Method for improving sensor performance in a system comprising a plurality of platforms, each incorporating at least one sensor and one electronic processing module
By sharing and evaluating sensor information across platforms and adjusting configurations based on environmental conditions, the solution optimizes sensor performance and detection quality in diverse environments.
Patent Information
- Application Number
- FR2023011297
- Authority / Receiving Office
- FR · FR
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-10-19
- Publication Date
- 2025-04-25
- Estimated Expiration
- 2043-10-19
AI Technical Summary
Existing sensor systems lack the ability to optimally utilize the performance of diverse sensors across multiple platforms, particularly in environments with varying conditions, leading to suboptimal mission outcomes due to inadequate adjustment and adaptation to environmental factors.
A process and system that enhances sensor performance by sharing and evaluating information between platforms, adjusting sensor configurations based on the performance of other sensors, and incorporating environmental conditions to improve detection quality and relevance.
The solution enables improved sensor performance by optimizing sensor settings and adapting to environmental conditions, enhancing detection accuracy and quality across decentralized platforms.
Smart Images
Figure 00000018_0000 
Figure 00000019_0000 
Figure 00000020_0000
Abstract
Description
Title of the invention: Method for improving the performance of sensors in a system comprising a plurality of platforms each carrying at least one sensor and an electronic processing module Technical field
[0001] The invention lies in the field of platform systems carrying sensors operating in a common area of interest. The invention applies in particular to the case where all of these sensors constitute sensor chains, also called sensor suites. The sensors considered comprise radars, for example having GMTI / SAR and / or ISAR modes, EO / IR optronic balls and / or radar detectors. These systems are preferably airborne, but use on the ground on mobile vehicles is possible.
[0002] Such systems are found in the field of combat aircraft, maritime combat missions, satellite acquisitions or on-board aeronautical missions for example. Prior art
[0003] Sensors on platforms perform a set of actions essential to the success of the mission. Their functions cover very diverse mission objectives, ranging from image capture for intelligence to the survival and security functions of their platform or platforms. Faced with this set of functions, a system operator attempts to make the best possible use of the set of available sensors in order to successfully complete the mission and optimize its results, i.e. to achieve the main objectives and accomplish as many opportunity objectives as possible.
[0004] The optimal use of sensors to fulfill a mission requires knowledge of the performance of these sensors in order to make the right decisions. The performance of these sensors depends on parameters such as the physical location of the sensors on the airborne platforms, the technologies used but also the environment in which the sensors are used (weather, etc.). Classically, performance models based on equations are defined.
[0005] Radars are also known to have self-adaptive modes to improve clutter rejection. Usually, depending on the target sizes and their kinematics, there are preset modes to get the best out of the sensor. However, this is not enough.
[0006] there is a need to improve the quality and relevance of the detections carried out at using such systems. Summary of the invention
[0007] To this end, according to a first aspect, the present invention describes a method for improving the performance of sensors in a system comprising a plurality of platforms each carrying at least one sensor and one electronic processing module, the platforms being connected by remote transmission links, said method comprising the following steps: - capturing a first image by a first sensor on a first platform of the system and programmed according to a first configuration of the first sensor; - determination, by a first processing block on the first platform, of a first set of information, as a function of the first captured image, said set of information comprising at least one piece of information on the location of an object detected on said first image and one piece of information indicating the size of said object; - remote transmission of said first set of information from the first platform to a second platform; - reception, by the second platform comprising a second sensor and a second processing block, of said first set of information and determination, by the second processing module, of a first configuration of the second sensor as a function of the information in the set of information; - capturing a second image comprising said object by the second sensor programmed according to said first configuration of the second sensor determined; - determination, by the second processing block, of a second set of information, as a function of the second captured image, said second set of information comprising at least one piece of information on the location of said object detected on said second image and one piece of information indicating the size of said object; - evaluation, based on the first and second sets of information, by one of the first and second processing modules and selection, based on this evaluation, of one of the first and second sets of information; - triggering the programming, according to a second configuration depending on the selected set of information, of that of the first and second sensors not programmed according to said selected set of information selected and capture of a new image by said sensor programmed according to said second configuration.
[0008] The invention allows the improvement of sensor performance, in particular in a context of decentralized collaborative platforms and in particular allows the improvement of the quality of service of optics (EO / IR) by the use of RADAR means (GMTI & SAR) by considering common analysis zones. Targeted measurements and statistical measurements can be carried out by zone to take into account the specificities of the respective measurement conditions.
[0009] The measurements made by the sensors are used to enable better identification of the object, for example either by contour measurement or by cast shadow measurement. Depending on the measurements made by one of the sensors (depending on their quality for example) in a platform, the technical parameters of at least one other of the sensors, located on another platform, are adjusted to increase its acquisition performance.
[0010] The invention therefore proposes an automatic solution for adjusting the sensor according to the parameters and acquisitions of another sensor.
[0011] In embodiments, such a method will further comprise at least one of the following features:
[0012] - said set of information further comprises at least one piece of information relating to image capture among: sensor aperture, resolution, distance to object, object classification;
[0013] - the acquisition spectra of the first and second sensors are separated, by example one of the first and second sensors is of the radar type and the other of the first and second sensors is of the EO / IR type;
[0014] - the steps of the method are implemented cyclically between distinct pairs of system platforms;
[0015] - said method further comprises triggering the programming, according to a third configuration depending on the selected set of information, a sensor of a platform distinct from the first and second platforms and capture of a new image by said sensor programmed according to said third configuration.
[0016] According to another aspect, the invention describes a system comprising a plurality of platforms each carrying at least one sensor and one electronic processing module, the platforms being connected by remote transmission links, adapted to implement the following operations: a first sensor on a first platform of the system and programmed according to a first configuration of the first sensor is adapted to capture a first image by; a first processing block on the first platform, is adapted to determine a first set of information, based on the first captured image, said set of information comprising at least one piece of information on the location of an object detected on said first image and one piece of information indicating the size of said object; the first platform is adapted to transmit said first set of information to a second platform; said second platform comprising a second sensor and a second processing block, is adapted to receive said first set of information and the second processing module is adapted to determine a first configuration of the second sensor as a function of the information in the set of information; the second sensor is adapted to, programmed according to said first configuration of the second sensor determined, capture a second image comprising said object; the second processing block is adapted to determine a second set of information, as a function of the second captured image, said second set of information comprising at least one piece of information on the location of said object detected on said second image and one piece of information indicating the size of said object; one of the first and second processing modules is adapted to carry out an evaluation, as a function of the first and second sets of information, and to select, as a function of this evaluation, one of the first and second sets of information; said system being adapted to trigger the programming, according to a second configuration depending on the selected set of information, of that of the first and second sensors not programmed according to said selected set of information and to trigger the capture of a new image by said sensor programmed according to said second configuration.
[0017] In embodiments, such a system will further comprise at least one of the following features:
[0018] - said set of information further comprises at least one piece of information relating to image capture among: sensor aperture, resolution, distance to object, object classification;
[0019] - the acquisition spectra of the first and second sensors are separated, by example one of the first and second sensors is of the radar type and the other of the first and second sensors is of the EO / IR type;
[0020] - said system is adapted to cyclically implement said operations between distinct pairs of system platforms;
[0021] - said system is further adapted to trigger the programming, according to a third configuration depending on the selected information set, a sensor of a platform distinct from the first and second platforms and capture of a new image by said sensor programmed according to said third configuration. Brief description of the drawings
[0022] The invention will be better understood and other characteristics, details and advantages will appear more clearly on reading the following description, given without limitation, and thanks to the appended figures, given by way of example.
[0023] [Fig-1] [Fig.l] schematically represents a system of platforms of sensors embedded in one embodiment of the invention;
[0024] [Fig.2] [Fig.2] represents steps of a performance improvement process forms of sensors embedded in an embodiment of the invention;
[0025] [Fig.3] [Fig.3] illustrates the steps of calculating the contours of an object in an image using the Sobel filter;
[0026] [Fig.4] Figure 4 is a representation of the apparent surface '
[0027] [Fig.5] [Fig.5] illustrates a shadow calculation of an object.
[0028] Identical references may be used in different figures when they designate identical or comparable elements. Description of the embodiments
[0029] [Fig.l] schematically represents a system 1 of on-board sensor platforms in an embodiment of the invention. The system 1 comprises n platforms 10, named PF1, ..., PFn, which are mobile or not, with n greater than or equal to 2. According to the embodiments, one or more platforms are of the aircraft or land vehicle or maritime vessel type, satellites, land mobile elements. For example, in the case considered, the platforms are drones.
[0030] Each PFi platform 10, i = 1 to n, comprises at least one sensor 20, CPTi, and a processing module 30, TRTi. According to the embodiments, one or more sensors 20 are of the radar, lidar or sonar type or of the optical imaging type, for example of the ElectroOptical or Infrared (EO / IR) or hyperspectral (HSI) type. A radar sensor 20 is for example of the GMTI (Ground Moving Target Indicator) type, SAR (Synthetic Aperture Radar) type, ISAR (Inverse Synthetic Aperture Radar) type, etc.
[0031] Each sensor 20, CPTi, is adapted to carry out measurements of physical quantities depending on its technology (echoes for radars, sonars, lidars; brightness, temperature, etc. for others) and to convert the measurements into an electrical output signal and provide them as input to the processing module 30, TRTi.
[0032] The processing module 30, TRTi, is adapted to carry out algorithmic processing based on this electrical signal provided as input. It is particularly adapted to implement the steps described below with reference to [Fig.2] and which are its responsibility. In one embodiment, the processing module 30, TRTi comprises a memory and a processor (not shown). The steps mentioned are for example implemented by the execution on the processor of software instructions stored in the memory. Alternatively, they are implemented by dedicated hardware, typically a digital integrated circuit, either specific (ASIC) or based on programmable logic (for example FPGA / Field Programmable Gate Array).
[0033] In the case considered, the sensor 20 CPT1 is of the RADAR type in SAR mode and the sensor 20 CPT2 is of the EO / IR type. The sensors 20 target a common area of interest 60, Z.
[0034] There are teletransmission links (not shown), wired or wireless, between the platforms 10, in particular between any two platforms involved in the method described below.
[0035] A method 100 for improving the performance of the on-board sensors of the system 1 is now described in an embodiment of the invention, with reference to [Fig.2],
[0036] In a step 100_l, a primary detection is carried out by the sensor 20 CPT1 (RADAR) following a phase of analysis of the echoes received following a radar emission: the presence of an object is determined on the detection zone Z 60 as a function of these echoes by the sensor 20 CPT1.
[0037] In a step 101, the sub-steps 101a, 101b, 101c, 101b are then implemented within the platform PF1.
[0038] In sub-step 101a, the radar 20 CPT1 performs SAR imaging on the detection zone Z 60. In sub-step 101b, from the SAR image obtained, the processing module 30 TRTI determines a location of the object (in the form of geographical coordinates) and performs algorithmic processing to determine contours of the object on the SAR image. In sub-step 101c, as a function of the determined contours, the processing module 30 TRTI then determines the dimensions of the object, for example here in the form of the calculation of the Equivalent Surface of the object and deduces therefrom the pair of determined data {Location; dimensions of the object (here Equivalent Surface)}.
[0039] In sub-step 1Old, the processing module 30 TRTI transmits to the processing module 30 TRT2 the pair of determined data {Location; dimensions of the object (here Equivalent surface)}.
[0040] In a step 102, the sub-steps 102a, 102b, 102c are then implemented within the platform PF2.
[0041] On receipt of the pair of data {Location; Equivalent surface} transmitted, in sub-step 102a, the processing module 30 TRT2 automatically calculates the line of sight of the 20 CPT2 sensor (EO / IR) and the FOV (Field of View, which includes the resolution calculation), depending on this location and the dimensions of the object (here Equivalent Surface) so that the entire object is included in the FOV. The values of the EO / IR image capture parameters of the 20 CPT2 sensor are adjusted so as to respect this line of sight and this FOV, and an EO / IR image of the object is then captured by the 20 CPT2 sensor.
[0042] In sub-step 102b, from the EO / IR image obtained, the processing module 30 TRT2 determines a location of the object (in the form of geographical coordinates, performs an algorithmic processing to determine the contours of the object on the EO / IR image. In sub-step 102c, as a function of the determined contours, the processing module 30 TRT2 then determines the dimensions of the object, for example here in the form of the calculation of the Equivalent Surface of the object.
[0043] These same calculations as those previously carried out in step 101 thus make it possible at the end of step 102c to determine a new value of the location and a new value of the dimensions, i.e. a pair {Location; Equivalent surface}.
[0044] A set 200 of steps is then implemented, comprising steps 201 and 202. In the embodiment considered, this set of steps is implemented by the processing module TRT2 30 of the platform PF2. In another embodiment, it is carried out by the processing block of the platform PF, TRT1 30 after remote transmission by the platform PF2 to the platform PF1 of the data resulting from the processing of the platform PF2.
[0045] In step 201, the pairs {Location; Equivalent Surface] calculated respectively by the processing modules 30 TRT1, TR2, i.e. respectively {LocationCPTi; Equivalent SurfaceCPTi}, {LocationCPT2; Equivalent SurfaceCPT2}. are compared.
[0046] Then in the performance evaluation step 202, the processing module TRT2 30 evaluates, based on this comparison, which of the values of the components of the two pairs are optimal, i.e. which corresponds to better localization and / or better image resolution (sensor performance analysis).
[0047] Here, the term “best location” means the location provided by the sensor with the best precision (it is generally provided by the manufacturer who manufactures the equipment; it generally varies from 1 to 10 degrees relative to the actual position of the target depending on the equipment used).
[0048] The performance evaluation includes updating the values of the “location” and “equivalent surface” components by retaining for each component, the optimal value in terms of location precision (location component) and resolution of the observed object (equivalent surface component). Following the comparison of the results (step 201), the pair {Location, Equivalent Surface] is updated (and then stored in a memory) if at least the value of one of the two components from CPT2 is evaluated as "better" than those calculated by CPT1. For example, if CPT2 gives a better resolution or Equivalent Surface of the object, but not a more precise location, the new Pair becomes {LocalizationCPi; Equivalent SurfaceCPT2}-
[0049] Thus, it is determined between the two locations LocationCPTi and LocationCPT2 which corresponds to the best location accuracy and it is evaluated whether the new measurement of the Equivalent Surface means an improvement in the resolution and / or the size of the object. For example, CPT1 can give the equivalent surface but the resolution can be improved by CPT2.
[0050] The resolution is considered better when the number of pixels characterizing the same object, obtained following the detection of contours, is greater.
[0051] The best location accuracy corresponds to the best sensor accuracy between those of CPT1 and CPT2, which are known.
[0052] In one embodiment, this comparison / evaluation is carried out on more than two pairs, for example N pairs coming from N platforms, with N >2.
[0053] The optimal torque is: - returned to the processing module 30 of the one of the two platforms PF1, PF2 which was not set according to the optimal torque, to determine a programming adaptation of the CPT sensor 20 according to this optimal torque and to implement the programming adaptation of the CPT sensor 20 to thus have a better quality of the image; for example, here the optimal torque is the torque determined in step 102c in the platform PF2, this torque is transmitted to the platform PF1 for implementation of the corresponding settings on the sensor CPT1; in particular if the optimal torque is that of the platform PF2, it is provided to the platform PF1 to be taken into account for the programming; and / or - provided to (at least) a third platform to perform its sensor programming based on this optimal torque.
[0054] For example, if the comparison shows that the torque provided by CPT1 is better, we keep the precision / resolution proposed by CPT1 (which will be the one given to CPT3, if applicable), are kept unchanged by CPT1.
[0055] In one embodiment, this process is implemented cyclically between distinct pairs of platforms from all the platforms constituting the system 1. Thus, for example, the optimal pair is provided as input to at least one PFi platform, i being an integer between 2 and n and a step 10i, similar to step 102, is implemented in the PFi platform, followed by a set of steps 200. This allows the PFi platform to benefit from the improvements determined using the two PFI platforms, PF2.
[0056] In one embodiment, instead of the pair {location; dimensions of the object], a set of data is transmitted by the platform PF1 in step 101, then considered in the set 200 of steps comprising, in addition to the location and the dimensions, one or more additional elements having an impact on the adjustment parameters of the sensors 20, for example the sensor aperture, the resolution, the distance to the object, classification of the object.
[0057] In the above, a radar image is first taken. In another embodiment, an EO / IR image is first taken, then a radar image. In another embodiment, the sensors CPT1 and CPT2 capture images of the same type, i.e. radar type or EO / IR type).
[0058] The performance of EO / IR sensors is highly dependent on the environment and shooting conditions. These conditions can change with:
[0059] - weather: day / night, sunshine, clouds, fog, mist, snow / frost, rain, wind ;
[0060] - the shooting conditions, in particular the interaction of the wearer's movements, the height and distance to the observed target;
[0061] - the adjustment conditions of the optical sensor, possibly the added filters can affect the quality of videos and therefore the possibilities of detection, analysis and automatic recognition.
[0062] These conditions do not have the same effects on electromagnetic sensors, in particular day / night conditions, sunlight, etc. Unlike EO / IR sensors, the radar in SAR mode illuminates the scene. It is therefore less dependent on external conditions. It undergoes attenuation depending on the rain, but it is also capable of measuring the amount of rain and estimating performance in these conditions. Transmission and reception are co-located (except in the special case of bi-static and passive radars). The shadows cast by the radar are therefore deterministic.
[0063] A method according to the invention is relevant for processing acquisitions in a detection zone Z common to heterogeneous platforms / sensors. When the sensors 20 CAPT1 on PF1 and CAPT 2 on PF2 point towards this zone, they each make it possible to recover, in addition to the information specific to the detections, secondary information which corresponds to degradations in the detection quality caused by the environment (for example, in maritime surveillance, the SAR can estimate a wind speed as a function of the detection noise caused by the waves).
[0064] In one embodiment, by exploiting, or even merging, the secondary information from at least one of the types of sensors, or even both types of sensors, it is determined, by any of the processing modules 30, for example TRT1, TRT2, a state of the weather environment of zone Z. When the weather conditions of the zone are known (or estimated based in particular on secondary information from a sensor, for example the radar sensor), it is possible to calculate a performance forecast (estimation of degradations due to weather conditions) for another, for example 1 EO / IR sensor, as well as the optimal settings (or treatments) associated with it.
[0065] Secondary radar sensor information includes, for example, detection noise caused by waves and target velocity vectors. They are measured by the radar sensor, for example. Knowing the environment, it is possible to correct certain undesirable effects on image quality.
[0066] Secondary information from the EO / IR sensor includes, for example, the measurement of the outside temperature and knowledge of the deformation of the image as a function of this temperature; the sensor can use correction curves on the image if available.
[0067] As an illustration, the radar sensor has a strong speckle degrading the equivalent surface of the object. Similarly, if there is strong heat, the EO / IR image is strongly degraded due to atmospheric deformations. To improve the accuracy of the determination of the equivalent surface of the object based on the measurements made on the captured image, it is further determined based on a calculation of the dimensions of the shadows cast on the image.
[0068] In the context of the method represented in [Fig.2], this action can be carried out in the processing phase operated by TRT1 30, respectively TRT2 30 according to the sensor CPT1, respectively CPT2.
[0069] The action triggers on the measurement or knowledge of the external temperatures which statistically must modify the shape of the object. Similarly, for the RADAR knowing the angle of incidence of the acquisition, we can presuppose knowing the degradation on the SAR image.
[0070] The settings and / or the location / dimensions pair of CPT1 transmitted to the sensor of the other platform, CPT2, are used to program the CPT2 sensor automatically and possibly improve its detection quality. A feedback loop then analyzes the new detection quality in order to verify that it has been improved as expected or, if this is not the case, to use the new secondary information obtained in order to improve the acquisition of data from CPT1 (radar) of the area and thus reproduce the process of optimizing the settings.
[0071] The advantage of the solution is that it can transmit the characteristics of the targets (size, shape, etc.) to each drone to another drone, which will allow the programming of the on-board sensor to be optimized. The solution makes it possible to increase the autonomy of the on-board systems and to automatically optimize performance. sensors on remote platforms, which is particularly useful in areas where human-system interactions are not systematically available.
[0072] In one embodiment of the invention, the sensor chain 20 of the system 1 is adapted to optimize the sensors and estimate the possible performances according to the operational constraints (climatic, interference, possible breakdowns, under-optimization of the settings, etc.).
[0073] Depending on the characteristics of the target sought and the estimated performances, it is determined, in one embodiment, by any of the processing modules 30, whether the detection mission is possible or utopian.
[0074] Methods for calculating object dimensions
[0075] Methods for calculating dimensions of the object are now described which are for example used by the processing modules in steps 101, 102, or 10i, i between 1 and n.
[0076] In order to be able to define the characteristics of the object and improve its detection, a post-processing of the acquired image is therefore applied, as indicated with reference to [Fig.2], to allow the contours of the object considered to be defined. This contour calculation phase makes it possible to improve the quality of the image by freeing itself from environmental interference.
[0077] In the literature, most of the treatments for contour detection are based on edge detection, and more particularly on the Sobel filter, and rely on the study of the gradient of the image: we are in the presence of a contour if in one of the two directions of the image we have a strong variation in the amplitude of the pixels.
[0078] A Sobel filter is generally applied to calculate the contours (thus also making it possible to overcome aberrations due to the sensors: chromatic aberrations, thermal aberrations, speckle); then hysteresis filtering of the contours is implemented. A Harris detection algorithm then detects the objects of interest (it makes it possible to detect the corners in an image); a measurement of the dimensions of the detected objects of interest is then implemented.
[0079] In the context of the method 100, the size of the object of interest is estimated from the contours and the resolution of the image taken by a sensor. This size information makes it possible to automatically program the pointing and opening of another sensor.
[0080] [Fig.3] illustrates describes steps for calculating the contours of objects of interest from of an EO / IR image (the steps are similar in the case of a SAR image).
[0081] Calculating the contour allows the observed object to be isolated from its environment in order to have an effective measurement of: its length, Lohj, its width lobj and its apparent surface, 7^' • The apparent surface of an object of interest is calculated from the number of pixels (X,Y) and the radial resolution of the image, ôr.
[0082] Thus:
[0083] Lobj = X*Ôr(l)
[0084] and
[0085] lohj = Y*ôr&
[0086] Figure 4 illustrates this calculation. The apparent surface, Tobj' along an axis corresponds to the surface of the white rectangle.
[0087] Calculation of cast shadows
[0088] In the same way, to improve the performance of each sensor, it is possible to know the size and shape of the object based on the shadow projected by each sensor. Indeed, as indicated above, speckle effects for RADAR or image deformations based on weather conditions (for example, atmospheric deformations) can have consequences on contour measurements (image too blurred, poor measurement of the object by SAR, etc.). To improve the measurement of the object, it is possible to measure the shadow cast by the object by each sensor, thus allowing a better appreciation of the object of interest.
[0089] As a result, processing of the shadows also makes it possible to validate the heights of the targets (buildings, vehicles, etc.) and allows a correlation of the 3D information between the zenithal lighting of the optics and the shadows cast by the radar (radial relative to the lighting): see [Fig.5] comprising an image of the same object of interest (pick-up) in SAR imaging (top) and in EO / IR (bottom).
[0090] Knowing the height of the shot, the size of the object is determined based on the measurement of the shadow cast.
[0091] In optics, depending on the respective positions of the sun and the wearer relative to the target, the height of the target can be determined by calculating the size of its projected shadow.
[0092] A concatenated image with additional information from the consolidated and conjugated shadows allows the 3D representation to be validated ([Fig.5]).
[0093] Changing sensor settings: determining camera resolution and aperture
[0094] Knowing the resolution of the image, it is possible to determine the size of the object. Knowing the altitude of the carrier making the acquisition, it is possible to determine the aperture of the camera necessary for effective pointing at the object.
[0095] A camera has the following elements:
[0096] The size of the output image in X*Y Pixels (X and Y are respectively the number of horizontal and vertical pixels in an image) - the pixel size: xx pm (micrometer) - the focal length of the optics, F - Instantaneous Field of View (IFOV) in rad.
[0097] The goal is to determine the size of an object, knowing the number of pixels, the distance to the object.
[0098] Let / 5 correspond to the size in radians of the object in the image [°0"] P_ jj “ ttiri ll
[0100] with
[0101] D: distance to target
[0102] T^'f ' fault along an axis (or Apparent Surface) (what we are looking for)
[0103] To obtain the pixel size of the object in the image, we use the IFOV of the camera, such as :
[0104] Re / =_A_(5) JPOV
[0105] Thus, knowing the distance to the object, we can determine the corresponding size:
[0106] = DW (Repxl*IFOV) (6)
[0107] Performance calculation (processing carried out during step 202)
[0108] It is possible to use statistical methods for performance measurement.
[0109] In this case, the optical part will detect the movements of the targets in the widest possible area.
[0110] The radar part will isolate the targets detected in the same area.
[0111] A processing of the speeds is carried out: the tangential speeds are less precise in radar, the radial speeds are more difficult to estimate in optics.
[0112] The quantity of targets detected in each of the sensors CPT1, CPT2 is compared taking into account the intrinsic parameters of the sensors. A sorting according to the speed vectors will be carried out and will make it possible to take into account the differences between the sensors.
[0113] For example, vectors at + / -45° will be weighted more heavily because the information is comparable.
[0114] A target detection density will be performed by the two sensors with a comparison by zones.
[0115] The acquisition conditions of the optical part will be taken into account for this comparison.
[0116] From these measurements, it is possible to determine the exact shape and / or size(s) of the object which will make it possible to improve the performance of the sensors by modifying the programming parameters of the sensors according to the exact shape and / or size(s).
Claims
Claims
1. Method for improving the performance of sensors (20) in a system (1) comprising a plurality of platforms (10) each carrying at least one sensor (20) and an electronic processing module (30), the platforms being connected by remote transmission links, said method comprising the following steps: - capturing a first image by a first sensor (20) on a first platform (30) of the system and programmed according to a first configuration of the first sensor; - determination, by a first processing block (30) on the first platform (10), of a first set of information, as a function of the first captured image, said set of information comprising at least one piece of information on the location of an object detected on said first image and one piece of information indicating the size of said object; - remote transmission of said first set of information from the first platform (10) to a second platform (10); - reception, by the second platform (10) comprising a second sensor (20) and a second processing block (30), of said first set of information and determination, by the second processing module (30), of a first configuration of the second sensor (20) as a function of the information of the set of information; - capturing a second image comprising said object by the second sensor (20) programmed according to said first configuration of the second sensor determined; - determination, by the second processing block (30), of a second set of information, as a function of the second captured image, said second set of information comprising at least one piece of information on the location of said object detected on said second image and one piece of information indicating the size of said object; - evaluation, as a function of the first and second sets of information, by one of the first and second processing modules (30) and selection, as a function of this evaluation, of one of the first and second sets of information; - triggering of programming, according to a second configuration as a function of the selected set of information, of that of the first and second sensors (20) not programmed according to said selected set of information and capture of a new image by said sensor programmed according to said second configuration.
2. Method for improving the performance of sensors (20) in a system (1) according to claim 1, according to which said set of information further comprises at least one piece of information relating to image capture among: sensor aperture, resolution, distance to the object, classification of the object.
3. A method of improving the performance of sensors (20) in a system (1) according to claim 1 or 2, wherein the acquisition spectra of the first and second sensors are separated, for example one of the first and second sensors is of the radar type and the other of the first and second sensors is of the EO / IR type.
4. A method of improving the performance of sensors (20) in a system (1) according to any one of the preceding claims, wherein the steps of the method are implemented cyclically between distinct pairs of platforms (10) of the system.
5. A method of improving the performance of sensors (20) in a system (1) according to any one of the preceding claims, further comprising triggering the programming, according to a third configuration depending on the selected set of information, of a sensor (20) of a platform (10) distinct from the first and second platforms and capturing a new image by said programmed sensor according to said third configuration.
6. System (1) comprising a plurality of platforms (10) each carrying at least one sensor (20) and an electronic processing module (30), the platforms being connected by remote transmission links, said system comprising: - a first sensor (20) on a first platform (10) of the system and programmed according to a first configuration of the first sensor is adapted to capture a first image by; - a first processing block (30) on the first platform, is adapted to determine a first set information, based on the first captured image, said set of information comprising at least one piece of information on the location of an object detected on said first image and one piece of information indicating the size of said object; the first platform (10) is adapted to transmit said first set of information to a second platform (10); said second platform (10) comprising a second sensor (20) and a second processing block (30), is adapted to receive said first set of information and the second processing module is adapted to determine a first configuration of the second sensor as a function of the information of the set of information; the second sensor (20) is adapted to, programmed according to said first configuration of the second sensor determined, capture a second image comprising said object; the second processing block (30) is adapted to determine a second set of information, as a function of the second captured image, said second set of information comprising at least one piece of information on the location of said object detected on said second image and one piece of information indicating the size of said object; one of the first and second processing modules (30) is adapted to carry out an evaluation, as a function of the first and second sets of information, and to select, as a function of this evaluation, one of the first and second sets of information; said system (1) being adapted to trigger the programming, according to a second configuration depending on the selected set of information, of that of the first and second sensors not programmed according to said selected set of information and to trigger the capture of a new image by said sensor programmed according to said second configuration.
7. System (1) according to claim 6, wherein said set of information further comprises at least one piece of information relating to image capture among: sensor aperture, resolution, distance to object, object classification.
8. System (1) according to claim 6 or 7, in which the acquisition spectra of the first and second sensors are separate, for example one of the first and second sensors is of the radar type and the other of the first and second sensors is of the EO / IR type.
9. System (1) according to any one of claims 6 to 8, adapted to cyclically implement said operations between distinct pairs of platforms of the system.
10. System (1) according to any one of the preceding claims 6 to 9, further adapted to trigger the programming, according to a third configuration depending on the selected set of information, of a sensor (20) of a platform (10) distinct from the first and second platforms and capture of a new image by said sensor (20) programmed according to said third configuration.