WO2025008599A1 - Procede et dispositif de suivi de satellites dans une sequence d'images - Google Patents
Procede et dispositif de suivi de satellites dans une sequence d'images Download PDFInfo
- Publication number
- WO2025008599A1 WO2025008599A1 PCT/FR2024/050918 FR2024050918W WO2025008599A1 WO 2025008599 A1 WO2025008599 A1 WO 2025008599A1 FR 2024050918 W FR2024050918 W FR 2024050918W WO 2025008599 A1 WO2025008599 A1 WO 2025008599A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- measurements
- trajectories
- satellite
- sequence
- trj
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/62—Extraction of image or video features relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Definitions
- the present invention relates to the fields of image analysis and object tracking.
- the present invention relates to a method and device for tracking satellites in an image sequence, as well as an associated monitoring system, computer program and information medium.
- the present invention finds a particularly advantageous, although in no way limiting, application for the implementation of space monitoring systems.
- the invention falls within the particular context of outer space surveillance systems. Such systems aim to locate satellites in orbit and provide their respective trajectories. To this end, these systems use optical sensors to capture images of satellites orbiting the Earth. However, these images include satellites, stars, and noise (e.g. hot pixels).
- the present invention aims to remedy all or part of the drawbacks of the prior art, in particular those set out above.
- a method for tracking satellites in a sequence of images comprising: a step of obtaining a set of measurements associated with spatial objects observed in the images of the sequence, each measurement comprising a position of an object in an image and a temporal index of the image; a step of detecting satellite trajectories in the sequence of images, by a detector comprising at least one neural network and using all or part of the set of measurements obtained; and a step of classifying the detected satellite trajectories, by a classifier comprising at least one other neural network.
- the present invention makes it possible to extract the respective trajectories of the satellites observed in the image sequence.
- the present invention makes it possible to associate the measurements belonging to the same satellite to describe its trajectory in the image sequence. It is important to note that the present invention makes it possible not only to track satellites in geostationary orbit, but also satellites in medium or low Earth orbit.
- the present invention allows for classifying the detected trajectories.
- the class assigned to each satellite trajectory may indicate the type of orbit of the satellite.
- the list of assignable classes may include: fixed geostationary trajectory; drifting geostationary trajectory; scrolling type trajectory (i.e. satellite in medium or low Earth orbit moving at high speed); and trailing type trajectory (i.e. satellite in low Earth orbit moving at very high speed).
- the present invention makes it possible to improve the reliability of tracking observed satellites within an image sequence. It also has reduced implementation complexity (i.e. reduced execution times, less configuration).
- the present invention does not require a priori knowledge of the orbits to track satellites in an image sequence.
- the use of one or more neural networks to implement the detector makes it possible to obtain reliable detection of the trajectories, in terms of correctly detected trajectories.
- the neural network(s) are trained automatically from training data (i.e. reference image sequences) to optimize the reliability of detection of the satellite trajectories. This makes it possible in particular to obtain significantly improved performance compared to analytical solutions.
- the use of one or more neural networks Multiple neural networks can significantly reduce the computing resources required, particularly in terms of execution time.
- the classifier comprising one or more neural networks, which makes it possible to improve the reliability of classification of the trajectories (i.e. correctly identified trajectories). This also makes it possible to reduce the necessary computing resources (i.e. execution time).
- the joint use of one or more neural networks to detect the satellite trajectories, and of one or more other neural networks to classify these trajectories makes it possible to obtain reliable tracking of the satellites observed in the image sequence.
- the present invention makes it possible to track satellites in an image sequence reliably and with minimal implementation complexity.
- the detector comprises at least one convolutional neural network, and determines frames each encompassing (surrounding) a satellite trajectory in the images of the sequence.
- the classifier comprises at least one classification (i.e. segmentation) neural network, in particular at least one PointNet type neural network. And, for each frame encompassing a satellite trajectory, the classifier determines (i.e. associates) the measurements located in the frame belonging to this trajectory, and assigns a class to this trajectory.
- convolutional neural network we refer here to a neural network comprising one or more convolution layers.
- PointNet neural network we refer to a neural network dedicated to processing point clouds, and in particular to classification and segmentation applications. PointNet networks and their implementation are for example described in the following document: Qi et al., “PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation”, 2017.
- This implementation method proposes to use one or more convolutional neural networks to detect satellite trajectories, and to use one or more PointNet type neural networks to classify (segment) these trajectories.
- the convolutional neural network(s) enable the detector to exploit the spatiotemporal characteristics of the image sequence to detect the satellite trajectories. More specifically, it exploits the spatiotemporal characteristics of the image sequence to determine bounding boxes (i.e., "bounding boxes") for the satellite trajectories.
- bounding boxes i.e., "bounding boxes"
- PointNet neural networks to implement the classifier is particularly advantageous.
- the PointNet network is optimized to classify and segment point clouds.
- the PointNet network processes separately each of the frames encompassing the satellite trajectories. For each frame encompassing a trajectory, it takes as input all the measurements located inside the bounding box in the image sequence.
- these PointNet type networks make it possible to reliably separate the measurements belonging to the trajectory from those not belonging to the trajectory, and to assign a class to this trajectory.
- this mode of implementation makes it possible to reliably extract the respective trajectories of the different satellites observed in the sequence of images and to classify them.
- the detector comprises at least one Transformer-type neural network, and associates the measurements belonging to the same satellite trajectory in the images of the sequence.
- the classifier comprises at least one classification network and assigns a class to each of the satellite trajectories.
- Transformer type networks are optimized to process sequential data and make it possible to exploit the temporal characteristics of the input data.
- the use of a Transformer type network allows the detector to exploit the temporal characteristics of the set of measurements to extract the respective trajectories of the satellites observed in the image sequence.
- This mode of implementation is particularly advantageous in that it makes it possible to improve the reliability of the tracking of the satellites observed in a sequence of images.
- the use of a Transformer type network makes it possible to reduce the computing resources (i.e. the execution time) necessary for the implementation of the proposed solution.
- the detector determines latent vectors representative of said measurements using said at least one Transformer-type neural network; and the detector associates measurements with the same satellite trajectory as a function of distances between the latent vectors.
- This implementation mode proposes to use the Transformer network type network to represent in a latent space the measurements associated with the spatial objects observed in the image sequence. For each of the measurements provided as input, the Transformer neural network determines a latent vector.
- the Transformer-type neural network makes it possible to learn, on the basis of training data, a latent space in which the measurements belonging to the same trajectory are close. This network thus makes it possible to represent the measurements in a latent space so as to minimize the distance between the latent vectors of measurements belonging to the same trajectory. trajectory, and to maximize the distances between the latent vectors of measurements belonging to different trajectories.
- This mode of implementation therefore makes it possible, by using the distances between the latent vectors, to reliably associate (i.e. to group) the measurements belonging to the same trajectory.
- the detector comprises a plurality of neural networks respectively dedicated to the detection of satellite trajectories of different speeds; and/or the classifier comprises a plurality of neural networks respectively dedicated to the classification of satellite trajectories of different speeds.
- This implementation mode makes it possible to adapt the detection and/or classification of trajectories according to different types of trajectories (e.g. fixed or drifting geostationary orbits, medium and low Earth orbits). This advantageously makes it possible to improve the reliability of the proposed solution.
- trajectories e.g. fixed or drifting geostationary orbits, medium and low Earth orbits. This advantageously makes it possible to improve the reliability of the proposed solution.
- the method comprises a step of filtering the set of measurements to eliminate the measurements associated with (i.e. originating from) the stars, and in which the detector only uses the filtered set of measurements to carry out the step of detecting satellite trajectories in the sequence of images.
- This implementation mode makes it possible to eliminate the stars present in the image sequence.
- the stars are in large number compared to the satellites in an image sequence, which can interfere with the detection of satellite trajectories.
- this implementation mode contributes to improving the reliability of the proposed solution for tracking satellites in an image sequence.
- the proposed solution for eliminating the measurements corresponding to stars does not require a priori knowledge of their positions.
- the step of filtering the set of measurements to eliminate the measurements associated with stars comprises: a determination of star displacement information from measurements whose brightness is greater than a threshold in the images of the sequence; and a step of eliminating (i.e. deleting) the measurements considered to be associated with stars from the determined displacement information.
- This mode of implementation is advantageous in that it allows stars to be reliably eliminated in the images of the sequence.
- the inventors have observed that the majority of the brightest measurements in an image sequence are stars. Also, it is proposed here to exploit the brightest measurements to determine the star displacement information (eg speed of movement of the stars per image, or rotation angle of the stars per image). Then, these displacement information is used to determine which measurements correspond to stars, and eliminate these.
- star displacement information eg speed of movement of the stars per image, or rotation angle of the stars per image.
- the method comprises: one or more steps of merging (i.e. combining) the satellite trajectories to group together the measurements associated with several detected satellite trajectories and belonging to the same satellite; and/or a step of filtering the satellite trajectories so that a said satellite trajectory comprises only one measurement per image of the sequence.
- This implementation mode proposes to carry out different post-processing steps on the satellite trajectories detected by the proposed solution. It thus contributes to improving the reliability of the proposed satellite tracking solution.
- a device for tracking satellites in a sequence of images comprising: an obtaining module configured to obtain a set of measurements associated with spatial objects observed in the images of the sequence, each measurement comprising a position of an object in an image and a temporal index of the image; a detector configured to detect satellite trajectories in the sequence of images, the detector comprising at least one neural network and using all or part of the set of measurements obtained; and a classifier configured to classify the detected satellite trajectories, the classifier comprising at least one other neural network.
- the satellite tracking device may be configured to carry out any of the modes of implementation of a method according to the invention.
- the device may comprise a module configured to implement this step.
- module can correspond to both a software component and a hardware component or a set of hardware and software components.
- a software component corresponds to one or more computer programs or subprograms, or more generally to any element of a program capable of implementing a function or a set of functions as described for the modules concerned.
- a hardware component corresponds to any element of a hardware assembly capable of implementing a function or a set of functions for the module concerned.
- a surveillance system comprising: a satellite tracking device according to the invention; and a device for acquiring a sequence of images.
- a computer program comprising instructions for implementing the steps of a method according to the invention, when the computer program is executed by at least one processor or one computer.
- the computer program may be formed of one or more subparts stored in the same memory or in separate memories.
- the program may use any programming language, and be in the form of source code, object code, or intermediate code between source code and object code, such as in a partially compiled form, or in any other desirable form.
- a computer-readable information medium comprising a computer program according to the invention.
- the information carrier may be any entity or device capable of storing the program.
- the carrier may comprise a storage means, such as a non-volatile memory or ROM, for example a CD-ROM or a microelectronic circuit ROM, or a magnetic recording means, for example a floppy disk or a hard disk.
- the storage medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by a telecommunications network or by a computer network or by other means.
- the program according to the invention may in particular be downloaded onto a computer network.
- the information carrier may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the method in question.
- the proposed satellite tracking device, monitoring system, computer program and information medium have the advantages described above in connection with the proposed satellite tracking method.
- Figure 1 represents an exemplary architecture of a monitoring system according to an embodiment of the invention
- Figure 2 represents an example of architecture of a satellite tracking device in a sequence of images according to an embodiment of the invention
- Figure 3 represents steps of a method for tracking satellites in a sequence of images according to one embodiment of the invention
- Figure 4 shows steps of a method according to one embodiment of the invention for eliminating stars in a sequence of images
- Figure 5 represents an example of architecture of a satellite tracking device in a sequence of images according to an embodiment of the invention
- Figure 6 shows examples of images processed by a satellite tracking device according to one embodiment of the invention.
- Figure 7 shows an example of architecture of a satellite tracking device in an image sequence according to an embodiment of the invention.
- Figure 8 represents steps of a method according to an embodiment of the invention for performing post-processing operations on satellite trajectories.
- the present invention applies, in particular, to space surveillance systems.
- the following description of the invention will refer to this particular context of application, which is given only as an illustrative example and should not limit the invention.
- Figure 1 represents an exemplary architecture of a monitoring system according to an embodiment of the invention. This figure is described below to introduce the present invention and to exemplify an application context.
- the monitoring system SYS comprises at least one acquisition device SENS and a satellite tracking device APP.
- the system SYS may also comprise a restitution and/or storage device OUT.
- SYS system comprising a single SENS acquisition device.
- the SYS system comprises a plurality of SENS acquisition devices, located at different locations on Earth and/or of different types.
- the SENS device is configured to acquire (ie capture) a sequence of images IMG_SEQ.
- This SENS device comprises one or more optical sensors, such as a camera, or an infrared sensor. As illustrated in FIG. 1, the SENS device is configured to provide the sequence of images IMG_SEQ to the APP device.
- the image sequence IMG_SEQ comprises a plurality of consecutive images covering a portion of the sky. Each image of the sequence IMG_SEQ is associated with an acquisition time. Typically, within the images of the sequence IMG_SEQ, there are both satellites orbiting the Earth, stars (in large numbers compared to the satellites), and measurements from noise (eg hot pixels).
- the satellites observed in the image sequence IMG_SEQ may include fixed geostationary satellites, drifting geostationary satellites, and moving satellites.
- the observed satellites may have geostationary orbits GEO (acronym for “Geosynchronous Earth Orbit” in English), medium Earth orbits MEO (acronym for “Medium Earth Orbit” in English), or low Earth orbits LEO (acronym for “Low Earth Orbit” in English).
- the APP device is configured to track the satellites observed in the image sequence IMG_SEQ. In other words, it is configured to determine the respective TRJ trajectories of the different satellites present in the image sequence IMG_SEQ.
- a "satellite trajectory” is defined by a set of measurements associated with the satellite, each of these measurements comprising a position of the satellite (eg x and y coordinates of the satellite in the image), and a time index (eg index of the image in the sequence, or instant of acquisition of the image).
- the APP device is configured to provide the TRJ satellite trajectories to the OUT device.
- the OUT device may be a rendering device comprising a screen for displaying the TRJ satellite trajectories.
- the OUT device could be a storage device used to record the TRJ satellite trajectories.
- the APP device comprises, according to the embodiment illustrated by FIG. 1: at least one processing unit or processor PROC; and at least one memory MEM.
- the APP device has, according to one embodiment, the hardware architecture of a computer.
- the memory MEM constitutes an information medium in accordance with the invention, that is to say readable by the processor PROC and on which a computer program PROG in accordance with the invention is recorded.
- the program PROG comprises instructions for carrying out steps of a satellite tracking method in accordance with the invention, when the program PROG is executed by the processor PROC.
- the program PROG defines in particular the functional modules of the APP device represented in FIG. 2, which rely on or control the hardware elements of the latter.
- the APP device has a communication module COM configured to communicate with the acquisition device SENS and/or the OUT device. No limitation is attached to the nature of the communication interfaces between these devices, which may be wired or wireless, and may implement any protocol known to those skilled in the art.
- Figure 2 and Figure 3 respectively represent an example of architecture of a satellite tracking device and steps of a satellite tracking method according to an embodiment of the invention. These figures present the operation of the APP device introduced with reference to Figure 1.
- the proposed method comprises all or part of the steps S100 to S600 described below, these being implemented by the device APP.
- the proposed method may comprise a plurality of iterations of each of its steps.
- the steps of the method may be reiterated at a given frequency, for example every night, or every week.
- step S100 the obtaining module OBT of the APP device obtains the image sequence IMG_SEQ. From this, it obtains a set of measurements Si N associated with the spatial objects observed in the images of the sequence IMG_SEQ.
- the APP device receives the image sequence IMG_SEQ from the acquisition device SENS, via its communication module COM.
- the monitoring device APP could also read the image sequence IMG_SEQ on a memory shared with the acquisition device SENS.
- the APP device processes the image sequence IMG_SEQ to obtain the set of measurements Si N associated with the different spatial objects observed in the sequence IMG_SEQ.
- the APP device can use any known technique for detecting points in images. For example, it can use a high-pass filter, or a neural network.
- the filtered set of measurements S F LT does not include the measurements considered as corresponding to stars.
- step S200 we describe in more detail the implementation of step S200 with reference to FIG. 4.
- step S300 the DET detector of the APP device detects, from the filtered measurement set SFLT, the satellite trajectories TRJ in the image sequence IMG_SEQ.
- the DET detector comprises (i.e. implements) at least one artificial neural network.
- step S400 the CLA classifier of the APP device classifies the detected satellite trajectories TRJ.
- the CLA classifier comprises (i.e. implements) at least one artificial neural network.
- the APP device obtains satellite trajectories TRJ.
- the APP device assigns a class C m to this trajectory T m .
- the classes assigned to the trajectories by the APP device indicate the types of trajectories (i.e. orbits) of the satellites.
- the list of assignable classes may include: fixed geostationary trajectory; drifting geostationary trajectory; scrolling type trajectory (i.e. satellite in medium or low Earth orbit moving at high speed); and trailing type trajectory (i.e. satellite in low Earth orbit moving at very high speed); and noise.
- steps S300 and S400 are described in more detail with reference to FIGS. 5 and 7. More specifically, FIG. 5 illustrates a first embodiment of steps S300 and S400, while FIG. 7 illustrates a second embodiment of these steps.
- step S500 the processing module PP of the device APP performs (analytical) operations called post-processing of the detected satellite trajectories TRJ.
- step S500 makes it possible in particular to improve the results of the proposed solution for tracking satellites in a sequence of images IMG_SEQ.
- step S500 is detailed below with reference to FIG. 8.
- step S600 the device APP provides the satellite trajectories TRJ to the device OUT, via its communication module COM.
- each of the satellite trajectories TRJ provided by the APP device comprises: a set of measurements T m describing the positions of a satellite in the images of the sequence IMG_SEQ; and a class C m indicating the type of trajectory of the satellite.
- the OUT device can for example display the trajectories of satellites TRJ on a screen.
- this OUT device can be used to store the trajectories of satellites TRJ.
- These TRJ trajectories can be further analyzed by a separate device to catalog the orbits of the observed satellites.
- Figure 4 shows steps of a method according to an embodiment of the invention for eliminating stars in a sequence of images. This figure details the implementation of the filtering step S200 for eliminating the measurements associated with stars.
- step S200 comprises all or part of steps S210 to S240 described below and implemented by the FLT filter of the APP device.
- step S210 the APP device selects, from the set of measurements Si N , the measurements whose brightness is greater than a threshold in the images of the sequence IMG_SEQ.
- the APP device selects the brightest points in the images of the sequence IMG_SEQ.
- step S220 the APP device determines the star displacement information from selected measurements.
- the APP device determines whether the trajectories of the stars are substantially rectilinear in the image sequence IMG_SEQ, or whether they are substantially circular. To do this, it can determine whether the normals of the trajectories of the stars intersect near an observation center (i.e. whether the intersection of these normals is close to the observation center).
- the APP device determines the average speed of the stars per image. In other words, it determines the displacement vector of the stars from one image to the next (e.g. -5.1 pixels/image along the axis, and -1.2 pixels/image along the y axis). To determine this displacement vector, the APP device calculates the average of the distances between each of the measurements and its closest neighbor in the following image.
- the APP device determines the center of rotation of the stars (i.e. the intersection of the normals), as well as the rotation angle of the stars from one image to another (eg 3.2 degrees/image). [0101] In step S230, the APP device eliminates, using the star displacement information, the measurements considered to come from stars.
- the APP device performs step S230 in the following manner. For each observed spatial object, the APP device predicts its positions in the other images of the sequence by assuming that this object is a star. If these predictions prove to be correct, then the APP device considers that the object is a star and eliminates the associated measurement.
- the APP device predicts the positions of the object in the other images of the sequence using the star displacement information. If the number of predicted positions considered correct is greater than a threshold, the APP device eliminates the measurement considered. A predicted position is considered correct when there is a measurement present in the image and whose distance to the predicted position is less than a threshold.
- step S240 the APP device eliminates, among the remaining measurements, the measurements isolated in the image sequence IMG_SEQ.
- the APP device identifies measurements that do not have close neighbors, and eliminates these measurements.
- a measurement is considered to have no close neighbors when there is no measurement located near it, i.e. no measurement included below a distance threshold.
- the APP device considers in particular the image in which the measurement is located, but also the previous and following images (e.g. the two previous images, and the two following images).
- the APP device can also eliminate measurements located on the edges (i.e. included below a given distance from the edges of the images) in the first images or the last images of the IMG_SEQ sequence.
- the APP device obtains a filtered set of measurements S F LT, in which the measurements considered to be associated with stars have been eliminated.
- This set S F LT is then used by the APP device to detect the satellite trajectories TRJ in the image sequence IMG_SEQ.
- the APP device uses this set S F LT to detect the satellite trajectories TRJ in the image sequence IMG_SEQ.
- Figure 5 represents an example of architecture of a satellite tracking device in an image sequence according to an embodiment of the invention. This figure illustrates a first embodiment of the DET detector and the CLA classifier, previously introduced with reference to Figures 2 and 3.
- the detector DET detects satellite trajectories TRJ in the image sequence IMG_SEQ from the filtered measurement set S FL T. Then, during step S400, the classifier CLA classifies (i.e. segments) the trajectories TRJ.
- This first embodiment proposes to use convolutional neural networks to implement the DET detector, and to use PointNet type neural networks to implement the CLA classifier.
- the DET detector comprises, according to this first embodiment, several convolutional neural networks CNN LS , CNN H s , and CNN V HS - Each of these networks takes as input an image in which all the filtered measurements S F LT are superimposed, and provides as output frames BB encompassing detected satellite trajectories (i.e. "bounding boxes" in English).
- the CNN LS , CNN HS , and CNN VH neural networks are respectively dedicated (i.e. specialized, trained) to detecting satellite trajectories of different speeds.
- the CNN LS network is trained to detect satellite trajectories in substantially fixed or drifting geostationary orbit (i.e. low-speed trajectories).
- the CNN H network is trained to detect satellite trajectories in medium or low orbit and moving at high speed (i.e. scrolling type trajectories).
- the CNNVHS network is trained to detect satellite trajectories in low orbit and moving at very high speed (i.e. trailing type trajectories).
- the CNN network L s can use a Faster R-CNN type neural network with an Inception V2 type feature extraction network (i.e. "backbone network”).
- Faster R-CNN type networks and their implementation are for example described in the following document: Ren et al., "Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks", 2016.
- Inception V2 type networks and their implementation are for example described in the following document: Szegedy et al., "Rethinking the Inception Architecture for Computer Vision", 2015.
- the CNN H s and CNN V HS networks can respectively use a Faster R-CNN type neural network with a Resnet 101 type feature extraction network.
- the ResNet 101 type networks and their implementation are for example described in the following document: He et al., “Deep Residual Learning for Image Recognition”, 2015.
- Training of the DET detector neural networks is performed using simulated training data, i.e. images comprising computer-generated satellite trajectories to which noise is added.
- the neural networks can also be pre-trained using public datasets, such as the “COCO” dataset.
- the CNN LS network To dedicate the CNN LS network to the detection of satellite trajectories in substantially fixed or drifting geostationary orbit (i.e. low-speed trajectories), it is trained using only these types of trajectories. Similarly, the CNN H s network is trained using scrolling type trajectories, and the CNNVHS network is trained using trailing type trajectories.
- the CNNLS neural network processes in parts the image in which the set of filtered measurements S FL T is superimposed.
- the CNNHS and CNNVHS networks take the entirety of this image as input.
- the CLA classifier comprises, according to this first embodiment, several PN L s and PN HS neural networks of PointNet type. These PN L s and PN HS networks are respectively dedicated to the classification (i.e. the segmentation) of trajectories of TRJ satellites of different speeds.
- the PN LS network is trained to classify trajectories of satellites in substantially fixed or drifting geostationary orbit (i.e. low-speed trajectories). As illustrated in FIG. 7, the PN LS network takes as input the BB frames provided by the CNN LS network, dedicated to the detection of low-speed trajectories.
- the PN LS network processes each of the frames BB separately. For each frame B m encompassing a trajectory T m , it takes as input all the measurements located inside the frame B m in the image sequence IMG_SEQ. Note also that the measurements located in the frame B m are recentered before providing them to the PN LS network so that the coordinates x m and y m of the center of the frame are equal to 0.
- the PN LS network provides as output a label to each of the measurements located in the frame B m , this label indicating whether or not the measurement belongs to the trajectory T m .
- the PN L s network makes it possible to separate the measurements belonging to the trajectories TRJ from the noise.
- it provides as output a class C m attributed to the trajectory T m .
- the list of classes assignable by the PN LS network is: fixed geostationary trajectory; drifting geostationary trajectory; double drifting geostationary trajectories (two drifting trajectories are present in the same frame); and noise.
- the PN HS network is trained to classify trajectories of satellites in medium or low orbit (i.e. high-speed trajectories).
- the PN H S network takes as input the BB frames provided by the CNN HS and CNN VHS networks, dedicated to the detection of high-speed trajectories.
- the PN H s network processes each of the BB frames separately. For the PN H s network, only the measurements around the diagonals of the BB bounding frames (i.e. included below a threshold of distance from the diagonals) are provided.
- the PN HS network provides as output a label to each of the measurements located in the frame B m , this label indicating whether or not the measurement belongs to the trajectory T m . It also provides as output a class C m attributed to the trajectory T m .
- the list of classes assignable by the PN HS network is: high-speed trajectory (scrolling-type trajectory or trailing-type trajectory); and noise. Then, the APP device can distinguish the scrolling-type trajectories and the trailing-type trajectories as follows.
- the scrolling-type trajectories comprise on average less than 1.5 measurements per frame, while the trailing-type trajectories comprise on average more than 1.5 measurements per frame.
- PointNet type networks and their implementation are for example described in the previously cited Qi et al. document.
- PointNet type PN LS and PN H s networks are trained using simulated data.
- the PN L s network To dedicate the PN L s network to the classification (segmentation) of satellite trajectories in substantially fixed or drifting geostationary orbit (i.e. low-speed trajectories), it is trained using only these types of trajectories. Similarly, the PN H s network is trained using scrolling type trajectories and trailing type trajectories (i.e. high-speed trajectories).
- Figure 6 shows examples of images processed by a satellite tracking device according to one embodiment of the invention.
- the first image IMG_SUP represents a superposition of the images of the sequence IMG_SEQ obtained by the obtaining module OBT during step S100.
- the second image IMG_SUP_FLT represents within the same image the set of filtered measurements S F LT obtained by the FLT filter following step S200.
- the third image IMG_TRJ_DET represents within the same image the BB frames encompassing the satellite trajectories detected by the DET detector following step S300.
- the fourth image IMG_TRJ_CLA represents within the same image the trajectories of TRJ satellites obtained by the CLA classifier following step S400.
- the fifth image IMG_TRJ_PP represents within the same image the trajectories of satellites TRJ obtained by the processing module PP following step S500.
- Figure 7 represents an example of architecture of a satellite tracking device in an image sequence according to an embodiment of the invention. This figure illustrates a second embodiment of the DET detector and the CLA classifier, previously introduced with reference to Figures 2 and 3.
- the DET detector comprises, according to this second embodiment, at least one Transformer-type TRF neural network, i.e. a network comprising an attention mechanism.
- Transformer-type TRF neural networks and their implementation are for example described in the document A. Vaswani et al. previously cited.
- the TRF network of the DET detector takes as input the measurements of the set S F LT, and provides a latent vector LV for each of these measurements. Then, a CLST association module of the DET detector associates (i.e. groups together) the measurements belonging to the same satellite trajectory T m according to the Euclidean distances between the latent vectors LV.
- the CLST association module uses a technique called agglomerative hierarchical clustering with a defined distance threshold.
- agglomerative hierarchical clustering technique uses a technique called agglomerative hierarchical clustering with a defined distance threshold.
- the details of implementing the agglomerative hierarchical clustering technique are known to those skilled in the art.
- the CLST association module uses other grouping algorithms to associate the measurements belonging to the same satellite trajectory from the distances between the latent vectors.
- the CLA classifier comprises, according to this second embodiment, at least one classification neural network CN.
- This CN neural network processes the different detected satellite trajectories TRJ separately.
- the classification network CN takes as input the measurements associated with this trajectory T m , and provides as output a class C m attributed to this trajectory T m .
- the CN network may be a multilayer perceptron, comprising several dense layers with ReLU type activation functions, and an output layer with a Softmax type activation function.
- TRF and CN neural networks can be trained from simulated (i.e. computer-generated) training data.
- the DET detector comprises a plurality of Transformer-type neural networks respectively dedicated to the detection of TRJ satellite trajectories of different speeds.
- the CLA classifier could comprise a plurality of classification neural networks respectively dedicated to the classification of TRJ satellite trajectories of different speeds.
- the device APP performs during the step S500 various post-processing operations of these trajectories TRJ.
- the step S500 makes it possible to improve the reliability of the satellite tracking carried out by the device APP.
- Figure 8 represents steps of a method according to an embodiment of the invention for performing post-processing operations of satellite trajectories. This figure details the implementation of step S500 introduced with reference to Figures 2 and 3.
- step S500 comprises all or part of steps S510 to S550 described below and implemented by the processing module PP of the device APP.
- step S510 the APP device merges (i.e. combines) satellite trajectories TRJ to group the measurements associated with several detected satellite trajectories TRJ and belonging to the same satellite. Thus, if measurements belonging to the same satellite have been associated with several trajectories by the APP device, then these measurements are grouped within the same trajectory.
- the APP device compares TRJ satellite trajectories two by two, and merges them or not depending on the result of the comparison.
- the APP device compares the number of measurements common to two trajectories T m and T p . If this number is greater than a given percentage (eg 20%) of the number of measurements of the smallest of the two trajectories, then the measurements of the two trajectories T m and T p are grouped within a single trajectory. In other words, the APP device then achieves the union of the two trajectories T m and T p .
- a given percentage eg 20%
- the APP device does not compare all types of trajectories together.
- fixed geostationary type trajectories are compared only with each other.
- drifting geostationary trajectories, scrolling type trajectories, and trailed type trajectories are compared together.
- step S520 the APP device performs a completion of the TRJ satellite trajectories of scrolling types, i.e. high-speed trajectories comprising on average less than 1.5 measurements per image.
- This completion uses three-dimensional linear regressions.
- the completion carried out by the APP device makes it possible to eliminate measurements associated with the trajectory T m , but not belonging to this satellite. In addition, this completion makes it possible to add measurements belonging to the satellite and which would not have been associated with the trajectory T m .
- the APP device predicts the position of the object using three-dimensional linear regression determined from the measurements of the trajectory T m . If there is another measurement closer to the predicted position than the measurement considered, then the measurement considered is eliminated from the trajectory T m , and is replaced by the other closer measurement.
- the APP device For each image i k of the sequence IMG_SEQ, the APP device proceeds as follows. If the trajectory T m includes a measurement on this image i k , and if the distance between this measurement and a predicted position is greater than a threshold, then the measurement is eliminated from the trajectory T m . If the trajectory T m does not include a measurement on this image i k , and if the distance between the predicted position and the measurement closest to the predicted position is less than a threshold, then this measurement is associated (i.e. added) to the trajectory T m .
- step S530 the APP device filters the trail-type TRJ trajectories, i.e. the high-speed trajectories comprising on average more than 1.5 measurements per image.
- the APP device determines the slope and the ordinate at the origin of the linear regression line formed by the measurements of the trajectory T m . It then eliminates, among the measurements of the trajectory T m , those which deviate beyond a given distance from the linear regression line.
- the APP device thus eliminates measurements that are not aligned along the two-dimensional trajectory. In this sense, it performs a polishing of the trail-type trajectories.
- step S540 the APP device merges the trail-type trajectories. This makes it possible to prevent the same trail-type trajectory from being detected in several disjoint pieces by the APP device. To this end, the APP device compares the trail-type trajectories with each other, and merges them or not depending on the result of the comparison. [0166] In fact, the device compares two by two the linear regression lines of the trajectories of the drag type.
- the APP device groups the measurements of the two trajectories T m and T p within a single trajectory.
- step S550 the APP device filters the satellite trajectories TRJ so that each of the trajectories comprises only one measurement per image of the sequence IMG_SEQ.
- a trajectory T m comprises several measurements on an image of the sequence IMG_SEQ
- the APP device combines these measurements to obtain a single measurement.
- the position of the single measurement obtained by the APP device may correspond to the barycenter of the positions of the measurements of the trajectory T m included in this image.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
Description
Claims
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| AU2024291267A AU2024291267A1 (en) | 2023-07-06 | 2024-07-05 | Method and device for tracking satellites in an image sequence |
| EP24745480.4A EP4740187A1 (fr) | 2023-07-06 | 2024-07-05 | Procede et dispositif de suivi de satellites dans une sequence d'images |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| FRFR2307195 | 2023-07-06 | ||
| FR2307195A FR3150891B1 (fr) | 2023-07-06 | 2023-07-06 | Procédé et dispositif de suivi de satellites dans une séquence d'images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2025008599A1 true WO2025008599A1 (fr) | 2025-01-09 |
Family
ID=88689997
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/FR2024/050918 Ceased WO2025008599A1 (fr) | 2023-07-06 | 2024-07-05 | Procede et dispositif de suivi de satellites dans une sequence d'images |
Country Status (4)
| Country | Link |
|---|---|
| EP (1) | EP4740187A1 (fr) |
| AU (1) | AU2024291267A1 (fr) |
| FR (1) | FR3150891B1 (fr) |
| WO (1) | WO2025008599A1 (fr) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3120950A1 (fr) * | 2021-03-17 | 2022-09-23 | Share My Space | Système de détection de la trajectoire d’objets mobiles |
-
2023
- 2023-07-06 FR FR2307195A patent/FR3150891B1/fr active Active
-
2024
- 2024-07-05 WO PCT/FR2024/050918 patent/WO2025008599A1/fr not_active Ceased
- 2024-07-05 EP EP24745480.4A patent/EP4740187A1/fr active Pending
- 2024-07-05 AU AU2024291267A patent/AU2024291267A1/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR3120950A1 (fr) * | 2021-03-17 | 2022-09-23 | Share My Space | Système de détection de la trajectoire d’objets mobiles |
Non-Patent Citations (8)
| Title |
|---|
| A. VASWANI ET AL., ATTENTION IS ALI YOU NEED, 2017 |
| HE ET AL., DEEP RESIDUAL LEARNING FOR IMAGE RECOGNITION, 2015 |
| JARRED JORDAN ET AL: "Quasi Real-Time Autonomous Satellite Detection and Orbit Estimation", ARXIV.ORG, CORNELL UNIVERSITY LIBRARY, 201 OLIN LIBRARY CORNELL UNIVERSITY ITHACA, NY 14853, 13 April 2023 (2023-04-13), XP091483482 * |
| LIU TONG ET AL: "Photometric space object classification via deep learning algorithms", ACTA ASTRONAUTICA, PERGAMON PRESS, ELMSFORD, GB, vol. 185, 8 May 2021 (2021-05-08), pages 161 - 169, XP086602851, ISSN: 0094-5765, [retrieved on 20210508], DOI: 10.1016/J.ACTAASTRO.2021.05.008 * |
| PENG HAO ET AL: "Comparative evaluation of three machine learning algorithms on improving orbit prediction accuracy", ASTRODYNAMICS, vol. 3, no. 4, 1 December 2019 (2019-12-01), pages 325 - 343, XP093125074, ISSN: 2522-008X, DOI: 10.1007/s42064-018-0055-4 * |
| QI ET AL., POINTNET : DEEP LEARNING ON POINT SETS FOR 3D CLASSIFICATION AND SEGMENTATION, 2017 |
| REN ET AL., FASTER R-CNN : TOWARDS REAL-TIME OBJECT DETECTION WITH REGION PROPOSAL NETWORKS, 2016 |
| SZEGEDY ET AL., RETHINKING THE INCEPTION ARCHITECTURE FOR COMPUTER VISION, 2015 |
Also Published As
| Publication number | Publication date |
|---|---|
| FR3150891B1 (fr) | 2025-08-01 |
| AU2024291267A1 (en) | 2026-02-05 |
| FR3150891A1 (fr) | 2025-01-10 |
| EP4740187A1 (fr) | 2026-05-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2400460B1 (fr) | Procédé d'évaluation de la vitesse horizontale d'un drone, notamment d'un drone apte au vol stationnaire autopiloté | |
| FR3073312A1 (fr) | Procede d'estimation de pose d'une camera dans le referentiel d'une scene tridimensionnelle, dispositif, systeme de realite augmentee et programme d'ordinateur associe | |
| EP1694058A1 (fr) | Procédé et dispositif de capture d'images comprenant une mesure de mouvements locaux | |
| FR2814312A1 (fr) | Procede de segmentation d'une surface image video par objets elementaires | |
| EP2795580A1 (fr) | Procédé d'estimation de flot optique à partir d'un capteur asynchrone de lumière | |
| EP2316082A1 (fr) | Procede d'identification d'un objet dans une archive video | |
| EP2930659B1 (fr) | Procédé de détection de points d'intérêt dans une image numérique | |
| WO2018185104A1 (fr) | Procede d'estimation de pose, dispositif, systeme et programme d'ordinateur associes | |
| EP3709049B1 (fr) | Systeme de traitement radar et procede de debruitage associe | |
| EP0961227B1 (fr) | Procédé de détection de la profondeur relative entre deux objects dans une scène à partir d'un couple d'images issues de prises de vues différentes | |
| EP3070643B1 (fr) | Procédé et système de reconnaissance d'objets par analyse de signaux d'image numérique d'une scène | |
| FR2916562A1 (fr) | Procede de detection d'un objet en mouvement dans un flux d'images | |
| EP4740187A1 (fr) | Procede et dispositif de suivi de satellites dans une sequence d'images | |
| EP4439478A1 (fr) | Procédé de suivi d'objets multiples | |
| EP3701492A1 (fr) | Procede de restauration d'images | |
| EP2943935B1 (fr) | Estimation de mouvement d'une image | |
| EP4191530A1 (fr) | Procédé de localisation et cartographie simultanées intégrant un masquage temporel auto-supervisé et modèle d'apprentissage automatique pour générer un tel masquage | |
| FR3127837A1 (fr) | Procédé d’analyse de la surface d’un toit d’un bâtiment | |
| EP4066204A1 (fr) | Procédé et dispositif de traitement d'images | |
| EP3274914A1 (fr) | Procédé de dénombrement d'objets dans une zone spatiale prédéterminée | |
| FR3146534A1 (fr) | Procede d’estimation d’une carte de profondeur d’une image monoculaire et dispositifs associes | |
| EP4439484A1 (fr) | Procédé de classification de données multidimensionnelles fortement résolues | |
| FR3128015A1 (fr) | Système d’aide à la navigation d’un véhicule. | |
| WO2024009026A1 (fr) | Procede et dispositif de classification et de localisation d'objets dans des sequences d'images, systeme, programme d'ordinateur et support d'informations associes | |
| EP1426901B1 (fr) | Procédé pour détecter des cibles ponctuelles et système de veille mettant en oeuvre ce procédé de détection de cibles |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 24745480 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 2026500177 Country of ref document: JP Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: AU2024291267 Country of ref document: AU |
|
| ENP | Entry into the national phase |
Ref document number: 2024291267 Country of ref document: AU Date of ref document: 20240705 Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2024745480 Country of ref document: EP |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| ENP | Entry into the national phase |
Ref document number: 2024745480 Country of ref document: EP Effective date: 20260206 |
|
| WWP | Wipo information: published in national office |
Ref document number: 2024745480 Country of ref document: EP |