300 can be performed outside the image acquisition device, that is, before the optical lens or camera module is mounted in the image acquisition device. Alternatively, the characterization phase 300 can be performed inside the image acquisition device, after the optical lens or camera module is mounted. In this case, some or all of the steps of the characterization phase are performed within the image acquisition device.

[0121] Phase 300 includes a step 302 of emitting a test light beam towards the optical lens, and in particular towards the camera module comprising the optical lens and the image sensor. The test light beam can be described by a matrix of values, denoted MT.

[0122] The test light beam passes through the optical lens and is received at the image sensor. The illumination that has passed through the optical lens is measured by the optical sensor in step 304. This illumination received at the sensor contains information about the optical aberrations introduced by the optical lens in each captured image. The illumination observed and measured at the image sensor can be described by a matrix of values ​​denoted MTOBS.

[0123] Knowing the MT matrix and measuring the MTOBS matrix, it is possible to determine, in step 306, an aberration matrix, MA, as a convolution of the matrices between MT and MA, such that MTOBS = MT * MA

[0124] Optionally, but particularly advantageously, the aberration matrix can be obtained individually for different DOC distances j between the optical lens and the image sensor, with j=1,...,J and J≥2. In this case, at a step 308, the DOC distance is modified, for example by moving the image sensor closer to or further from the optical lens, and steps 302-306 are repeated for each DOC distance j.

[0125] Optionally, but particularly advantageously, the aberration matrix can be obtained individually for different DOS distances k between the optical lens and the image sensor, with k=1,...,K and K≥2. In this case, at a step 310, the DOS distance is modified, for example by moving the source of the test light beam closer to, or further away from, the optical lens, and steps 302-308 are repeated for each DOS distance k.

[0126] The aberration matrix can be obtained in one go for the entire extent of the image sensor.

[0127] Alternatively, the aberration matrix can be obtained individually for different regions of the image sensor, Ri with i=1,...,I ​​and I≥2, each corresponding to one or more pixels on said image sensor, in the plane of the image sensor. In this case, in step 312, the test region is modified and steps 302-310 are repeated for each region Ri individually.

[0128] Thus, the characterization phase 300 provides at least one MA aberration matrix. In the following, and without loss of generality, we consider that the characterization phase 300 provides a number NB = I x J x K MA aberration matrices ijk for different regions R i, different DOC distances j and different DOS distances k.

[0129] In an optional step 314, the aberration matrices can be stored in a database.

[0130] Phase 300 of the FIGURE 3 may be part of an image acquisition method according to the invention, such as the examples of methods that will be described later. However, the image acquisition method according to the invention does not necessarily include such a characterization phase.

[0131] At least one aberration matrix can be determined by optical measurement on the actual optical objective using an optical measuring device comprising an optical beam source and, optionally, a processing unit. The optical measuring device may include a Modulation Transfer Function (MTF) measuring device, or an Optical Transfer Function (OTF) measuring device, or a Point Spread Function (PSF) measuring device. Such devices are well known to those skilled in the art.

[0132] Alternatively, at least one aberration matrix can be determined by simulation in a numerical simulator on a numerical model of the optical lens. A numerical model of the optical lens can be defined in a simulator, for example, from any combination of the geometric parameters described with reference to the FIGURE 2 and / or individual parameters described with reference to the FIGURE 1 In this case, the optical lens and image sensor are modeled in software, such as ZEMAX® software, in which it is possible to simulate MTF, OTF or PSF functions, by simulating the emission and propagation of a test light beam and measuring the illumination received on the image sensor.

[0133] There FIGURE 4 is a schematic representation of a non-limiting example of an embodiment of a method according to the invention for image acquisition.

[0134] The 400 process of the FIGURE 4 includes at least one iteration of a 402 phase of image acquisition with a camera module, and in particular with the 200 camera module of the FIGURE 2 .

[0135] Phase 402 includes a step 404 of image capture. The captured image is in the form of a matrix of values, provided by the image sensor, and denoted IMa. The IMa matrix includes numerical values ​​for each pixel of the sensor. For example, for an RGB image, the IMa matrix includes three values ​​for each pixel, one for each color. The acquired image, and therefore the IMa matrix, includes the optical aberrations introduced by the lens, such as optical blur or shift.

[0136] Optionally, but particularly advantageously, the acquisition phase 402 may include a step 406 of determining, by measurement or calculation, the DOC distance between the optical lens and the image sensor during image acquisition. The correction of the captured image can then be performed by selecting, or calculating, the correction matrix corresponding to said DOC distance.

[0137] Optionally, but particularly advantageously, the acquisition phase 402 may include a step 408 of determining, by measurement or calculation, at least one DOS distance between the optical lens and the imaged scene during image acquisition. In particular, a DOS distance may be measured / calculated for each region R i of the image sensor. The correction of the captured image may then be performed by selecting, or calculating, the correction matrix(s) corresponding to said DOS distance, or to said DOS distances for each region R i.

[0138] During step 410, at least one correction matrix is ​​selected, or optionally calculated, based on the DOC and / or DOS distances determined during optional steps 406 and 408. When the correction matrix(s) is / are pre-calculated and stored in a database, then step 410 simply selects the correction matrix(s) from that database. When the correction matrix(s) is / are not pre-calculated, then step 410 performs an on-the-fly calculation of the correction matrix(s).

[0139] The captured image, and in particular the matrix IMa representing the captured image, is corrected, in part or in full, during step 412. To do this, at least some of the values ​​of the matrix IMa are corrected using at least one correction matrix selected / calculated during step 410. In the following, and without loss of generality, it is assumed that a correction matrix, denoted MCijk, is used for each region Ri of the sensor, for the DOC distance j determined during step 406, and for the DOS distance k determined during step 408 for said region Ri. Denoting IMc as the matrix representing the corrected image, the image correction can be performed by convolving the matrix IMa with each correction matrix MCijk, for each region Ri of the image sensor: IM c = IM a * MC ijk , avec i = 1 , . . , I et I ≥ 2

[0140] In the corrected image thus obtained, the aberrations introduced by the optical lens are corrected, or at least attenuated.

[0141] There FIGURE 5a is a schematic representation of another non-limiting example of an embodiment of a method according to the invention.

[0142] The 500 process of the FIGURE 5a includes the acquisition phase 402 of process 400 of the FIGURE 4 .

[0143] The process 500 further includes, prior to the acquisition phase 402, a step 502 for calculating a kernel aberration matrix from the aberration matrices determined for the optical lens. Indeed, in order to reduce the storage resources required for storing the aberration matrices, it can be advantageous to calculate a kernel aberration matrix which will then be used to calculate, on the fly, the aberration matrix for each region R i of the image sensor, optionally based on the DOC distance and / or the DOS distance(s) measured for each captured image.

[0144] In this case, during step 410, each MA ijk aberration matrix for each region R i is derived on the fly from the kernel aberration matrix. Then, each MA ijk aberration matrix is ​​used to calculate the MC ijk correction matrix for each region R i of the image sensor.

[0145] The kernel aberration matrix can be determined as described above, using a basis of functions.

[0146] There FIGURE 5b is a schematic representation of another non-limiting example of an embodiment of a method according to the invention.

[0147] Process 510 of the FIGURE 5b includes the acquisition phase 402 of process 400 of the FIGURE 4 .

[0148] The process 510 further includes, prior to the acquisition phase 402, a step 512 for calculating and storing all the MC ijk correction matrices that can potentially be used to correct an image during the acquisition phase. In other words, step 512 performs a calculation, for each region R i, and possibly for each DOC distance j and / or each DOS distance k, of an MC ijk correction matrix from the MA ijk aberration matrix determined during the characterization phase. Each of these MC ijk correction matrices is then stored in a database.

[0149] In this case, step 410 of the acquisition phase 402 does not perform any correction matrix calculation. During this step 410, the database storing the correction matrices MC ijk is read to select, for each region R i of the sensor, the desired correction matrix to correct the acquired image, possibly for the DOC distance j and / or the DOS distance k measured for said region R i.

[0150] There FIGURE 5c is a schematic representation of another non-limiting example of an embodiment of a method according to the invention.

[0151] The 520 process of the FIGURE 5c includes the acquisition phase 402 of process 400 of the FIGURE 4 .

[0152] The process 520 further includes, before the acquisition phase 402, a step 522 of calculation of all the correction matrices MC ijk which can potentially be used to correct an image during the acquisition phase 402. In other words, the step 522 performs a calculation, for each region R i, and possibly for each DOC distance j and / or each DOS distance k, a correction matrix MC ijk from the aberration matrix MA ijk determined during the characterization phase.

[0153] The process 520 further includes, before the acquisition phase 402, a step 524 of calculating a correction kernel matrix from the MC ijk correction matrices determined for the optical lens during step 522. Indeed, in order to reduce the storage resources required to store all the MC ijk correction matrices, it may be advantageous to calculate a correction kernel matrix which will be used to calculate, on the fly, each MC ijk correction matrix for each region R i of the image sensor, optionally as a function of the DOC distance and / or the DOS distance(s) measured for each image captured during the acquisition phase 402.

[0154] In this case, during step 410 of the acquisition phase, each correction matrix for each region R i is deduced, on the fly, from the correction kernel matrix, and optionally as a function of the DOC distance and / or the measured DOS distance(s).

[0155] The kernel correction matrix can be determined as described above, using a basis of functions.

[0156] In the examples described with reference to FIGURES 5a, 5b et 5c , each of the steps 502, 512, 522 and 524 can be implemented either in the imaging device performing the image capture, i.e. the device equipped with the camera module, or outside of said imaging device.

[0157] For example, in process 500 of the FIGURE 5a Step 502, calculation of the aberration kernel matrix, can be performed either within the imaging device or outside of it, for example, in an optical lens characterization device, for instance, implementing phase 300 of the characterization. FIGURE 3 .

[0158] In process 510 of the FIGURE 5b Step 512, the calculation of correction matrices, can be performed either within the imaging device or outside of it, for example, in an optical lens characterization device, for instance, implementing characterization phase 300 of the FIGURE 3 .

[0159] In process 520 of the FIGURE 5c Step 522, calculation of the correction matrices, and step 524, calculation of the correction kernel matrix, can be performed either within the imaging device or outside the imaging device, for example, in an optical lens characterization device, for example, implementing characterization phase 300 of the FIGURE 3 .

[0160] There FIGURE 6 is a schematic representation of a non-limiting example of an implementation of a device for characterizing an optical objective.

[0161] The 600 device of the FIGURE 6 can be used to characterize any type of optical lens used for image acquisition, and more specifically the 200 lens of the FIGURE 2 .

[0162] The 600 device of the FIGURE 6 can be used to implement a characterization phase of an optical objective, and in particular the 300 characterization phase of the FIGURE 3 , with a view to determining, at least, aberration matrices representative of the aberrations introduced by an optical lens in an image captured with said optical lens.

[0163] The device 600 includes a source 602 for emitting one or more test light beams towards the optical objective, and in particular towards the camera module, such as, for example, the camera module 200 of the FIGURE 2 .

[0164] Device 600 may include a mechanism (not shown) for adjusting the position of source 602: in the Z direction of emission of the test light beam, so as to adjust the DOS distance and determine aberration matrices for different DOS distances, and / or in the plane, denoted XY, perpendicular to the Z direction of emission of the test light beam, to modify the region R i for which an aberration matrix is ​​determined.

[0165] Device 600 may include a mechanism (not shown) for adjusting the relative positions of the optical lens and image sensor in the Z direction, so as to adjust the DOC distance between the image sensor and the optical lens and determine aberration matrices for different DOC distances.

[0166] Device 600 may include a computing unit 604.

[0167] This computing unit 604 is designed to receive on the one hand a value matrix, denoted MT, representing the test light beam emitted by the source 602 and on the other hand a value matrix, denoted MTOBS, representing the illumination observed by the image sensor for this test light beam.

[0168] The computing unit 604 may include a computing module 606 designed to calculate, based on the MT and MTOBS matrices, an aberration matrix, denoted MA, for example using the following convolution relation: MTOBS = MT * MA When the aberration matrix is ​​determined for a region R i, a DOC distance j, and a DOS distance k for this region R i, the relationship used can be noted as follows: MTOBS ijk = MT ijk * MA ijk

[0169] The computing unit 604 may further include a computing module 608 intended to compute an aberration kernel matrix, for example by implementing step 502 of process 500, in which case this step 502 is not implemented in the imaging apparatus.

[0170] Alternatively, or in addition, the calculation module 608 can be configured to calculate: at least one correction matrix, for example by implementing step 512 of process 510, or step 522 of process 520, in which case this step is not implemented in the imaging device; and optionally, at least one correction kernel matrix, for example by implementing step 524 of process 520, in which case this step is not implemented in the imaging device.

[0171] The computing unit 604 can be in a physical form, such as a server, a computer, a processor, an electronic chip, etc. Alternatively, the computing unit 604 can be in a software form, such as one or more computer programs. According to yet another alternative, the computing unit 604 can be formed by any combination of at least one hardware component and one software component.

[0172] Modules 606 and 608 can each be an individual module, independent of the other. Alternatively, modules 606 and 608 can be integrated into a single module. Each of modules 606 and 608 can be in a physical form, such as a server, a computer, a processor, an electronic chip, etc. Alternatively, each of modules 606 and 608 can be in a software form, such as a virtual machine, one or more computer programs, etc. According to yet another alternative, each of modules 606 and 608 can be formed by any combination of at least one hardware component and one software component.

[0173] There FIGURE 7 is a schematic representation of a non-limiting example embodiment of an image acquisition device according to the invention.

[0174] The 700 device of the FIGURE 7 can be used to implement a method according to the invention for acquiring an image, and in particular any one of the methods 400, 500, 510 and 520 of the FIGURES 4 , 5a, 5b et 5c .

[0175] The 700 device includes a 702 camera module, which could, for example, be the 200 camera module of the FIGURE 2 comprising an image sensor, for example image sensor 202, and an optical lens, for example optical lens 204.

[0176] Optionally, the device 700 may include a sensor 706 for measuring the DOC distance between the optical lens 704 and the image sensor 702. Such a sensor 706 may be a capacitive sensor, a resistive sensor, or an optical sensor. This sensor 706 provides a value for the DOC distance, or a value of an electrical quantity representative of the DOC distance, such as a voltage, a current, etc.

[0177] Optionally, the device 700 may include a sensor 708 for measuring the DOS distance between the optical lens 204 and the imaged scene. Such a sensor 708 could be, for example, a LiDAR sensor. This sensor 708 provides a value for the DOS distance, or a value of an electrical quantity representative of this DOS distance, such as, for example, a voltage, a current, etc. Preferably, but in no way limitingly, this sensor 708 is intended to measure the DOS distance individually for different regions Ri of the image sensor 202.

[0178] The device 700 further includes a computing unit 710 configured to correct the image captured by the sensor 704, and in particular the matrix IM a representing the captured image, and to provide a matrix, denoted IM c, representing the corrected image, as a function of at least one correction matrix MC ijk.

[0179] The computing unit 710 can be a hardware unit such as a processor or computer chip. Alternatively, the computing unit can be a computer program or application.

[0180] According to embodiments, at least one correction matrix MC ijk can be read from a memory area, or a database, 712. In this case, the computing unit 710 reads said at least one correction matrix, optionally as a function of DOC and / or DOS distance(s) measured during image acquisition.

[0181] Depending on embodiments, the calculation unit 710 can also be configured to calculate, in particular on the fly, at least the correction matrix, as a function of: of a correction kernel matrix, at least one predetermined aberration matrix, or an aberration kernel matrix; stored in database 712. In this case, the computing unit reads said at least one matrix and calculates the at least one correction matrix from said at least one matrix read.

[0182] In the example shown, the camera module 702 consists of the optical lens 204 and the image sensor 202, and optionally the distance sensors 706 and 708. The camera module 702 may include other components than those shown, such as for example a focus adjustment mechanism (not shown) modifying the distance between the image sensor 704 and the optical lens 702.

[0183] The computing unit 710, and optionally the database 712, can be integrated into a photo and / or video application, 714, installed or run within a device such as a Smartphone, tablet, computer, etc.

[0184] There FIGURE 8 is a schematic representation of a non-limiting example embodiment of an image acquisition device according to the invention.

[0185] The 800 unit can include all the elements of the 700 unit of the FIGURE 7 .

[0186] Depending on the embodiment, the 800 device can be a camera, a tablet, a smartphone, a computer, a surveillance camera, and more generally a camera module intended to be integrated into another device, etc.

[0187] In the example shown, device 800 is a Smartphone comprising a display screen 802 on the front and the camera module 702 opening onto its rear, application 714 integrating the computing unit 710 and the database 712.

[0188] Of course, the invention is not limited to the examples that have just been described.

Claims

1. A method (400;500;510;520) for image acquisition with an apparatus (700;800) comprising a camera module (200;702) comprising an optical lens (204) associated with an image sensor (704), said method (400;500;510;520) comprising at least one iteration of a phase (402) of image acquisition by said apparatus (700;800) comprising the following steps: - capturing (404) an image with said camera module (200; 702); and - correcting (412) said captured image as a function of at least one correction matrix calculated, directly or indirectly, from at least one matrix, called aberration matrix, determined for said optical lens (204), and representative of optical aberrations introduced by said optical lens (204) into an image. the image acquisition phase (402) further comprises a step (408) of determining a distance, DOS, between the optical lens (204) and the scene, the at least one correction matrix being a function of said DOS distance.

2. The method (400;500;510;520) according to the preceding claim, characterized in that the image acquisition phase (402) comprises a step (406) of determining a distance, DOC, between the optical lens (204) and the image sensor (202), the at least one correction matrix being a function of said distance DOC.

3. The method (400;500;510;520) according any of the preceding claims, characterized in that it comprises a step (410; 512; 522) of calculating at least one correction matrix within said apparatus (700; 800).

4. The method (400;500;520) according to the preceding claim, characterized in that the step (410) of calculating the at least one correction matrix is carried out during the acquisition phase (402) so that said at least one correction matrix is calculated on the fly for each captured image.

5. The method (510) according to any one of claims 1 to 3, characterized in that the step (512) of calculating the at least one correction matrix is performed prior to the acquisition phase (402) so that said calculating step (512) is common to multiple iterations of the image acquisition phase (402).

6. The method (400;500;510;520) according to any one of claims 3 to 5, characterized in that at least one correction matrix is calculated from: - at least one aberration matrix; - at least one previously calculated correction kernel matrix comprising coefficients for deducing said at least one correction matrix; or - at least one previously calculated aberration kernel matrix, comprising coefficients enabling said at least one correction matrix to be deduced.

7. The method (400;500;510;520) according to any of the preceding claims, characterized in that it further comprises, prior to the first iteration of the acquisition phase, a characterization phase (300) comprising a determination of at least one aberration matrix.

8. The method (400;500;510;520) according to the preceding claim, characterized in that the characterization phase (300) comprises determining: - multiple aberration matrices for multiple regions of the image sensor; - multiple aberration matrices for multiple DOC distance values; and / or - multiple aberration matrices for multiple DOS distance values.

9. The method (400;500;510;520) according to any one of claims 7 or 8, characterized in that at least one aberration matrix is determined by optical measurement, on the actual optical lens, with an optical measuring apparatus.

10. The method (400;500;510;520) according to any one of claims 7 to 9, characterized in that at least one aberration matrix is determined by simulation in a digital simulator on a digital model of the optical lens.

11. The method (400;500;510;520) according to any one of the preceding claims, characterized in that at least one aberration matrix is: - a matrix, called PSF matrix, of Point Spread Function (PSF) values, and / or - a matrix, called OTF matrix, of Optical Transfer Function (OTF) values, and / or - a matrix of values obtained by wavefront analysis.

12. The method (500) according to any one of the preceding claims, characterized in that it comprises a step (502) of calculating a matrix, called aberration kernel matrix, comprising coefficients enabling at least one aberration matrix to be deduced.

13. An image acquisition system with a device (700;800) comprising a camera module (702;200) having an optical lens (204) associated with an image sensor (202), said system comprising: - a device (600) for characterizing said optical lens (204) to determine at least one matrix, called aberration matrix, representative of optical aberrations introduced by said optical lens (204) into an image captured by said camera module (702;200), - said camera module (702;200) for capturing an image, and - a calculating unit (710) in said apparatus (700;800); configured to implement the method (400;500;510;520) according to any one of claims 7 to 12.

14. An image acquisition apparatus (700;800) comprising: - a camera module (702;200) comprising an optical lens (204) associated with an image sensor (202), and - a calculating unit (710); configured to implement the steps of the method (400;500;510;520) according to any one of claims 1 to 6.