Adjustment method, particularly operational, of a detection system, especially by uncooled infrared imaging, and associated adjustment device.

Abstract

This method for adjusting a detection system (2) forming a succession of images of a scene (4), the detection system (2) comprising a sensor (6) equipped with an array of active detectors (8), a plurality of blind detectors (10), and a readout circuit (12) providing a response signal for each sensitive detector (8), includes calculating a representative value of at least a portion of the response signals, determining a deviation between said representative value and a target value, comparing the deviation to a tolerance threshold, and, depending on the result of the comparison, modifying a base voltage to reduce the deviation. Figure for the abstract: Fig 2

Description

Title of the invention: Method for adjusting, in particular operational, a detection system, in particular by uncooled infrared imaging, and associated adjustment device. technical field

[0001] The invention has as its technical field electromagnetic radiation detection systems, in particular bolometric detection systems for uncooled infrared imaging.

[0002] In particular, the present invention relates to a method for adjusting, especially an operational method, a detection system and an associated adjustment device. Prior techniques

[0003] The temperature of an object is characteristic of its spectral emission. Analyzing the electromagnetic radiation emitted from a scene containing the object can enable its detection and characterization. There are optical scene detection, monitoring, and / or surveillance applications that require distinguishing objects with similar temperatures and / or detecting the temperature of objects over a wide temperature range using the same detection system.

[0004] The detection system generally comprises an array of bolometric detectors sensitive to electromagnetic radiation from the scene, as well as a plurality of blind bolometers insensitive to said electromagnetic radiation.

[0005] The resistance of the matrix bolometric detectors depends on the scene temperature and the temperature of the detection system components. The resistance of the blind bolometers does not depend on the scene temperature and depends on the temperature of the detection system components.

[0006] The detection system also includes an analog-to-digital converter and a computer with a memory in which are stored a setting value for the bias voltage of the bolometric detector matrix, a setting value for the bias voltage of the blind bolometers and a setting value for a trans-impedance capacitive amplification gain, known as "CTIA gain".

[0007] The detection system provides a succession of images of the scene from measurements of the bolometric detector array, measurements of blind bolometers and CTIA gain.

[0008] The detection system's setting values ​​are sometimes unsuitable for conditions of implementation of the detection system, impairing the quality of the images of the scene formed, and therefore the quality of the information acquired on the object to be detected.

[0009] The measurement noise of the detection system may be too high in relation to the object to be detected.

[0010] The object to be detected may saturate the images of the scene, in particular by being too hot or too cold compared to the scene.

[0011] Furthermore, the setting values ​​of the detection system are generally defined at the factory at the beginning of the life of the detection system and therefore do not take into account the aging of the detection system and its electronics. Description of the invention

[0012] The present invention therefore aims to overcome all or part of the aforementioned drawbacks, to reduce measurement noise of a detection system observing a scene, to increase the accuracy of discrimination of objects of similar temperature in the scene, to enable the detection of the temperature of objects over an extended temperature range, to increase a measurement dynamic of the detection system with respect to the observed scene, to improve the robustness of the detection system to aging, to heating of the components of the detection system and to variations of the observed scene.

[0013] The present invention relates to a method for adjusting a detection system forming a succession of images of a scene, the detection system comprising a sensor equipped with an array of detectors sensitive to electromagnetic radiation from the scene, called "active", a plurality of detectors insensitive to said radiation, called "blind", as well as a reading circuit providing a response signal for each sensitive detector, the method comprising:

[0014] - the calculation of a representative value of the level of at least a portion of the signals of response;

[0015] - the determination of a difference between said representative value and a target value;

[0016] - the comparison of the deviation to a tolerance threshold; and

[0017] - according to the result of the comparison, the modification of a bias voltage applied to insensitive detectors, called "base voltage", in order to reduce the gap.

[0018] The set of response signals from the sensitive detectors forms a video stream comprising the succession of images of the scene.

[0019] The calculation of the representative value of the level of at least a portion of the response signals may include one of the following: the calculation of an instantaneous or time average value of at least a portion of the response signals, the calculation of an instantaneous or time median value of at least a portion of the response signals, or the calculation of another statistical function applied to at least a portion of the signals. answer.

[0020] Determining said deviation, comparing said deviation to the tolerance threshold and modifying the debasing voltage to reduce said deviation allows said representative value to be controlled to the target value by adapting the debasing voltage.

[0021] Modifying the base voltage according to the difference between the representative value of the level of at least part of the response signals and the target value makes it possible to reduce the risk of saturation of objects present in the scene in the image of the scene, in particular of objects whose temperature is far from the average temperature of the observed scene.

[0022] Adapting the base voltage to the measurement made by the sensor of the detection system makes it possible to adapt the dynamics of the sensor with respect to the scene observed during the implementation of the system, to make the system less sensitive to the aging of the system components, in particular less sensitive to the aging of the sensor, and to make the system robust to temperature variations during the implementation of the system.

[0023] The calculation of a representative value of the level of at least a part of the response signals may include the calculation of a voltage representative of voltage values ​​provided by the detection system for each active detector or the calculation of an intensity representative of intensity values ​​of the pixels of an image of the scene formed by the detection system.

[0024] The target value may correspond to a value between a minimum value of analog-to-digital conversion voltage of the detection system and a maximum value of analog-to-digital conversion voltage of the detection system.

[0025] The target value may correspond to a value close to the middle of the analog-to-digital conversion range of the detection system. For example, the target value may be between 0.9 and 1.1 times a value corresponding to the middle of the analog-to-digital conversion range of the detection system.

[0026] Choosing a value close to the middle of the conversion range allows for a similar margin between the minimum analog-to-digital conversion voltage of the system and the target value, and between the maximum analog-to-digital conversion voltage of the system and the target value. Thus, the risk of saturation of so-called "hot" objects is similar to the risk of saturation of so-called "cold" objects, as the choice of the target value takes into account the variation in luminance of the electromagnetic radiation of a black body as a function of the black body's temperature.

[0027] Advantageously, the target value is different from the value corresponding to the middle of the conversion range so as to increase the dynamics of the sensor towards high temperatures.

[0028] Said target value can be corrected according to an estimated operational measurement need in particular in order to correspond to a non-central value of the analog-to-digital conversion range of the detection system.

[0029] The estimated operational measurement requirement corresponds to a need to measure the temperature of objects whose temperature is far from the average temperature of the scene. For example, when the estimated operational measurement requirement corresponds to a need to measure the temperature of objects whose temperature is higher than the average temperature of the scene, said target value can be increased or decreased in order to reduce the risk of saturation of said objects in the scene image.

[0030] The value of the base voltage modification can be selected, at least as a function of said deviation, from a predetermined table of correspondence between a variation of the base voltage and a variation of the representative value of the level of at least a part of the response signals.

[0031] Thus, the modification of the base voltage is based on a previously determined value of correspondence between the variation of the base voltage and the variation of the representative value of the level of at least a part of the response signals.

[0032] Said correspondence table allows an accelerated convergence of the deviation towards a value below the tolerance threshold.

[0033] Said correspondence table is representative of a law of variation of the representative value of the level of at least a part of the response signals as a function of the value of the base voltage.

[0034] The modification of the base voltage can also be selected in said lookup table as a function of a current bias voltage of the active detector matrix of the sensor and / or as a function of a current value of trans-impedance capacitive amplification gain of the sensor and / or as a function of the temperature of the sensing system.

[0035] The said current values ​​correspond to the values ​​applied to the sensor of the detection system before the modification of the base voltage.

[0036] Said correspondence table can be established at the factory.

[0037] Said correspondence table can also be established for several values ​​of bias voltage of the active detector matrix of the sensor and / or for several values ​​of capacitive amplification gain trans-impedance of the sensor and / or for several temperatures of the sensing system.

[0038] Thus, the said correspondence table is established for several implementation parameters of the detection system, and allows an accelerated convergence of the deviation towards a value below the tolerance threshold.

[0039] The modification of the base voltage can be carried out between two successive image acquisitions of the detection system.

[0040] Thus, the appearance of an intensity gradient on the image of the observed scene is avoided, the intensity gradient corresponding to a sequential application, for example in the case of a "rolling shutter" type implementation in Anglo-Saxon terms of the detection system, of the base voltage to a plurality of blind detectors arranged by rows or by columns of the matrix of active detectors.

[0041] The method may include an initialization of the detection system during which a bias voltage value of the active detectors and / or a bias voltage value of the blind detectors and / or a trans-impedance capacitive amplification gain value of the sensor is selected at least as a function of a measurement of the system temperature.

[0042] System initialization corresponds, for example, to switching the system on.

[0043] The bias voltage value of the active detectors and / or the bias voltage value of the blind detectors and / or the trans-impedance capacitive amplification gain value of the sensor can be selected from a predetermined table of values ​​or be selected according to a predetermined function.

[0044] During system initialization, a current non-uniformity correction table in sensor gains can also be selected at least as a function of the system temperature measurement.

[0045] During system initialization, a current sensor additive non-uniformity correction table, also called an "offset" non-uniformity correction table in Anglo-Saxon terms, can also be selected at least according to the system temperature measurement.

[0046] The bias voltage value of the active detectors is selected according to the measured temperature of the system to improve a signal-to-noise ratio of the detection system by reducing sensor noise, in particular time-dependent sensor noise corresponding to variations in the response signals of the sensitive detectors when the detection system observes a fixed scene of constant temperature.

[0047] The signal-to-noise ratio of the detection system is improved with the constraint of encompassing, or at least matching, the estimated operational measurement requirement to the sensor dynamics. Thus, the bias voltage value of the active detectors, selected according to the measured system temperature, allows the system to be implemented for a wide range of detection system temperatures.

[0048] The bias voltage value of the blind detectors, selected according to the measured system temperature, allows for accelerated convergence of the deviation towards a value below the tolerance threshold, the current bias voltage of the blind detectors then being adapted to the system temperature. Indeed, when If the bias voltage value of the blind detectors is not selected based on temperature, the scene image could be completely saturated when calculating the representative value of the level of at least some of the response signals, thus slowing down the servoing of said representative value to the target value by adapting the base voltage.

[0049] The trans-impedance capacitive amplification gain value of the sensor selected according to the measured temperature of the system allows the use of a higher trans-impedance capacitive amplification gain value of the sensor at low temperature than at high temperature to reduce the impact of analog-to-digital conversion noise of the system.

[0050] The bias voltage value of active detectors and / or the bias voltage value of blind detectors and / or the trans-impedance capacitive amplification gain value of the sensor can also be selected according to the estimated operational measurement requirement.

[0051] The values ​​contained in said previously determined table of values ​​can be established for several temperatures of the detection system and by the use of several reference scenes in order to encompass, or at least match, the estimated operational measurement requirement to the sensor dynamics.

[0052] Said previously determined table of values ​​can be established by using a reference scene by temperature of the detection system, for example a constant temperature scene in particular equal to the temperature of the detection system, or can be established by using several reference scenes by temperature of the detection system, for example for several reference scenes at different temperatures.

[0053] The method may include, following the modification of the base voltage, in particular when said deviation is less than the tolerance threshold, the application of a non-uniformity correction table in additive of the sensor adapted to the modification of the base voltage.

[0054] The additive sensor non-uniformity correction table adapted to the base voltage modification can be equal to a current additive sensor non-uniformity correction table to which is added a term equal to the variation in the intensity level of the pixels of the scene images following the base voltage modification multiplied by a gain non-uniformity correction table.

[0055] The sensor additive non-uniformity correction table can be updated using a column-cutting and / or slitting algorithm.

[0056] The additive non-uniformity correction table for the sensor is thus adapted to the modified base voltage. Preferably, the additive non-uniformity correction table for the sensor is also adapted to the current bias voltage of the active detectors and the current value of the capacitive amplification gain transimpedance of the sensor. Indeed, the current additive non-uniformity correction table of the sensor, applied to the images of the scene before the modification of the base voltage, is no longer suitable for the new base voltage.

[0057] The present invention also relates to an adjustment device for a detection system forming a succession of images of a scene, the detection system comprising a sensor equipped with an array of detectors sensitive to electromagnetic radiation from the scene, called "active", a plurality of detectors insensitive to said radiation, called "blind", as well as a readout circuit providing a response signal for each sensitive detector, the device comprising: - a means of calculating a representative value of the level of at least a part of the response signals; - a means of determining the difference between said representative value and a target value; and - a means of comparing the deviation to a tolerance threshold; - a means of modifying a bias voltage applied to the insensitive detectors, known as "base voltage".

[0058] The device may include a temperature sensor configured to measure the temperature of the detection system and / or the sensor of the detection system.

[0059] The device may include a means for initializing a bias voltage value of the active detectors as a function of the temperature measured by the temperature sensor.

[0060] The device may include a means for initializing a bias voltage value of the blind detectors as a function of the temperature measured by the temperature sensor.

[0061] The device may include a means for initializing a trans-impedance capacitive amplification gain value of the sensor as a function of the temperature measured by the temperature sensor.

[0062] The device may include a means for selecting an estimated operational measurement need.

[0063] The device may include a means for applying a correction table for sensor non-uniformities, particularly in gain and additive form. This application means may be configured to estimate a correction table for the sensor's additive non-uniformities. Brief description of the drawings

[0064] Other objects, features and advantages of the invention will become apparent upon reading the following description, given solely as a non-limiting example and made with reference to the attached drawings on which:

[0065] [Fig.1] schematically illustrates a detection system comprising an adjustment device according to the invention;

[0066] [Fig.2] schematically illustrates an adjustment method according to the invention; and

[0067] [Fig.3] schematically illustrates the prior determination of value tables according to the adjustment process. Detailed description

[0068] Fig. 1 schematically represents a detection system 2 observing a scene 4, for example an uncooled infrared detection system, also called an IRNR detector.

[0069] The detection system 2 can be a monitoring and / or watch system.

[0070] The detection system 2 comprises an optical sensor 6 equipped with a matrix of Sensor 6 includes detectors 8 sensitive to electromagnetic radiation from scene 4, called "active" detectors 8, for example, an array of bolometric detectors. Sensor 6 also includes a plurality of detectors 10 insensitive to electromagnetic radiation from scene 4, called "blind" detectors 10, and a readout circuit 12 providing a response signal for each sensitive detector of sensor 6.

[0071] The detection system 2 provides, from the response signals of the active detectors 8, a succession of images of the scene 4 forming a video stream of the scene 4.

[0072] Electromagnetic radiation includes, in particular, infrared, even visible, terahertz or ultraviolet radiation.

[0073] The detection system 2 comprises an arrangement 14 of optical components for projecting the image of the scene 4 onto the array of active detectors 8 located in the focal plane of the arrangement 14 of optical components. Such a detection system 2 is known as a "focal-plane array" (FPA) in Anglo-Saxon terms.

[0074] Each active detector 8 includes a thermal resistance sensitive to electromagnetic radiation from scene 4. A current through the thermal resistance of each active detector 8 is representative of the heating of said active detector 8 due to electromagnetic radiation from scene 4.

[0075] Each blind detector 10 is a detector insensitive, or almost insensitive, to the radiation from the scene 4, generating a current, called compensation current, representative of the non-useful current in particular linked to heating of the elements of the detection system 2.

[0076] The blind detectors 10 are arranged so that several blind detectors 10 are associated with a row or column of the active detector matrix 8.

[0077] The difference between the current through the thermal resistance of each active detector 8 and the compensation current produced by the blind detectors 10 is calculated in order to extract a useful analog signal from scene 4.

[0078] This current difference is then integrated, in particular by filling a capacitance, over a period called "integration time", thus generating a voltage constituting a useful integrated analog signal associated with each active detector 8.

[0079] The sensor 6 reading circuit 12 allows multiplexing and sequentially providing the useful integrated analog signals associated with each active detector 8 of the matrix.

[0080] The reading circuit 12 includes a trans-impedance capacitive amplifier allowing a trans-impedance capacitive amplification gain of the sensor 6, called "CTIA gain", to be applied to the useful integrated analog signals.

[0081] The detection system 2 includes implementation electronics (not shown), referred to as "proximity electronics", equipped with at least one analog-to-digital converter configured to digitally convert the useful integrated analog signals to which the CTIA gain is applied in order to form the images of the scene 4, the images thus being formed from the signals of the active detectors 8.

[0082] The CTIA gain makes it possible in particular to reduce the contribution of digital conversion noise.

[0083] The proximity electronics is capable of biasing the active detectors 8 by providing a bias voltage for the active detectors 8. The proximity electronics is capable of biasing the blind detectors 10 by providing a bias voltage for the blind detectors 10, called the "base voltage".

[0084] The detection system 2 includes an adjustment device 16 comprising a temperature sensor 18 suitable for measuring the temperature of the detection system 2 or of the sensor 6 of the detection system 2. For example, the temperature sensor 18 is integrated into the sensor 6 of the detection system 2.

[0085] The adjustment device 16 includes a calculator 20 equipped with a memory.

[0086] The adjustment device 16 also includes a calculation means 22 of a value representative of the level of at least a portion of the response signals, a means 24 for determining a deviation between the representative value of the level of at least a portion of the response signals and a target value, a means 26 for comparing the deviation to a tolerance threshold, a means 28 for modifying the base voltage, a means 30 for initializing the bias voltage value of the active detectors 8 as a function of the temperature measured by the temperature sensor 18, a means 32 for initializing the base voltage value as a function of the temperature temperature measured by the temperature sensor 18, a means of initializing 34 the CTIA gain value as a function of the temperature measured by the temperature sensor 18, a means of selecting 36 an estimated operational measurement requirement and a means of applying 38 a gain non-uniformity correction table and an additive non-uniformity correction table of the sensor 6.

[0087] For example, the calculator 20 of the adjustment device 16 includes modules in the form of calculation codes, including a calculation module for a representative value of the level of at least a part of the response signals, a module for determining a deviation between the representative value of the level of at least a part of the response signals and a target value, a module for comparing the deviation to a tolerance threshold, a module for modifying the base voltage, a module for initializing the bias voltage value of the active detectors 8 as a function of the temperature measured by the temperature sensor 18, a module for initializing the base voltage value as a function of the temperature measured by the temperature sensor 18, a module for initializing the CTIA gain value as a function of the temperature measured by the temperature sensor 18,a module for selecting the estimated operational measurement requirement and a module for applying correction tables for non-uniformities in gains and additive values ​​of sensor 6. ,

[0088] The memory includes a pre-determined correspondence table between a variation of the base voltage and a variation of the representative value of at least part of the response signals, a pre-determined table of bias voltage values ​​of the active detectors 8, a pre-determined table of base voltage values, a pre-determined table of CTIA gain values ​​and pre-determined tables for correcting non-uniformities in gains and in additiveness of the sensor 6. Alternatively, said tables may be established in the form of functions.

[0089] The aforementioned predetermined memory tables are determined before the operational implementation of the detection system 2, particularly at the factory. The computer 20 includes program code instructions for implementing a method of adjusting the detection system 2.

[0090] Fig. 2 schematically represents the adjustment process of the detection system 2, the process corresponding in particular to the initialization and operational implementation of the detection system 2.

[0091] Optionally, the method for adjusting the detection system 2 begins with a step 40 for initializing the detection system 2. The initialization step 40 is performed when the detection system 2 is powered on or at regular intervals so that the detection system 2 adapts to the warming of its components and to variations in the observed scene 4.

[0092] For example, the temperature of the detection system 2 or of the sensor 6 of the detection system 2 is measured with the temperature sensor 18, then implementation parameters of the detection system 2 are selected according to the measured temperature, and then the implementation parameters are applied.

[0093] For example, it is estimated that the average temperature of the observed scene 4 is close to the temperature of the detection system 2 at its initialization.

[0094] Advantageously, the implementation parameters of the detection system 2 are also selected based on the estimated operational measurement requirement. The estimated operational measurement requirement corresponds to an estimated need to measure objects whose temperature is far from the temperature of the detection system 2 at its initialization. This requirement can be estimated manually by an operator using the operational measurement requirement selection means 36 or fixed by design of the detection system 2.

[0095] For this purpose, a current CTIA gain value is selected. For example, the CTIA gain value initialization means 34 selects from the predetermined CTIA gain value table a CTIA gain value suitable for the measured temperature and the operational measurement requirement to reduce digital conversion noise.

[0096] Next, a current value for the bias voltage of the active detectors 8 is selected. For example, the initialization means 30 for the value of the bias voltage of the active detectors 8 selects from a predetermined table of values ​​for the bias voltage of the active detectors 8 a value for the bias voltage of the active detectors 8 suitable for the measured temperature and the operational measurement requirement to reduce the measurement noise of the detection system 2, in particular the temporal noise of the sensor 6 corresponding to the variations in the response signals of the active detectors 8 when the detection system 2 observes a stationary scene. This measurement noise corresponds to the NETD, for "noise equivalent temperature difference", in Anglo-Saxon terms.

[0097] Then, a current non-uniformity correction table for the sensor 6 is selected, also known as "gain non-uniformity correction" in Anglo-Saxon terms. For example, the application means 38 of a non-uniformity correction table for the sensor 6 selects from the previously determined non-uniformity correction tables for the sensor 6 a non-uniformity correction table for the sensor 6 that is adapted to the measured temperature, and advantageously adapted to the current value of the bias voltage of the active detectors 8 and to the current value of the CTIA gain.

[0098] Then, a current base voltage value is selected. For example, the base voltage value initialization means 32 performs a selection in the previously determined table of base voltage values ​​of a base voltage value adapted to the measured temperature and the operational measurement requirement.

[0099] Alternatively, the selection order of the current CTIA gain value, the current bias voltage value of the active detectors 8, the current gain non-uniformity correction table of the sensor 6 and the current base voltage value may be different.

[0100] Next, an image acquisition step 42 of scene 4 is carried out during which the detection system 2 provides, from the response signals of the active detectors 8, a succession of images of scene 4.

[0101] Then, a step 44 of basing voltage control is carried out in order to adapt the measurement of the detection system 2 to the observed scene 4 containing, for example, hot and / or cold objects in motion.

[0102] We start by calculating a representative value of all or part of the response signals of the active detectors 8. For example, the representative value is calculated by the calculation means 22 of the detection system 2.

[0103] A deviation is determined between the representative value of all or part of the response signals of the active detectors 8 and a target value. For example, the deviation is calculated by the determination means 24 of the detection system 2.

[0104] The deviation is compared to a tolerance threshold. For example, the comparison is carried out by the comparison means 26 of the detection system 2.

[0105] If the deviation exceeds the tolerance threshold, the debase voltage is modified to reduce the deviation. For example, the debase voltage is modified by the modification means 28 of the detection system 2.

[0106] The target value corresponds, for example, to a voltage value between a minimum value of analog-to-digital conversion voltage of the detection system 2 and a maximum value of analog-to-digital conversion voltage of the detection system 2.

[0107] More specifically, the target value corresponds, for example, to a value close to the middle of the analog-to-digital conversion range of the detection system 2, advantageously eccentric more towards low scene temperatures than towards high temperatures, in order to meet the dynamic requirements of the sensor 6 with respect to the scene towards high and low temperatures.

[0108] For example, at the factory, for a detection system 2 temperature of +20°C, a minimum detection temperature of 0°C and a maximum detection temperature of +40°C are estimated, an object at the minimum or maximum detection temperature not to saturate the scene image 4 provided by the detection system 2. The current value of the bias voltage of the active detectors 8 and the Current CTIA gain values ​​allow the total temperature dynamics of scene 4, namely a temperature variation between the minimum value to be detected and the maximum value to be detected, here a total temperature dynamics of scene 4 of 40°C, to be encompassed, or at least matched, within the dynamics of sensor 6, namely the voltage range taken between the minimum value of the analog-to-digital conversion voltage of system 2 and the maximum value of the analog-to-digital conversion voltage of system 2. Furthermore, a target value close to the middle of the analog-to-digital conversion range of detection system 2 helps to avoid non-linear conversion phenomena for values ​​close to the minimum and maximum values ​​of the analog-to-digital conversion voltage of detection system 2.

[0109] For example, the minimum analog-to-digital conversion voltage of system 2 is 0V, the maximum analog-to-digital conversion voltage of system 2 is 2V, and the average temperature of the observed scene 4 is +20°C. If the calculated representative value for an image of scene 4 is 2V, objects above +20°C in scene 4 are saturated in the images of scene 4. If the tolerance threshold is 0.1V and the target level is IV, then the deviation exceeds the threshold. Therefore, the base voltage is adjusted so that the calculated representative value for subsequent images of scene 4 approaches 1.0V.

[0110] Alternatively, the control described here as being based on a voltage value can be based on a digital level value or any other quantity representative of the level of at least a portion of the response signals, for example, in least significant bits (LSB). Advantageously, the target value is corrected according to the estimated operational measurement requirement.

[0111] For example, it is estimated that it may be necessary to measure hotter objects, for example at +60°C. The target value is then modified so that the target value is significantly different from the average of the minimum value of the analog-to-digital conversion voltage of system 2 and the maximum value of the analog-to-digital conversion voltage of system 2 in order to reduce the risk of saturation of the hotter objects.

[0112] The calculation of the representative value of at least part of the response signals can be directly carried out by calculating the average intensity level of all or part of an image of the scene 4 provided by the detection system 2, each pixel of the image of the scene 4 corresponding to a response signal of an active detector 8.

[0113] Step 44 of basing voltage control allows the representative intensity level of the images of scene 4 to be controlled in order to reduce the risk of saturation of hot and cold objects in the images of scene 4.

[0114] Advantageously, the modification means 28 of the detection system 2 performs a selection of a base voltage modification value from the predetermined table of correspondence between a variation of the base voltage and a variation of the representative value of at least a part of the response signals.

[0115] Preferably, the modification of the debasing voltage is carried out between two successive image acquisitions of the detection system 2, i.e. in an image interframe or "dead time" between the acquisition of two successive images.

[0116] Performing the modification of the base voltage in the image interframe makes it possible to avoid the appearance of an intensity level gradient on the image of the scene 4 observed, in particular in the case of an implementation of the detection system 2 of the "rolling shutter" type in Anglo-Saxon terms, the base voltage being applied sequentially by row or by column according to the arrangement of the blind detectors 10.

[0117] When the difference between the representative value of the level of at least part of the response signals and the target value is less than the tolerance threshold, a step 46 of estimating a non-uniformity correction table is carried out by additively of the sensor 6, preferably facing the scene 4.

[0118] For example, during the estimation step 46, the detection system 2 carries out the implementation of a calibration process on a background that is as uniform as possible, in particular by closing a "shutter" or by defocusing an optic of the arrangement 14, or of a scene calibration process based on optical flow or based on statistical criteria for filtering textured elements of the scene.

[0119] Step 46 of estimating the additive non-uniformity correction table of sensor 6 performed with respect to scene 4 improves the image restoration quality of scene 4.

[0120] The gain non-uniformity correction table associated with the additive non-uniformity correction table of sensor 6 allows the computer 20 to perform the correction of the images from the active detectors 8 produced by sensor 6, digitized by the proximity electronics. This gain and additive correction of sensor 6 is called "2-point".

[0121] Preferably, once the additive non-uniformity correction table of sensor 6 has been estimated, steps 42 for image acquisition of scene 4 and 44 for servo control of the base voltage are carried out again, with the detection system 2 then being operational.

[0122] Advantageously, the non-uniformity correction table is regularly updated by adding the sensor 6.

[0123] Alternatively, the additive non-uniformity correction table adapted to the modification of the base tension is estimated from the current correction table additive non-uniformities of sensor 6. For example, the additive non-uniformity correction table adapted to the base voltage modification is equal to the current additive non-uniformity correction table of sensor 6 plus a term equal to the change in pixel intensity level of the scene images 4 following the base voltage modification, multiplied by the gain non-uniformity correction table. For example, the additive non-uniformity correction table of sensor 6 is updated using a column-cutting and / or row-cutting algorithm, particularly depending on the row or column architecture of the blind detectors 10.

[0124] Fig. 3 schematically represents steps for pre-setting the detection system 2, particularly at the factory, in order to determine and record in the memory of the computer 20 the tables previously determined at least as a function of the temperature of the detection system 2.

[0125] For different predetermined temperature segments of the detection system 2, a step 48 is performed of recording a value in the table of bias voltage values ​​of the active detectors 8. For several temperatures of the detection system 2 or of the sensor 6, and for several reference scenes, bias voltage values ​​of the active detectors 8 are established and recorded, selected to reduce the measurement noise of the detection system 2, in other words to reduce the NETD, while ensuring a minimum holding of the total temperature dynamics of the scene 4.

[0126] A step 50 is then performed of recording a value in the previously determined table of CTIA gain values. For several temperatures of the sensing system 2 or of the sensor 6, CTIA gain values ​​are recorded, selected to reduce the digital conversion noise of the sensing system 2.

[0127] Preferably, the highest CTIA gain values ​​allowing the total temperature dynamics of the scene 4 to be maintained will be selected and recorded simultaneously with the bias voltages of the active detectors 8 so as to jointly reduce the measurement noise and digital conversion values ​​of the detection system 2.

[0128] Advantageously, said tables of values ​​of bias voltages of the active detectors 8, or active pixels, and of CTIA gains are established for the highest temperature value of each predetermined temperature segment of the detection system 2, the highest temperature value being the value of strongest constraint in terms of total temperature dynamics of scene 4 to be held of the predetermined temperature segment of the detection system 2.

[0129] For example, for each high temperature in each temperature segment of the detection system 2, the available CTIA gain is first selected by The design of the weakest detection system 2 is followed by varying the bias voltage of the active detectors 8. For each bias voltage value of the active detectors 8, at least two distinct, known, uniformly constant temperature scenes are imaged. This allows, on the one hand, the evaluation of the noise of the detection system 2 and, on the other hand, the conversion of the desired or estimated total temperature dynamics of the scene 4 to the sensor space. The bias voltage value of the active detectors 8 that minimizes the noise of the detection system 2 is recorded. The selected bias voltage value of the active detectors 8 also allows the desired or estimated total temperature dynamics of the scene 4 to be encompassed, or at least matched, to the dynamics of the sensor 6.We then test whether a higher CTIA gain allows us to maintain the desired or estimated total temperature dynamic range of scene 4 in order to further reduce digital conversion noise. The highest possible CTIA gain while maintaining the total temperature dynamic range of scene 4 is recorded in the CTIA gain table.

[0130] For example, for a sensor 6 temperature of +20°C at the top of a segment [0°C, 20°C], the total temperature dynamic range of scene 4 is estimated to be from 0°C to +40°C, i.e., no saturation for scene elements between 0°C and 40°C. Two acquisitions in front of two distinct reference scenes presented to the detection system 2 allow us to deduce the range occupied by the excursion from 0°C to 40°C at the output of sensor 6, the excursion being, for example, expressed in volts or digital level. If this excursion does not fall within the dynamic range of sensor 6, corresponding to the difference between the maximum value of the analog-to-digital conversion voltage of the detection system 2 and the minimum value of the analog-to-digital conversion voltage of the detection system 2, then the tested biasing of the active bolometers is unsuitable.If this excursion is compatible with the dynamic range of sensor 6, then the noise is calculated on the reference scenes presented to the detection system 2. Once all the polarization values ​​of the active pixels have been tested, the value ensuring the lowest noise is retained, recorded in the table of active pixel polarizations, and associated with the segment [0°C, 20°C]. Finally, for the selected active pixel polarization, the available CTIA gains are evaluated to determine which ones can handle the total temperature dynamic range of scene 4, and the highest value is selected and recorded in the previously determined table of CTIA gain values.

[0131] This logic for finding the polarizations of the active pixels and the CTIA gain per predetermined temperature segment of the detection system 2 can be repeated at the factory for each detection system 2 to be produced or, more simply, be considered stable and pre-established from a few detection systems 2 to avoid having to re-estimate at each detection system 2, as the procedure is lengthy.

[0132] For each predetermined temperature segment of the detection system 2, a step 52 is also performed to record the polarization of the blind detectors 10, or basing pixels, in a predetermined table of basing voltage values. For several temperatures of the detection system 2 or of the sensor 6 and for several reference scenes, basing voltage values ​​are selected and recorded to adapt the representative value of at least a portion of the response signals to a target value, the target value being determined to allow for the maintenance of a high scene dynamic range and the maintenance of a low scene dynamic range, that is, to avoid saturating an estimated high temperature scene and to avoid saturating an estimated low temperature scene.For example, for each temperature of the detection system 2, with a fixed bias voltage of the active detectors 8 and a fixed CTIA gain value, a scene of known constant temperature is imaged, the value of the base voltage is varied, and the value of the base voltage is recorded such that the representative value of at least a part of the response signals corresponds to the target value.

[0133] During the initialization of the detection system 2, the application of a base voltage value adapted to the temperature of the detection system 2 or of the sensor 6 makes it possible to accelerate the convergence of the deviation towards a value below the tolerance threshold.

[0134] Advantageously, one can record not a single base voltage value applicable to the predetermined temperature segment of the detection system 2, but a function applicable to said predetermined temperature segment, for example polynomial, in particular dependent on the temperature of the detection system 2.

[0135] For each temperature segment of the detection system 2, a step 54 is performed to record a correlation value between a variation in the base voltage and the associated variation in the representative value of at least a portion of the response signals. For several temperatures of the detection system 2 or of the sensor 6, for each pair of bias voltage of the active detectors 8 and CTIA gain, the variation in the representative value of at least a portion of the response signals is recorded as a function of the variation in the base voltage. Used by an embedded control algorithm in the computer 20 of the detection system 2, this correlation value makes it possible to accelerate the convergence of the deviation towards a value below the tolerance threshold.

[0136] Advantageously, one can record not a single correspondence value applicable to the temperature segment of the detection system 2 but a function applicable to the predetermined temperature segment, for example polynomial, in particular dependent on the temperature of the detection system 2.

[0137] Finally, a step 56 is carried out of recording the non-uniformity correction tables in gain and additive of the sensor 6. Preferably for several bias voltage values ​​of the active detectors 8, preferably for several base voltage values ​​for the additive table, preferably for several CTIA gain values ​​and preferably for several temperature values ​​of the detection system 2 or of the sensor 6, a non-uniformity correction table in additive of the sensor 6 and a non-uniformity correction table in gain of the sensor 6 are established and recorded, in order to compensate for the spatial non-uniformities of the images of the scene 4 formed by the detection system 2 by the computer 20. The non-uniformity correction tables in additive and in gain of the sensor 6 are, for example, established by carrying out two acquisitions on uniform scenes with two distinct temperatures.

[0138] Alternatively, the order of steps 54 of recording a correspondence value between a variation of the base voltage and the variation of the representative value of at least part of the response signals and 56 of recording the tables of correction of non-uniformities in gains and in additive of the sensor 6 could be reversed.

[0139] Advantageously, one can search for the optimum of the polarizations of the active pixels at several temperatures of the detection system 2 or of the sensor 6 within a predetermined temperature segment of the detection system 2. Then, not a single value of the polarization voltage of the active pixels applicable to the predetermined temperature segment of the detection system 2, but a function applicable to the predetermined temperature segment, for example polynomial, in particular dependent on the temperature of the detection system 2.

[0140] Advantageously, the tables established for a predetermined segmentation of temperatures of the detection system 2, for example for temperatures between -40°C and -10°C, between ❖10°C and +40°C and between +40°C and +70°C, can be established in the form of functions in particular in order to reduce the amount of data recorded in the memory of the computer 20.

Claims

Claims

1. Method for adjusting a detection system (2) forming a succession of images of a scene (4), the detection system (2) comprising a sensor (6) provided with a matrix of detectors sensitive (8) to electromagnetic radiation coming from the scene (4), called "active", a plurality of detectors insensitive (10) to said radiation, called "blind", as well as a reading circuit (12) providing a response signal for each sensitive detector (8), characterized in that it comprises: - the calculation of a value representative of the level of at least part of the response signals; - the determination of a difference between said representative value and a target value; - the comparison of the difference with a tolerance threshold; and - according to the result of the comparison, the modification of a bias voltage applied to the insensitive detectors (10) called "base voltage", in order to reduce the difference.

2. The adjustment method according to claim 1, wherein the target value corresponds to a value between a minimum analog-to-digital conversion voltage value of the detection system (2) and a maximum analog-to-digital conversion voltage value of the detection system (2).

3. The adjustment method according to claim 2, wherein the target value corresponds to a value close to the middle of the analog-to-digital conversion range of the detection system (2).

4. Method according to any one of claims 1 to 3, in which the value of the modification of the basing voltage is selected, at least as a function of said difference, in a previously determined table of correspondence between a variation of the basing voltage and a variation of the value representative of the level of at least part of the response signals.

5. Method according to any one of claims 1 to 4, in which the modification of the basing voltage is carried out between two successive image acquisitions of the detection system (2).

6. A method according to any one of claims 1 to 5, comprising an initialization of the detection system (2) during which a bias voltage value of the active detectors (8) and / or a bias voltage value of the blind detectors (10) and / or a trans-impedance capacitive amplification gain value of the sensor (6) is selected at least as a function of a measurement of the temperature of the detection system (2).

7. A method according to claim 6, wherein the bias voltage value of the active detectors (6) and / or the bias voltage value of the blind detectors (10) and / or the trans-impedance capacitive amplification gain value of the sensor (6) is also selected based on an estimated operational measurement need.

8. Method according to one of claims 6 and 7, in which the bias voltage value of the active detectors (6) and / or the bias voltage value of the blind detectors (10) and / or the trans-impedance capacitive amplification gain value of the sensor (6) is selected from a previously determined table of values, the values contained in said previously determined table of values being established for several temperatures of the detection system (2) and by the use of several reference scenes.

9. Method according to any one of claims 1 to 8, comprising, following the modification of the base voltage, in particular when said deviation is less than the tolerance threshold, the application of a correction table for non-uniformities in additive of the sensor (6) adapted to the modification of the base voltage.

10. Method according to claim 9, in which the correction table for additive non-uniformities of the sensor (6) adapted to the modification of the basing voltage is equal to a current correction table for additive non-uniformities of the sensor (6) to which is added a term equal to the variation in the intensity level of the pixels of the images of the scene (4) following the modification of the basing voltage multiplied by a correction table for gain non-uniformities.

11. Method according to claim 10, in which the correction table of additive non-uniformities of the sensor (6) is updated using a decolumnizing and / or delineating algorithm.

12. Adjustment device (16) for a detection system (2) forming an image of a scene (4), the detection system (2) comprising a sensor (6) provided with a matrix of detectors sensitive (8) to electromagnetic radiation coming from the scene (4), called "active", of a plurality of detectors insensitive (10) to said radiation, called "blind", as well as a reading circuit (12) providing a response signal for each sensitive detector (8), characterized in that it comprises: - a means (22) for calculating a value representative of the level of at least part of the response signals; - a means (24) for determining a difference between said representative value and a target value; and - a means of comparing (26) the deviation to a tolerance threshold; - a means (28) for modifying a bias voltage applied to the insensitive detectors (10), called “base voltage”.

Images