Optical quality measurement method
The method employs random targets to measure MTF and point spread function locally in optronic systems, addressing phase shifting and noise issues, achieving precise and efficient optical quality assessment.
Patent Information
- Application Number
- FR2022012971
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-12-08
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2042-12-08
AI Technical Summary
Existing methods for measuring the modulation transfer function (MTF) in optronic systems are limited by issues such as phase shifting of source hole images, degraded signal-to-noise ratio, and the need for precise target positioning, especially in long focal length systems, and they often require averaging over the entire calculation area.
A method that uses random targets to measure MTF by displaying a sequence of random patterns, acquiring images, determining pixel pairings based on similarity measures, and calculating MTF locally without averaging, allowing precise measurement of MTF, point spread function, and distortion in optronic systems.
Enables accurate and efficient measurement of MTF and point spread function with high signal-to-noise ratio, capable of discerning aberrations and distortions in optronic systems without ambiguity, even in the presence of aberrations or blur.
Smart Images

Figure 00000015_0000 
Figure 00000015_0001 
Figure 00000016_0000
Abstract
Description
Title of the invention: Method for measuring optical quality Technical field of the invention
[0001] The invention relates to a method for measuring and quantifying the optical quality of, for example, an optronic system. State of the prior art
[0002] In the context of controlling the optical quality of an optical instrument and evaluating the imaging performance of an optoelectronic system, the measurement of the modulation transfer function (MTF) is known to quantify the optical performance. Indeed, the MTF represents for each of the spatial frequencies fx and f Y of the Fourier space, the attenuation of the contrast of the intensity of the image captured by the system to be characterized with respect to the object viewed.
[0003] In an optronic system, there are several contributors that influence the final quality of the image in addition to the optical system itself, such as the detector also called sensor, the electronics if applicable, vibrations, etc.
[0004] The FTM of the system, designated by FTM^t^ is the product of these different contributions, which can be calculated as follows:
[0005] [Math.l] FTMsystem = ....
[0006] where FTMoptigue is the FTM of the optical components of the system such as lenses, FTMcapteur is the FTM of the system sensor, FTMvibrcaio„ is the FTM induced by the vibrations that the system undergoes, FT M eiectronigue is the FTM due to the electronics embedded in the system such as a processing circuit.
[0007] In general, the measurement of MTF is restricted to the consideration of the MTF of the optics and the sensor, which are the main contributors to the intrinsic performance of an instrument, apart from image processing.
[0008] In the state of the art of MTF measurement, we can cite methods consisting of a measurement via the spatial impulse response h. By definition, the optical MTF is the Fourier transform of the squared modulus of the impulse response:
[0009] [Math.2] FTM = TF(\h\2) ... ...(1)
[0010] The spatial impulse response measurement, also called the Point Spread Function (PSF) is obtained experimentally by placing a target representing a source hole in an object plane of the system. However, due to problems such as the phasing of the source hole image on the sensor matrix, and the degraded signal-to-noise ratio, methods \FTM\~ Derivatives are used for measuring the spatial impulse response. Among these derivative methods, we can cite the calculation of the LSF (line spread function) and the ESF (edge spread function), which consist respectively of the use of a source slit pattern and a half-moon pattern. These methods have the disadvantage of using an average of the measurement over the entire calculation area, and of requiring precise positioning of the patterns.
[0011] There is another family of target types, namely random targets (“Random Target” or “Noiselike Target” in English). The advantage of these targets is that they are translation invariant and that they naturally and accurately sample the frequency spectrum. A random target contains in its power spectral density, designated by DSP or PSD in English for “Power Spectral Density”, all the spatial frequencies uniformly distributed, like white noise. Using such random targets, the MTF from equation (1) can be deduced by the following relationship:
[0012] [Math.3] DSP (muge ,z mireuléa1<
[0013] where DSPmire random is the DSP of the random pattern and DSPiffUtf,e is the DSP of the image captured by the random pattern system.
[0014] Random patterns can be generated by laser speckle techniques, with semi-transparent etched patterns or with LCD screens.
[0015] Such LCD screens are used to display a random pattern in the object plane and use equation (2) to calculate the MTF. Other methods propose to segment the MTF calculation areas over several regions of interest in order to have the MTF value over different parts of the system's field of view.
[0016] A limitation of these methods is the use of a spatial average over the imaged area. Indeed, the random pattern is not in reality the Fourier transform of a pulse or perfect Dirac. There are therefore frequencies in the PSD that are not represented. Such an average over the image makes it possible to standardize the spectral distribution of the random pattern. In addition, the signal-to-noise ratio of these methods is often a limit when the spread of the PSF is over a limited number of pixels.
[0017] The invention aims to improve the methods for determining the quality of an optronic system. Summary of the invention
[0018] To this end, the invention proposes a method for controlling an optronic system, the method comprising the steps of:
[0019] (a) displaying, in an object plane, a sequence of patterns by a plurality of pixels, each pixel emitting a sequence of signal values of said pattern sequence,
[0020] (b) acquiring a sequence of images of said target sequence, in an image plane of an optronic system sensor,
[0021] (c) associating for each pixel of the image plane, a corresponding pixel of the object plane, said pixel of the object plane emitting a signal captured by said pixel of the image plane,
[0022] (d) determining a quality parameter of the optronic system from the association of pixels in the image plane and pixels in the object plane.
[0023] The use of a sequence of patterns, in particular, for the association or matching of the pixels of the object plane and the image plane makes it possible to evaluate the quality of the optronic system locally and without requiring averaging of the measurements on the image.
[0024] The pattern sequence may be a randomly generated sequence.
[0025] The pixels in the object plane can be any type of light signal source, for example arranged in a grid to form a matrix of pixels.
[0026] The signal values of the pattern sequence may be the values of the light intensity emitted by the pixels.
[0027] The optronic system may comprise, in addition to the sensor, optical components such as lenses, objectives, etc., and an electronic circuit configured for example to control the sensor and / or process the images acquired by the sensor.
[0028] Step (a) may comprise displaying the pattern sequence by a screen comprising the plurality of pixels of the object plane and arranged in the object plane. Such a screen may be any type of spatial light modulator, for example a backlit liquid crystal display (abbreviated to LCD for liquid crystal display in English). The object plane may be the plane of said screen. The screen may be arranged opposite the sensor of the optronic system. The screen makes it possible to turn on and off each of the elementary pixels of the object plane independently of each other.
[0029] The pixels of the image plane can be the pixels of the sensor of the optronic system.
[0030] Step (c) may comprise a similarity measurement between the signal values of the pixels of the image plane and the signal values of the pixels of the object plane. Thus, for each pixel of the image plane a similarity measurement is carried out with each of the pixels of the object plane. A pixel of the object plane may be matched to a pixel of the image plane when the similarity measurement for these pixels is greater than a determined threshold.
[0031] Step (c) may consist of pairing pixels of the sensor ie of the image plane and pixels of the screen ie of the object plane, the pixels of the image plane having received the same light sequence emitted by the pixels of the object plane. Step (c) may be based on the determination of vectors of similar signal values between the object plane and the image plane.
[0032] This step makes it possible to match, to the nearest pixel, lines or columns of pixels of the object plane to rows or columns of pixels of the image plane.
[0033] For example, the similarity measure may be based on a correlation calculation. Alternatively, the similarity measure may be based on a calculation of the squared intensity difference (or "squared intensity differences" in English), or on a calculation of the absolute intensity difference (or "absolute intensity differences" in English), or on the calculation of the mean absolute difference (or "mean absolute difference" in English).
[0034] Step (a) may comprise a step of generating a plurality of random signals S2, • - • - ST , each signal S, having N rows and M columns. Said sequence may have a size T. The size T of the pattern sequence may be greater than or equal to 30, in particular greater than or equal to 50.
[0035] Said plurality of random signals can form the pattern sequence. The number of lines N and the number of columns M can correspond respectively to the number of lines and columns of the screen.
[0036] Step (a) may comprise the successive display of said random signals.
[0037] Each signal can be encoded in binary values or be encoded in gray level. The signal values can therefore be a sequence of 0 and 1, or integers between 0 and 255.
[0038] The sequence of signal values associated with a pixel of the object plane or the image plane may be unique and independent of the sequence of signal values of another pixel of the object plane or the image plane respectively, to ensure that the matching between the pixels of the image plane and the pixels of the object plane is unitary so as to achieve a bijection.
[0039] According to one example, the size of a pixel in the image plane may be greater than the size of a pixel in the object plane. This makes it possible to avoid an aliasing and aliasing effect. In particular, the size of a pixel in the image plane may be chosen to be greater than the size of a pixel in the object plane by taking into account the magnification of the optical components of the optronic system.
[0040] According to one embodiment, step (d) can be configured to determine the modulation transfer function as a quality parameter of the optronic system.
[0041] Step (d) can thus comprise the steps consisting of:
[0042] (dl) choosing a first row of pixels in the object plane and determining a corresponding second row of pixels in the image plane paired with the pixels of said first row,
[0043] (d2) for each of the first line and the second line, calculate the density power spectral,
[0044] (d3) determine the modulation transfer function as the ratio between the power spectral density of the second line and the power spectral density of the first line.
[0045] Step (a) makes it possible to associate an object pixel with an image pixel. It is therefore possible to associate a line of pixels of the object plane with a line of pixels of the image plane regardless of the imaged field or the magnification. Thus, the MTF can be obtained locally from any area of the image, without having to resort to pixel averaging in areas of the image plane, or homogeneous spectral distribution of the sequence of patterns, which is necessary in the case where there is no matching between the pixels.
[0046] Thus it is also possible to measure the MTF by zone, sagittal, tangential, of the center of the field, of the edge of the field, etc.
[0047] According to one embodiment, step (d) can be configured to determine the point spread function as a quality parameter of the optronic system.
[0048] Step (d) can thus comprise the steps consisting of:
[0049] (d4) choose a first pixel in the object plane and determine an image zone around it of a second pixel paired with the first pixel,
[0050] (d5) determining a similarity between the signal values of the first pixel and the signal values of each pixel in the image area,
[0051] (d6) determine a point spread function as a function of the similarity of completed at step (d5).
[0052] For example, the first pixel may have coordinates ( m) in the object plane, ; with n and m being integers. The first pixel may have the following sequence of signal values: P = [P^l, m), P2(n, m), ■■■ , P^n, m)].
[0053] The second pixel may be the theoretical position of the geometric image of P. The image area around the second pixel may comprise pixels with coordinates (n', nï), with n' and m' being integers. For example, n' and m may be between 20 and 80, in particular equal to 50.
[0054] The pixels of the image area may have the following sequence of signal values: P' = [P / jX m'), P^ri, m'), ■ ■ ■ , PT{fï, m')]-
[0055] Step (d5) may comprise calculating, for each pixel of the image area around the second pixel, the cross-correlation between P and P' of all the pixels of the image area, according to the following formula:
[0056] [Math.4]
[0057] with i and j being integers. The cross-correlation C(P, P ) of the vectors P and P' can quantify their similarity. If the vectors are identical then the correlation can be equal to or close to 1. The correlation C(P, P ) Pcut decreases as the vectors P and P' differ.
[0058] For example, a threshold can be set below which the vectors P and P' can be considered as being independent. The value of said threshold can be expressed experimentally and vary as a function of the length T. of the target sequence. The sensitivity of detection of the similarity varying in root of T, it is thus possible to obtain the profile of the spread of the spread function of the point with a very high signal-to-noise ratio and to geometrically discern the different aberration tasks of the optronic system.
[0059] According to one embodiment, step (d) can be configured to determine the distortion generated by the optronic system as a quality parameter of the optronic system.
[0060] Step (d) may comprise the steps of:
[0061] (d6) choose at least two first pixels in the object plane and determine second pixels in the image plane paired with said first two pixels of the object plane, respectively,
[0062] (d7) determine the ratio of the distance between the two second pixels and the distance between the first two pixels,
[0063] (d8) evaluate the distortion as a function of said determined ratio.
[0064] Step (d6) may comprise the selection of several first pixels in the object plane, for example between 6 and 20 first pixels. Said first pixels may be arranged in a grid pattern in the object plane. For example, the distances between two first pixels of the same horizontal line, respectively of the same vertical column, of the object plane, may be equal.
[0065] In step (d6), for each first pixel of the object plane, the pixel matching carried out in step (c) can be used to search in the image plane for the corresponding second pixel with, for example, a maximum correlation.
[0066] The distortion parameter may be the difference in metrics between two first pixels of the object plane and two second pixels of the associated image plane.
[0067] Thus, it is possible to measure the distortion of the optronic system, even in the presence of aberration or blur in the capture of images by the sensor.
[0068] The quality parameter of the optronic system can be any type of quality factor other than those already mentioned.
[0069] The method may comprise, prior to step (a), the steps consisting of:
[0070] - displaying a sequence in binary coding in the object plane,
[0071] - perform an association between a pixel zone of the object plane and a zone of pixels of the image plane based on the binary encoded sequence.
[0072] These preliminary steps make it possible to know in a more coarse manner the correspondence zone of the pixels of the object plane towards the pixels of the image plane, before making the precise calculation of the similarities in step (c). This makes it possible to reduce the resources of calculation necessary for the execution of the process.
[0073] The binary encoded sequence may be a Gray code or a predefined binary code to simplify the association between areas of the object plane and the image plane.
[0074] The invention also relates to a device for controlling an optronic system comprising a display screen and a processing circuit configured to implement the method as mentioned above.
[0075] The display screen can be arranged opposite the sensor of the optronic system Brief description of the figures
[0076] [Fig-1] is a schematic representation of a first example of embodiment of a method according to the invention,
[0077] [Fig.2] represents an example of a sequence of random patterns used in the method of [Fig.l],
[0078] [Fig.3] represents an example of implementation of the step of measuring the power spectral density in the method of [Fig.l],
[0079] [Fig.4] represents a first example of realization of the step of measuring the point spread function in the method of [Fig.l],
[0080] [Fig.5] represents a second example of realization of the step of measuring the point spread function in the method of [Fig.l],
[0081] [Fig.6] represents a third example of realization of the step of measuring the point spread function in the method of [Fig.l],
[0082] [Fig.7] represents an example of carrying out the distortion measurement step in the method of [Fig.l].
[0083] [Fig.8] is a schematic representation of a second example of embodiment of a method according to the invention,
[0084] [Fig.9] is a comparative schematic representation of the calculation of the modulation transfer function theoretically and the determination of the modulation transfer function according to the method of [Fig.3]. Detailed description of the invention
[0085] With reference to [Fig.l], the method 100 is configured to evaluate the quality of an optronic system comprising for example a sensor, optical components such as lenses, objectives, etc., and an electronic circuit configured for example to control the sensor and / or process the images acquired by the sensor. For this purpose, a display screen is arranged opposite the sensor, so that the sensor receives the signal emitted by the screen. The display screen comprises a plurality of pixels arranged in an object plane. The display screen may be a backlit liquid crystal display (abbreviated as LCD for liquid crystal display in English). Furthermore, the display screen may be any type of spatial light modulator comprising a plurality of light sources.
[0086] The size of a pixel in the image plane may be greater than the size of a pixel in the object plane. This makes it possible to avoid an aliasing and aliasing effect. In particular, the size of a pixel in the image plane is chosen to be greater to satisfy the following relationship:
[0087] [Math.5] pix , > pix . .G
[0088] with pixcapteur being the size of the elementary pixel of the sensor and being the size of the elementary pixel of the screen, and G being the magnification of the optical system defined by G = dim / d
[0089] With dim a unit size in the image plane and d^j a unit size in the object plane.
[0090] In the case of a long focal length optronic system with a hyperfocal length of several hundred meters, it is possible to use a collimator type optical system and place the screen at the focus of the collimator which is a tool for putting the display screen at infinity, in this case the magnification G is defined as ~. ~ f
[0091] With f the focal length of the sensor lens and f cnuimatetir the focal length of the collimator.
[0092] The method 100 comprises a step 102 of displaying on the screen a sequence of random patterns 5 = S{, S2, ... , ST where each signal St has N lines and M columns. An example of a sequence of random patterns S is shown on the left in Figure 2. Each signal Si is generated randomly. Said sequence has a size T, greater than or equal to 30, in particular greater than or equal to 50.
[0093] The number of lines N and the number of columns M correspond respectively to the number of lines and columns of the screen.
[0094] On the right of Figure 2, the intensity vector of a pixel is represented on the ordinate as a function of the image of the sequence of random patterns 5, on the abscissa.
[0095] Each signal can be encoded in binary values or be encoded in gray level. The signal values can therefore be a sequence of 0 and 1, or integers between 0 and 255.
[0096] The method 100 comprises a step 104 of acquisition by the sensor of the images £=5^2, ... ,Sr.
[0097] The method 100 comprises a step 106 for determining a similarity between the following sequences of signal values: P = ni), P^n, ni), • • ■ , P^n, ni)] emitted by the pixels of the display screen and the sequences of signal values P' = [Prn'), P^i, m'), • • • , PT'(ri, m')] received by the sensor pixels. The coordinates (n, m) are relative to the pixels in the object plane, i.e. of the display screen. The coordinates m') are relative to the pixels in the image plane, i.e. of the sensor.
[0098] Step 106 aims to pair pixels of the sensor, i.e., of the image plane, and pixels of the screen, i.e., of the object plane. We search for the pixels of the image plane that have received the same light sequence emitted by the pixels of the object plane.
[0099] The sequence of signal values associated with a pixel of the object plane (image plane) is advantageously unique and independent of the sequence of signal values of another pixel of the object plane (image plane) respectively to guarantee that the matching between the pixels of the image plane and the pixels of the object plane is unitary (bijection).
[0100] For example, it is possible to generate 2r different binary sequences, and for example when T is greater than 50, there are 2A50 different combinations, which is much greater than the number of pixels on a display screen, which may be, for example, 2e6 pixels for a screen of size 1920x1080 pixels. Based on a random signal, it is therefore very unlikely to display an identical sequence between pixels on such a display screen. The screen successively displays the random signals.
[0101] This step 106 makes it possible to match pixel-wise the lines, or columns, of pixels of the object plane to lines, or columns, of the image plane. This allows the calculation of the modulation transfer function (MTF) without having to resort to an average over the lines or columns of the images.
[0102] For example, the similarity measure can be based on a correlation calculation c=f(p,p}
[0103] The method comprises a step 108 in which the pixels are considered to be matched, if this measurement C is greater than a certain threshold. Several similarity methods can be used such as the calculation of the squared intensity difference (or "squared intensity differences" in English), or on the calculation of the absolute intensity difference (or "absolute intensity differences" in English), or on the calculation of the average deviation (or "mean absolute difference" in English).
[0104] Optionally, in order to optimize the calculation in step 106, the method 100, in the embodiment shown in FIG. 8, comprises a step 110 of displaying by the display screen a sequence in binary code C — Cp C2, ... , CT which can be a Gray code or a predefined binary code to simplify the association between zones of the object plane and the image plane. The method 100 then comprises:
[0105] - a step 110 of displaying the image sequence in binary code or gray code
[0106] - a step 112 of acquisition by the sensor of the binary code sequence ...,C'r,and
[0107] - a step 114 of reading the code C' for a pixel in the image plane to de terminate a reduced pixel calculation area in the object plane
[0108] - a step 106 of determining the similarity between the sequences of values S' of the image plane and the sequences of values S of the object plane for each sequence of values of the signal P' on each sequence of values of the signal P of the reduced calculation zone in the object plane.
[0109] This makes it possible to perform an association between a pixel area of the object plane and a pixel area of the image plane on the basis of the binary coding sequence.
[0110] These preliminary steps make it possible to know in a more coarse manner the correspondence of the pixels of the object plane to the pixels of the image plane, before making the precise calculation of the similarities in step 106 on a reduced zone of pixels in the object plane. This makes it possible to reduce the computational resources necessary for the execution of the method because, for each pixel of the image plane, a similarity measurement is carried out with each of the pixels of a reduced zone of the object plane obtained by binary or gray coding.
[0111] The method 100 comprises a step 103 of determining a quality factor of the optronic system.
[0112] The quality factor can be the modulation transfer function, the point spread function or the distortion generated by the optronic system.
[0113] Thus, step 103 comprises a step 116 of determining the modulation transfer function by determining the power spectral density. With reference to [Fig. 3], step 116 comprises selecting a first line 204 of pixels in an image 202 of the object plane and determining a corresponding second line 208 of pixels in an image 206 of the image plane paired with the pixels of said first line 204.
[0114] Then, for each of the first line 204 and the second line 208, the power spectral density is calculated.
[0115] The modulation transfer function (MTF) is determined by the ratio between the power spectral density 212 of the second line 208 and the power spectral density 210 of the first line 204.
[0116] [Fig.9] shows a simulation comparison of the calculation of the MTF from a theoretical model of an optic (solid line) and from the ratio of spectral power density of the image plane and the object plane (line —o) to vectors of the image plane as determined in step 116.
[0117] It is clear that the determination of the FTM according to the method 100 corresponds to the calculation of the FTM according to the theoretical model.
[0118] Thus, the MTF can be obtained locally from any area of the image, without having to resort to calculating the DSP on an averaged area of the pixels of the image plane, which amounts to spectrally homogenizing the sequence of object patterns.
[0119] It is also possible to measure the MTF by zone, sagittal, tangential, center of field, edge of field, etc.
[0120] Alternatively or in addition, step 103 comprises a step 118 of determining the point spread function. Exemplary embodiments of step 118 are shown in FIGS. 4, 5 and 6. Step 118 comprises selecting a first pixel 304 having coordinates (n, m) in the object plane, with n and m being integers. The first pixel 304 has, for example, the following sequence of signal values: P = [P / zz, m), P2(n, m), ■ ■ ■, P^n, w)L
[0121] Next, a second pixel 308 in the image plane 206 corresponding to the first pixel 304 is determined using the association or matching performed in step 108.
[0122] Step 118 also comprises calculating the similarity between the signal values of the first pixel 304 and the signal values of each pixel of an image area 307 around the second pixel 308.
[0123] The image area 307 around the second pixel comprises the pixels of coordinates ( nï), with n' and m' being integers. For example, n' and m are equal to 50 (for example).
[0124] The pixels of the image area 307 have the following sequence of signal values: P' = [P['(«', m'), P2(n\ m'), ••• , PT'(n\ m')]-
[0125] The similarity calculation may consist of calculating, for each pixel of the image zone 307, the cross-correlation between P and P' of all the pixels of the image zone, according to the following formula:
[0126] [Math.6] C(P, +J)
[0127] with i and j being integers. The cross-correlation p, p} of vectors P and P' quantifies their similarity. If the vectors are identical then the correlation is equal to or close to 1 (for a normalized correlation function C). The correlation C^P. P ) decreases as the difference between vectors P and P' increases.
[0128] For example, a threshold is set below which the vectors P and P' are considered to be independent. The value of said threshold is chosen experimentally in order to avoid possible false associations and varies according to the length T of the pattern sequence.
[0129] Displaying the correlation matrix c(P, P ) allows the image of the spread function of point 300-1, 300-2, 300-3 to be visualized.
[0130] The method thus makes it possible to obtain the form of the spread of the point spread function with a very high signal-to-noise ratio and to geometrically discern the different aberration tasks of the optronic system.
[0131] For example, in [Fig.4] the spread function of point 300-1 represents a optronic system having diffraction-limited optics.
[0132] In [Fig.5], the point spread function 300-2 represents an optronic system having optics with a strong homogeneous defocus aberration on the sensor.
[0133] Thus, in this case of calculation with aberration, although the signal to noise ratio is less than 1 in the case of strong defocus, the estimation of the task of the point spread function can be done without ambiguity.
[0134] In [Fig.6], the point spread function 300-3 represents an optronic system having optics with field aberrations such as astigmatism and coma, combined with low defocus. Curve 310 represents the point spread function in 3D. Peak 312 corresponds to the value of the correlation of the second pixel 308.
[0135] Alternatively or in addition, step 103 comprises a step 120 of determining the distortion. Step 120 comprises the selection of a plurality of first pixels 404 in the object plane 202, for example 15 first pixels 404 as in the example of [Fig.7]. The first pixels 404 are in a grid pattern in the object plane 202. For example, the distances between each two first pixels 404 of the same horizontal line, respectively of the same vertical column, of the object plane, are substantially equal.
[0136] In step 120, for each first pixel 404 of the object plane 202, the pixel matching carried out in step 108 is used to search in the image plane 206 for the second pixel 408 corresponding to said first pixel 404, for example with a maximum correlation.
[0137] Then, the distortion generated by the optronic system is calculated as the difference in metrics between two first pixels 404 of the object plane 202 and two second pixels 408 of the associated image plane 206. For example, the distortion is the ratio of the distance between the two second pixels 408 and the distance between the two first pixels 404.
[0138] Thus, it is possible to measure the distortion of the optronic system, even in the presence of aberration or blur in the capture of images by the sensor.
Claims
Claims
1. Method (100) for controlling an optronic system, the method comprising the steps of: (a) displaying (102), in an object plane (202), a sequence of patterns (S) by a plurality of pixels, each pixel emitting a sequence of signal values of said sequence of patterns, (b) acquiring (104) a sequence of images (5") of said sequence of patterns, in an image plane (206) of a sensor of the optronic system, (c) associating (108) for each pixel of the image plane, a corresponding pixel of the object plane, said pixel of the object plane emitting a signal captured by said pixel of the image plane, (d) determining a quality parameter of the optronic system from the association of the pixels of the image plane (206) and the pixels of the object plane (202).
2. The method (100) of the preceding claim, wherein step (a) comprises displaying the pattern sequence by a screen comprising the plurality of pixels of the object plane and arranged in the object plane (202).
3. Method (100) according to one of the preceding claims, in which step (c) comprises a similarity measurement (C) between the signal values of the pixels of the image plane (206) and the signal values of the pixels of the object plane (202), a pixel of the object plane being matched to a pixel of the image plane when the similarity measurement for these pixels is greater than a determined threshold.
4. A method according to one of the preceding claims, wherein step (d) comprises the steps of: (dl) choosing a first row (204) of pixels in the object plane (202) and determining a corresponding second row (208) of pixels in the image plane (206) matched to the pixels of said first row, (d2) for each of the first row and the second row, calculating the power spectral density (210, 212), (d3) determining the modulation transfer function as the ratio of the power spectral density (212) of the second row to the power spectral density (210) of the first row.
5. Method (100) according to one of claims 1 to 3, in which step (d) comprises the steps of: (d4) choosing a first pixel (304) in the object plane (202) and determining an image zone (307) around a second pixel (308) matched at the first pixel, (d5) determining a similarity (C(P,P')) between the signal values of the first pixel and the signal values of each pixel of the image area, (d6) determining a point spread function (300-1,300-2,300-3) based on the similarity determined in step (d5).
6. Method (100) according to one of claims 1 to 3, wherein step (d) comprises the steps of: (d6) choosing at least two first pixels (404) in the object plane (202) and determining second pixels (408) in the image plane (206) paired with said first two pixels of the object plane, respectively, (d7) determining the ratio of the distance between the two second pixels and the distance between the two first pixels, (d8) evaluating the distortion as a function of said determined ratio.
7. Method (100) according to one of the preceding claims, comprising prior to step (a), the steps of: - displaying (110) a binary coded sequence (C') in the object plane, - performing (114) an association between a zone of pixels of the object plane and a zone of pixels of the image plane on the basis of the binary coded sequence.
8. Method (100) according to one of the preceding claims, in which step (a) comprises a step of generating a plurality of random signals Sp S2, ■... ST , each signal ST having N rows and M columns, and said sequence having a size T, said plurality of random signals forming the pattern sequence (S), and in which step (a) comprises the successive display of said random signals.
9. Method (100) according to one of the preceding claims, wherein the size of a pixel in the image plane (206) is greater than the size of a pixel in the object plane (202).
10. Device for controlling an optronic system comprising a display screen and a processing circuit configured to implement the method (100) according to one of the preceding claims.
