FR3055423A1 - METHOD FOR IDENTIFYING A VESSEL - Google Patents

Abstract

The invention relates to a method for identifying a ship, the method comprising obtaining a first synthetic radar image of inverse opening of a ship to be identified (8) and the relative position of the ship to be identified. and radar, the determination of a set of candidate ships for the vessel to be identified, the generation of a second synthetic reverse aperture radar image for each candidate vessel, the comparison of the first synthetic aperture radar image inverse with each second generated inverse aperture synthetic radar image, to obtain for each second synthesized inverse aperture generated radar image a similarity score, and the ranking of the similarity scores.

Description

055 423

01285 ® FRENCH REPUBLIC

NATIONAL INSTITUTE OF INDUSTRIAL PROPERTY © Publication number:

(to be used only for reproduction orders) (© National registration number

COURBEVOIE © Int Cl ⁸ : G 01 S 13/53 (2017.01)

PATENT INVENTION APPLICATION

A1

Date of filing: 31.08.16. (© Applicant (s): THALES Société anonyme - FR. © Priority : (43) Date of availability of the request: 02.03.18 Bulletin 18/09. @ Inventor (s): VEYER ANTOINE, GARREC PATRICK, GENIN JEAN BAPTISTE and CORRETJA VINCENT. (56) List of documents cited in the preliminary search report: See the end of this brochure © References to other related national documents: ©) Holder (s): THALES Société anonyme. o Extension request (s): (© Agent (s): CABINET LAVOIX Simplified joint-stock company.

104 / METHOD FOR IDENTIFYING A VESSEL.

FR 3 055 423 - A1 (5 /) The invention relates to a method of identifying a ship, the method comprising obtaining a first synthetic radar image of reverse aperture of a ship to be identified (8 ) and the relative position of the vessel to be identified and the radar, the determination of a set of candidate vessels for the vessel to be identified, the generation of a second reverse aperture synthetic radar image for each candidate vessel, the comparison of the first reverse aperture synthetic radar image with each second generated reverse aperture synthetic radar image, to obtain for each second generated reverse aperture radar radar image a similarity score, and the classification of the similarity scores .

Method for identifying a ship

The present invention relates to a method of identifying a ship using a reverse aperture synthetic radar and an associated computer program product.

Radar imagery is a technique for calculating representative images of elementary reflector maps. Radar imagery is based on the processing of backscatter signals from the object to be observed and recorded by a coherent radar system.

When the radar system has a broadband analyzer and a relative rotational movement between the radar and the object is present (synthetic aperture), it is possible to use imagery using the ISAR technique. The acronym ISAR refers to the English terminology "Inverse Synthetic Aperture Radar" translated into French by "synthetic reverse aperture radar". The opposite term indicates that movement is the movement of the object.

ISAR imagery provides two-dimensional images with good resolution of a given object. Such images are used in particular in maritime patrol aircraft to classify and identify ships.

Such a classification is carried out by an operator, which takes a long time and can generate errors.

There is therefore a need for a method of identifying a vessel which is simpler to implement and is more precise than the methods of identifying the state of the art.

To this end, the description describes a method of identifying a ship, the method comprising the steps of obtaining a first synthetic aperture radar image of a ship to be identified and of the relative position of the ship to identify and radar, determine a set of candidate vessels for the vessel to be identified, generate a second reverse aperture synthetic radar image for each candidate vessel, compare the first synthetic radar image of reverse aperture with each second reverse aperture synthetic radar image generated, to obtain for each second reverse aperture synthetic radar image generated a similarity score, and classification of similarity scores, the step of generating each second reverse aperture synthetic radar image comprising at least the steps of providing a geometric image comprising the candidate ship seen from the radar, the candidate ship and the radar being in the relative position of the ship to be identified and of the radar, for each point of the geometrical image corresponding to a part of the candidate ship, of determining the distance between the part of the ship candidate considered and the radar and the radial speed of the part of the candidate ship, for each point of the geometric image corresponding to a part of the candidate ship, simulating the signal amplitude received by the radar by random drawing of one value among a plurality of values, the construction of the image to be generated, each point of the image to be generated being associated in a one-to-one manner with a point of the geometric image and having as coordinates the determined distance, the speed determined radial and the value drawn randomly from the point of the geometric image considered.

According to particular embodiments, the method comprises one or more of the following characteristics, taken alone or according to all technically possible combinations:

- the determination involves a selection of candidate ships from a set of candidate ships according to a selection criterion.

- the selection criterion is a length interval of the candidate vessel.

- the method involves receiving an identifier of the nature of the vessel to be identified, the selection criterion consisting in eliminating the candidate vessels having a different identifier.

the comparison step is implemented by calculating a correlation between the first reverse aperture synthetic radar image with each second generated reverse aperture synthetic radar image, the similarity score being the value of the correlation .

- the method includes the display of second reverse aperture synthetic radar images generated for each candidate ship with a decreasing order of similarity score.

the step of generating each second synthetic aperture radar image includes the detection of the coordinates of each point of the geometric image corresponding to a part of the candidate ship and the step of generating each second synthetic radar image of reverse opening comprises the display of the geometrical image by a graphics engine comprising a depth buffer memory and the detection comprising the comparison of the values of the depth buffer memory before display and after display.

- the detection also includes the conversion of the depth into coordinates.

- the geometric image also includes the environment of the ship and the step of generating each second reverse-opening synthetic radar image includes a step of eliminating the points corresponding to the environment and the display of the image geometric by a graphics engine comprising a color buffer memory, the elimination step comprising the comparison of the values of the color buffer memory before display and after display.

The description also proposes a computer program product suitable for implementing a process for identifying a ship as described above.

The description also describes a computer medium storing instructions from the previous computer program product.

Other characteristics and advantages of the invention will appear on reading the following description of embodiments of the invention, given by way of example only and with reference to the drawings which are:

FIG. 1, a schematic view of a situation in which a ship is to be identified by control aircraft, FIG. 2, a schematic view of a graphical display engine and a screen forming part of the control aircraft,

FIG. 3, a flow chart of an example of implementation of a method for identifying the vessel to be identified, the method comprising a generation step,

- Figures 4 and 5, examples of views of ships, Figure 6, a flowchart of an example of implementation of part of the generation step,

- Figure 7, a schematic view of a geometric image of a ship seen by a radar,

FIG. 8, a view illustrating the passage matrix between several coordinate systems, FIG. 9, a graph showing the decomposition of the rotation in roll in a reference frame,

- Figure 10, a graph showing the decomposition of the rotation in pitch in a frame,

FIG. 11, a graph of the distribution of points in a distance and radial speed reference frame,

- Figure 12, a synthetic reverse aperture radar image of a ship seen by a radar, the image being obtained by implementing the determination method according to Figure 6, and Figures 13 to 15, examples of images obtained at various stages of the identification process according to FIG. 3.

FIG. 1 illustrates a schematic view of a situation in which a control aircraft 6 must identify a vessel to be identified 8.

In such an example, the control aircraft 6 is an aircraft for monitoring a fishing zone ZP.

The ZP fishing area is regulated, so that only certain types of vessels are allowed to fish in the ZP fishing area.

Therefore, it is desirable to be able to identify the vessel to be identified 8, especially since a direct interrogation of the vessel to be identified 8 should be confirmed by observation of the vessel to be identified 8 by the control aircraft 6.

The control aircraft 6 is thus capable of implementing a method of identifying the vessel to be identified 8.

According to the example in FIG. 1 and as visible in FIG. 2, the control aircraft 6 is equipped with a radar, not shown, with a screen 10 and with a graphic display engine 12.

Screen 10 is suitable for displaying images.

Screen 10 allows you to define a display window size. Usually, the size of the display window corresponds to the number of pixels controlled by the screen 10.

The graphical display engine 12 is more simply denoted graphical engine 12 below.

The graphics engine 12 is a set of components suitable for implementing an image display.

According to the example of FIG. 1, the graphics engine 12 comprises a processor 14, a depth buffer memory 16 and a color buffer memory 18.

The processor 14 is able to control each component of the graphics engine 12 and to implement calculation operations.

The depth buffer 16 is also called according to the English name "Z-buffer" meaning "buffer Z".

The depth buffer memory 16 is a memory memorizing a table associating for each pixel of an image the distance from the pixel to the observer of the image.

More precisely, the table of the depth buffer memory 16 has an identical size to the display window and gives for each pixel of the display window the distance from the pixel to the observer.

The color buffer memory 18 is also called according to the name "color buffer".

The color buffer memory 18 is a memory memorizing a table associating for each pixel of an image the color of the pixel in a colorimetric base.

The RGB base or the TSL base are two examples of colorimetric bases.

The operation of the control aircraft 6 is now described with reference to an exemplary implementation of a process for identifying a ship in accordance with the flow chart of FIG. 3.

The identification method comprises a reception step 30, a step of obtaining the relative position of the ship to be identified and of the radar 32, a step of obtaining a first reverse-synthesized radar image of a vessel to be identified 36, a determination step in interaction with a library 20, a generation step 34, a comparison step 38 and a classification step 40.

In the reception step 30, an identifier of the nature of the vessel to be identified is received 8.

For example, the vessel to be identified 8 provides a so-called AIS identifier, the identifier providing information giving the name of the vessel, the position of the vessel, the route of the vessel and the dimensions of the vessel.

In the step of obtaining 32 of the relative position of the vessel to be identified 8 and of the radar, information is provided on the relative position of the vessel to be identified 8 and of the radar.

According to the particular example illustrated, the following information is also provided in the step of obtaining 32: for the radar, position, speed, course and direction of aiming and for the ship, position, attitude, speed, course as well as the pitch and roll speeds.

According to another embodiment, the obtaining step 32 also includes the supply of the direction of sight of the radar, the distance between the ship and the radar and the angle of presentation of the ship.

The angle of presentation is defined as the angle between the direction of sight of the radar and the center line of the ship.

As a variant or in addition, the obtaining step 32 also includes the supply of information relating to sea movements, wind direction or wind strength.

In the obtaining step 36, it is obtained from a first synthetic radar image of reverse aperture of the vessel to be identified 8.

In the example described, the obtaining step 36 is implemented by acquisition of an image by the radar of the control aircraft 6.

As a variant, the vessel to be identified 8 is imaged by a radar belonging to another vessel or an aircraft working in interaction with the control aircraft 6.

In each of the embodiments, at the end of the obtaining step 36, a first reverse-opening synthetic radar image of the vessel to be identified 8 is obtained, this first image corresponding to an image obtained under real conditions of shooting.

During the determination step, a set of candidate vessels is determined for the vessel to be identified 8.

For this, the library 20 is used which stores geometric images of a set of candidate ships as shown in Figures 4 and 5.

According to the example of FIG. 3, the determination comprises a selection of candidate ships from a set of candidate ships according to a selection criterion.

By way of illustration, the selection criterion is an interval of the length of the candidate vessel.

According to another example or in addition, several criteria can be used, the selection criterion consisting in eliminating the candidate vessels having a different identifier, in particular a different AIS identifier.

At the end of the determination step, the number of candidate ships available in library 20 is reduced.

In some embodiments, the number of candidate ships is reduced by

90%.

The number of candidate vessels is generally less than a few hundred. A step of generating a second reverse aperture synthetic radar image for each candidate ship is then implemented.

According to the particular example, the generation step is carried out by implementing a generation method which will now be described with reference to FIGS. 6 to 11.

The generation process generates a synthetic reverse aperture radar image of a target seen by the radar.

The target has several parts, at least two parts having a distinct radial speed. The radial speed is defined with respect to the radar, the radial speed being the projection of the speed on the line of sight of the radar. The line of sight of the radar is the straight line passing through the radar and the center of the target.

In the present example, the target is a candidate ship which would be seen by the radar in the relative position of the ship to be identified 8 and of the radar.

The generation process includes a supply phase P1, a processing phase P2 and a generation phase P3 proper.

Only the processing phase P2 is illustrated diagrammatically in FIG. 6.

The supply phase P1 comprises a step of supplying a geometric image comprising the candidate ship seen from the radar, the candidate ship and the radar being in the relative position of the ship to be identified 8 and of the radar.

A geometric image is an image of the ship seen from the radar such that the ship would be seen by a camera operating in the visible.

The geometrical image makes it possible to visualize at least the exterior of the ship, in particular its hull and its superstructures.

According to one embodiment, the geometrical image makes it possible to visualize only an emerged part of the ship.

According to another embodiment, the geometric image results from a reconstruction.

For example, it is assumed that the providing step includes providing a geometric image including the ship seen from the radar, the ship and the radar being in the desired relative position.

The relative positioning is carried out by the processor 14 from a geometric image accessible in particular by the shipbuilder.

For the rest, a reference frame is defined. For example, the origin of the coordinate system is vertical to the radar and at zero altitude, the axes being an X axis _ref towards the North, a Y axis _ref towards the East and a Z axis _ref down.

A benchmark linked to the ship is also defined. The coordinate system linked to the ship has an origin at the center of the ship and three axes which are a forward Xnavire axis, a _ship Y axis to the right and a _ship Z axis downwards.

More precisely, during the reconstruction, the ship is inserted at zero altitude using its relative position with the radar as well as its three attitude angles (heading, pitch and roll).

In addition, an environment is also inserted to obtain a horizontal flat surface of predefined color, of zero altitude, representing the sea.

An orthographic projection of the entire ship and the environment is then implemented.

The orthographic projection depends on the size of the display window and the distance of the section planes along the line of sight (near plane and far plane).

All of the above operations are, according to one embodiment, implemented by the graphics engine 12.

An example of a geometric image obtained at the end of the implementation of the supply step is illustrated in particular in FIG. 7.

In this image, the vessel is noted N and the environment which is the sea is noted

E.

The processing phase P2 comprises, according to the example of FIG. 6, a step 50 of displaying an image, a step of obtaining 52 of particular pixels of the image, a step of calculation 54 and a step of simulation 56.

In display step 50, the geometric image obtained in the supply step is displayed.

The geometric image is displayed using the graphics engine 12.

In the display step, the depth buffer 16 and the color buffer 18 are initialized to a predefined value corresponding to an erasure value.

The display changes the depth value for certain pixels in the depth buffer 16 and gives a color value for each pixel in the color buffer 18.

In the step of obtaining 52 pixels, the points of the geometric image corresponding to a part of the ship N are obtained.

This eliminates the pixels that are part of environment E or the parts of the ship that are masked by environment E.

For this, during the obtaining step, it is detected coordinates of each point of the geometric image corresponding to a part of the ship N.

According to the example illustrated, the detection comprises the comparison of the values of the depth buffer memory 16 before display of the geometric image and after display of the geometric image.

This makes it possible to obtain pixels for which the value of the depth buffer 16 is modified by the display which are pixels corresponding to a part of the ship and if necessary to the environment.

As the geometric image also includes the environment E of the ship N, in the obtaining step 52, a step of eliminating the points corresponding to the environment E. is implemented.

According to the example illustrated, the elimination step comprises the comparison of the values of the color buffer memory before display and after display.

This makes it possible to obtain candidate pixels from which a selection is made to select the pixels corresponding to a predefined color.

For example, the predefined color is blue for the sea.

Also, according to an embodiment using an RGB colorimetric base, the candidate pixels selected are the pixels having a preponderant component of blue (for example, greater than 75%, the sum of the red and green components being less than 25%).

The selected candidate pixels are then eliminated.

Thus, in the step of obtaining 52 pixels, the points of the geometric image corresponding to a part of the ship N are obtained.

The calculation steps 54 and simulation 56 are implemented for each point of the geometric image corresponding to a part of the ship N.

This is illustrated by test 58, as long as a pixel corresponding to a part of the ship N has not been processed in the image, the calculation steps 54 and simulation 56 are repeated, as shown by arrow 60 on FIG. 6. When all the pixels corresponding to a part of the ship N have been processed, the result is obtained, which is shown by the result 62 in FIG. 6.

Calculation step 54 comprises on the one hand the determination of the distance between the part of the ship N considered and the radar and on the other hand the determination of the radial speed of the part of ship N.

First, the distance between the part of the ship N considered and the radar is determined.

For example, for the pixels corresponding to ship N, the position of each of the pixels is calculated.

For this, the value V of the depth buffer memory 16 is read to provide the distance between the radar and the part of the ship N considered.

The value read V is a normalized value, for example between 0 and 1.

The distance is then calculated using the following formula:

D = dp + V x (d, - dp)

Or :

• D is the orthographic distance, • dp and d | define the distances of the two section planes used in the orthographic projection of the scene.

It is then determined the position of a pixel in the display plane perpendicular to the line of sight. Such a determination is illustrated in particular in FIG. 8.

The position is given by three coordinates X, Y and Z, these coordinates being given by the following relationships:

(x = D

L C * L

Y = - - + <2 L (in pixels)

H R * H

Z = - + v 2 H (in pixels)

Or :

• R and C are the positions of the pixel in coordinates on the screen 10 (the origin being located in the bottom left corner), • L indicates the horizontal dimension of the orthographic projection of the rendering of the scene in meters while L (in pixels) gives the horizontal dimension of the screen 10, and • H denotes the vertical dimension of the orthographic projection of the rendering of the scene in meters while H (in pixels) gives the dimension of the vertical screen of the screen 10.

It then converts the three X, Y and Z coordinates to pixel coordinates.

For example, such a conversion uses the fact that in the reference frame, the line of sight is defined by A the azimuth angle (angle of rotation in the horizontal plane) and S the site angle (angle of rotation in the vertical plane) as well as the fact that the projection is orthographic.

As a result, the conversion is implemented by applying the following pass formula:

—Sin (yl) cos (? L) 0

Where X _r , Y _r and Z _r denote the coordinates of the pixel in the reference frame R.

The distance D between the radar and the pixel is then calculated by applying the following formula:

D = ^ x _r ² + Yr ² + zr ²

In a second step, the radial speed of the part of the ship is determined.

The ship has a speed v _x along the axis X of its reference frame and rotation speeds n _t in pitch and n _r in roll. The pitch rotation vector is carried by the Y axis, the roll rotation vector by the X axis. The components of the pixel speed which are noted by the points P and Q in FIGS. 9 (relative to the roll rotation) and 10 (relative to pitch) can be expressed as follows in the coordinate system linked to the ship:

V = V _x X + 0ΡΩ _Γ u + 0Qn _t v

This speed is then expressed in the reference frame using the inverse matrix of the passage matrix from the reference frame to the ship's frame.

Let us call M the passage matrix from the reference frame to the ship's frame. The matrix M is calculated in a conventional manner using the three angles of attitude of the ship heading, pitch and roll. The coordinates of the pixel in the coordinate system linked to the ship are given by:

/ Xp \ / ^ oo M ₀₂ \ / X _r -X _n \

Y _p = M _lo Mu M ₁₂ Y _r - Y _n \ zj \ M ₂₀ M ₂₁ M ₂₂ z \ Z _r - Z _n / where (x _n , Y _n , Z _n ) represents the coordinates of the origin of the ship in the reference frame.

By using in this coordinate system the speed of the pixel V _p and the speed of the radar V _o which were previously supplied, the radial speed is expressed by the equation:

v _R = (Vp - Vo) ^x 5 o?

Next, simulation step 56 is implemented.

For this, the signal amplitude received by the radar is simulated by randomly drawing a value from a plurality of values.

For this, the plurality of values comprises more than 10 distinct values, preferably more than 100 distinct values.

For example, according to the example described, the plurality of values comprises 256 values, that is to say the whole values comprised in the broad sense between 0 and 255.

According to a first embodiment, the draw is a uniform random draw.

According to a second embodiment, the drawing follows a normal or Gaussian law.

According to a third embodiment, the printout is weighted by the height of the pixels relative to the sea level.

Symbolically, as illustrated in FIG. 11, the whole of the implementation of the simulation steps 56 corresponds for the graph of the distribution of points in a distance and radial speed reference frame to choose a value for each of the points represented. In this graph, the origin of the horizontal axis coincides with the distance from the origin of the ship and the origin of the vertical axis corresponds to a zero radial speed.

The generation phase P3 comprises a step of construction of the image to be generated, each point of the image to be generated being associated in a one-to-one manner with a point of the geometric image and having as coordinates the determined distance, the speed determined and the value drawn randomly from the point of the geometric image considered

A reverse aperture synthetic radar image is thus obtained for each candidate ship as illustrated diagrammatically in FIG. 12.

The comparison step 38 is then implemented. During the comparison step 38, the first reverse aperture synthetic radar image is compared with each second generated reverse aperture synthetic radar image. It is thus obtained for each second reverse aperture synthetic radar image generated a similarity score.

According to an example, the comparison step is implemented by calculating a correlation between the first synthetic image with reverse aperture synthesis with each second radar image with synthetic reverse aperture generated, the similarity score being the value of the correlation.

As a specific example, the correlation uses a specific metric. The Kullback-Leiber divergence or the Itakura-Saito distance are two examples of metrics that can be advantageously used in this context.

Then, the classification step 40 is implemented.

For this, the similarity scores obtained in comparison step 38 are classified.

According to a particular example, the method comprises the display of second reverse aperture synthetic radar images generated for each candidate ship with a decreasing order of similarity score.

The operator then only has to confirm the recognition carried out automatically.

The efficiency of the process can easily be understood by comparing Figures 13 to 15. For the same candidate ship, Figure 13 corresponds to an ISAR image from any angle obtained by the generation process while Figure 14 corresponds to an ISAR image generated with the right angle by implementing the proposed process. The visual comparison with the image in FIG. 15 which corresponds to an actual ISAR image clearly shows that the identification is easier with FIG. 14 than with FIG. 13.

This shows that the method of identifying a vessel is simpler to implement and is more precise than the methods of identifying the state of the art.

Simplicity results in particular from the generation process used in the generation step.

The general principle of the generation process is as follows: display via the graphic engine 12 of the ship N lit by the radar while respecting its relative position and its angles of presentation; for each pixel in the image of ship N, calculation of its position and its distance from the radar; knowing the speed of rotation in pitch and roll of the ship and the position of the pixel, calculating its speed, then its radial speed; random drawing of a value representing the level (coded from 0 to 255 for example) of the signal returned by this pixel and knowing the distance and the radial speed, insertion of the point in the synthetic radar image of reverse aperture.

In summary, the generation method makes it possible from the geometric image of a vessel N displayed via a graphic engine 12 according to the angle of presentation of the vessel N and the angle of sight of the radar to obtain the distance of the points of the ship lit by the radar and therefore their position.

The generation process thus simplifies the implementation of the generation of reverse aperture synthetic radar images.

The generation process avoids the use of the radar wave propagation equation.

The generation process uses only the geometry of the vessel contained in a file in standard commercial format. Reference is no longer made to the materials or to the reflection coefficients of the ship's superstructures.

The generation process is implemented using the capabilities of a graphics engine 12. In fact, the calculations related to the display are entirely performed by the graphics engine 12. An additional advantage is that these calculations are offset to the graphics engine 12. This makes it possible to avoid using a dedicated processor.

Thus, the method for generating a synthetic image of reverse aperture radar is simpler to implement while at the same time preserving at least the precision of the generation methods based on a decomposition into elementary facets of the object.

This proves that the method of identifying a ship is simpler to implement and is more precise than the methods of identifying the state of the art.

Other variants corresponding to combinations of the embodiments described above can also be envisaged for the method of identifying the vessel to be identified 8.

In particular, it is possible to implement the process of identifying the ship with a computer program system and product. The interaction of the computer program product with the system enables this process to be implemented.

The system is a computer.

More generally, the system is an electronic computer capable of manipulating and / or transforming data represented as electronic or physical quantities in system registers and / or memories into other similar data corresponding to physical data in memories, registers or other types of display, transmission or storage devices.

The system includes a processor including a data processing unit, memories and an information carrier. The system also includes a keyboard and a display unit.

The computer program product contains a readable medium.

A readable medium of information is a medium readable by the system, usually by the data processing unit. The readable information medium is a medium suitable for storing electronic instructions and capable of being coupled to a bus of a computer system.

By way of example, the readable information medium is a floppy disk or flexible disk (from the English name of "floppy disk"), an optical disk, a CD-ROM, a magneto-optical disk, a ROM memory, a RAM memory, EPROM memory, EEPROM memory, magnetic card or optical card.

On the readable information carrier is stored a computer program comprising program instructions.

The computer program is loadable on the data processing unit and is adapted to cause the implementation of a method as previously described when the computer program is implemented on the data processing unit .

Claims (10)

1 Process for identifying a ship, the process comprising the steps: - obtaining a first synthetic radar image of reverse aperture of a vessel to be identified (8) and of the relative position of the vessel to be identified (8) and of the radar, - determining a set of candidate vessels for the vessel to be identified (8), - generation of a second reverse aperture synthetic radar image for each candidate vessel, - of comparison of the first synthetic image of reverse aperture synthesis with each second radar image of synthetic reverse aperture generated, to obtain for each second radar image of synthetic reverse aperture generated a score of similarity, and - classification of similarity scores, the step of generating each second reverse aperture synthetic radar image comprising at least the steps of: - supply of a geometrical image comprising the candidate ship seen from the radar, the candidate ship and the radar being in the relative position of the ship to be identified and the radar, - for each point of the geometrical image corresponding to a part of the candidate ship, determination of the distance between the part of the candidate ship considered and the radar and of the radial speed of the part of the candidate ship, - for each point of the geometric image corresponding to a part of the candidate ship, simulation of the signal amplitude received by the radar by random drawing of a value from a plurality of values, - construction of the image to be generated, each point of the image to be generated being associated in a one-to-one manner with a point of the geometric image and having as coordinates the determined distance, the determined radial speed and the value drawn randomly from the point of the geometric image considered. 2, - Method according to claim 1, wherein the determination comprises a selection of candidate ships from a set of candidate ships according to a selection criterion. 3. - Method according to claim 2, wherein the selection criterion is an interval of length of the candidate ship. 4. - Method according to claim 2 or 3, wherein the method comprises receiving an identifier of the nature of the vessel to be identified (8), the selection criterion consisting in eliminating the candidate vessels having a different identifier. 5. - Method according to any one of claims 1 to 4, wherein the comparison step is implemented by calculating a correlation between the first synthetic aperture radar image with each second synthetic aperture radar image of reverse aperture generated, the similarity score being the value of the correlation. 6. - Method according to any one of claims 1 to 5, in which the method comprises the display of second synthetic radar images of reverse aperture generated for each candidate ship with a decreasing order of score of similarity. 7. - Method according to any one of claims 1 to 6, in which the step of generating each second reverse aperture synthetic radar image comprises the detection of the coordinates of each point of the geometric image corresponding to a part of the candidate ship and the step of generating each second reverse aperture synthetic radar image comprises the display of the geometric image by a graphic engine (12) comprising a depth buffer memory (16) and the detection comprising the comparison of the values of the depth buffer memory (16) before display and after display. 8. - Method according to claim 7, wherein the detection also comprises the conversion of the depth into coordinates. 9. - Method according to any one of claims 1 to 8, wherein the geometrical image also comprises the environment (E) of the ship and the step of generation of each second radar image with reverse aperture synthesis comprises a step of eliminating the points corresponding to the environment (E) and the display of the geometric image by a graphics engine (12) comprising a color buffer memory (18) and in which the elimination step comprises the comparison of the values of the color buffer '18) before display and after display. 10. - A computer program product suitable for implementing a process for identifying a ship according to any one of claims 1 to 9. 1/6