Processing data acquired by a lidar sensor

Abstract

Examples present a method for processing data acquired by a lidar sensor, a computer configured to perform this processing, a vehicle carrying a computer configured for this purpose, a computer program product, and a non-transitory recording medium readable by a computer. Abstract figure: [Fig. 3]

Description

Title of the invention: Processing of data acquired by a lidar sensor Technical field

[0001] The present disclosure relates to the field of data processing. Prior art

[0002] The processing of data acquired by sensors is increasingly used in real time to assess situations and make decisions in light of these situations.

[0003] For example, there are methods which make it possible to follow, in real time, characteristics of an object identified on a succession of images acquired by a camera so as to estimate a relative movement of the object with respect to the camera, which can in particular make it possible to anticipate the movement of the object.

[0004] These methods are extremely useful, particularly in the context of driving assistance or autonomous driving functions, since they make it possible to follow, in real time, the motor vehicles surrounding the vehicle on which the cameras are mounted and possibly to anticipate the trajectory of these vehicles.

[0005] Methods for tracking characteristics of an object can however be improved. Summary

[0006] In this regard, a method is proposed, implemented by a computer, for processing data acquired by a lidar sensor, the method comprising: - obtaining a first and a second matrix of points from acquisitions by a lidar sensor, the points of the first and second matrices respectively representing the environment of the lidar sensor at a first and a second instant of time; each point of the matrix of points being associated with coordinates in a three-dimensional space and with an intensity value; - a projection onto a two-dimensional projection plane of the point matrices so as to obtain, for each of the matrices, an intensity image and a depth image; for each of the first and second point matrices: * a determination of at least one characteristic belonging to an element of the environment of the lidar sensor on the point matrix, from the three-dimensional coordinates and the intensity values ​​of the points of the point matrix, a characteristic thus being associated with a point of the matrix; then for at least one characteristic determined on each matrix: * a determination, in the intensity image, of a first window of neighboring points linked to the characteristic and comprising the point associated with the characteristic, from the two-dimensional coordinates of the points of the matrix and their intensity value in the intensity image; * a determination of a second window of neighboring points linked to the characteristic, the neighboring points of the second window of neighboring points corresponding to the neighboring points of the first neighboring window which have a distance with the point representing the characteristic less than a predetermined threshold in the three-dimensional space or in the depth image; then - a comparison of a second window of neighboring points of the first point matrix to a second window of neighboring points of the second point matrix; then - an association between a characteristic linked to a second window of neighboring points of the first point matrix and a corresponding characteristic linked to a second window of neighboring points of the second point matrix from the comparison.

[0007] Optionally, the method may further comprise a determination of a displacement of the characteristic belonging to the element of the environment of the lidar sensor between the first instant of time and the second instant of time, from the two second windows of neighboring points linked to the corresponding associated characteristics.

[0008] Optionally, the determination of a displacement of the characteristic belonging to the element of the environment of the lidar sensor between the first instant of time and the second instant of time may comprise: - a determination of a two-dimensional displacement of the characteristic between the first instant of time and the second instant of time, in the intensity image, from the two-dimensional coordinates and the intensity values ​​of the points of the two second windows of points associated with the characteristic in the intensity image; and - a determination of a three-dimensional displacement of the characteristic between the first instant of time and the second instant of time, from the two-dimensional displacement in the intensity image, and from the two-dimensional coordinates and depth values ​​of the points of the two second windows of points associated with the characteristic in the depth image.

[0009] Optionally, the method may further comprise for a second window of neighboring points linked to a characteristic: - a determination of a descriptor of the second window of neighboring points, a descriptor corresponding to a characteristic value of the window of neighboring points determined from the coordinates and the intensity values ​​of the points of the second window of points in the intensity image; and and the association between a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points can be carried out from a distance between the descriptor associated with the second window of neighboring points of the first matrix and the descriptor associated with the second window of neighboring points of the second matrix.

[0010] Optionally, the operations of determining a descriptor of a second window of neighboring points and of associating a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points from a distance between the descriptors of these windows can be implemented from a method of robust and independent binary elementary characteristics.

[0011] Optionally, the operations of determining a descriptor of a second window of neighboring points and of associating a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points can be implemented from a Lucas-Kanade method. In this option, the displacement of the characteristic of the element between the first instant of time and the second instant of time can be determined from a minimization of the distance between the descriptors of the two windows of neighboring points linked to the corresponding associated characteristics.

[0012] The application also relates to a computer configured to implement any one of the data processing methods presented by the present disclosure and to a vehicle carrying a computer having one of these configurations.

[0013] The application further relates to a computer program product comprising instructions for implementing any of the methods presented by the present disclosure when this program is executed by a processor.

[0014] Finally, the application relates to a non-transitory recording medium readable by a computer on which is recorded a program for implementing any of the methods presented by the present disclosure when this program is executed by a processor.

[0015] The method according to the present disclosure therefore makes it possible, from information acquired from a lidar sensor, to follow a characteristic of an element of the environment of the lidar sensor between two point matrices, possibly more, to trace a characteristic of the element of the environment of the lidar sensor over time. Thus, in applications in which the lidar sensor is on board a motor vehicle, the traced characteristic may for example belong to another motor vehicle so that motor vehicles traveling near the vehicle carrying the lidar sensor can be tracked. The method makes it possible, in particular, to track characteristics over time even in the event of occlusion situations. Thus, in options, the method can in particular make it possible to determine a relative displacement of the characteristic over time. Brief description of the drawings

[0016] Other characteristics, details and advantages will appear on reading the detailed description below, and on analyzing the attached drawings, in which:

[0017] [Fig.l] schematically represents an example of a data processing device allowing the implementation of a method for processing data acquired by a lidar sensor.

[0018] [Fig.2] schematically represents an example of a vehicle comprising a device data processing and a lidar sensor.

[0019] [Fig.3] schematically represents an example of a method for processing data acquired by a lidar sensor.

[0020] [Fig.4] schematically represents an example of two point matrices presenting an occlusion situation. Description of the embodiments

[0021] An example of a data processing device 1 is now described with reference to [Fig. 1] for carrying out a method for processing data acquired by a lidar sensor 10 and in particular the example of a data processing method presented later with reference to [Fig. 3].

[0022] The data processing device 1 can be adapted to be mounted on a vehicle 2.

[0023] The data processing device 1 comprises a computer 11 and a memory 12. The device is configured to process data acquired by a lidar sensor 10 (Light Detection And Ranging in English).

[0024] The memory 12 can store the code instructions executed by the computer 11 and making it possible to control the acquisition of data by the lidar sensor 10 as well as their processing. The computer therefore has access to the information stored in memory. The memory 12 may also be adapted to store data acquired by the lidar sensor 10.

[0025] The memory 12 may for example comprise a ROM (Read-Only Memory), a RAM (Random Access Memory), an EEPROM (Electrically Erasable Programmable Read-Only Memory) or any other type of suitable storage means. The memory may for example comprise optical, electronic or magnetic storage means.

[0026] The data processing device 1 may be included in a motor vehicle, as shown in the example of [Fig. 2]. In this figure, a lidar sensor 10 is also shown allowing the acquisition of the data processed by the data processing device.

[0027] A lidar sensor 10 (Light Detection And Ranging in English) is a sensor emitting light waves and determining, from the reflection of these light waves, a matrix of points representing an environment of the lidar sensor.

[0028] The lidar sensor 10 is adapted to acquire point matrices. Each point is associated with three-dimensional coordinates (x, y, z) representing the environment of the lidar sensor 10 and with an intensity value. With regard to the three-dimensional coordinates, the x coordinate of a point corresponds to an abscissa coordinate of the point relative to the sensor 10. The y coordinate of a point corresponds to an ordinate coordinate of the point relative to the sensor 10. The z coordinate of a point corresponds to a depth coordinate of the point relative to the lidar sensor 10. In the example shown in [Fig. 2], and assuming the contours of the vehicle 2 and the lidar sensor 10 are shown in top view, the abscissa X and depth Z axes of the lidar sensor 10 are shown. The Y ordinate axis of the sensor 10 is perpendicular to the X and Z axes. In the example illustrated in [Fig.2], the lidar sensor 10 is mounted at the front of the vehicle and is oriented to emit light beams in the direction of travel of the vehicle.

[0029] Each point acquired by the lidar sensor 10 is also associated with an intensity value. This is the intensity received by the lidar sensor after reflection of the light beam on a surface.

[0030] In first examples, the lidar sensor 10 of the present disclosure may correspond to a scanner-type lidar sensor, also referred to by the terminology lidar scanner. This is a lidar sensor which acquires a matrix of points representing its environment in which the points of the matrix are each associated with a different instant of time. More precisely, each point of the matrix of points is acquired via the emission of a distinct light beam so that there is a time difference between the acquisition of each of the points of the matrix of points representing the environment of the lidar.

[0031] In second examples, the lidar sensor 10 of the present disclosure may correspond to a flash lidar, also referred to by the terminology flash lidar. Unlike the scanner lidar sensor, a flash lidar sensor acquires a matrix of points representing its environment by emitting a single light beam, of wide section, so that each point of the matrix of points is acquired at the same instant of time.

[0032] There is now presented, with reference to [Fig.3], an example of method 100 of processing of data acquired by a lidar sensor 10. The method can for example be implemented by the computer 11 of the data processing device 1.

[0033] It should be noted that [Fig. 3] is only an illustration of the example of the method 100 representing, by blocks, the different operations possibly included in the method and described in the remainder of the document. As such, this illustration does not reflect any sequentiality between the operations except when such sequentiality is expressly explained by the present disclosure. In other words, the operations described with reference to [Fig. 3] are not necessarily implemented one after the other and may be implemented in a different order from that shown in [Fig. 3]. Furthermore, it is not necessary for each operation to be implemented once before the same operation is repeated a second time. The frequency of implementation of each operation is specific to it and is not necessarily linked to the implementation of the other operations.

[0034] As illustrated in [Fig. 3], the method 100 comprises an operation 110 of obtaining a first matrix and a second matrix of points. These matrices of points are obtained from acquisitions by a lidar sensor 10. The points of the first and second matrices of points respectively represent the environment of the lidar sensor at a first and a second instant of time. As explained previously, each point of the matrices of points is associated with a three-dimensional coordinate and an intensity value.

[0035] In the first examples in which the lidar sensor 10 is a scanner-type lidar sensor, obtaining the first and second matrix of points may comprise a correction of the three-dimensional coordinates of the points of each of these matrices so as to compensate for the time shift existing between the respective points of each matrix. Indeed, in examples in which the lidar scanner 10 moves during the acquisition of the points of a matrix, for example when it is mounted on a motor vehicle, the time shift between the acquisition of each point of the matrix will result in a shift in the three-dimensional coordinates of the points which must be corrected so that all of the points of a matrix are considered to be acquired at the same instant of time. Methods for correcting this shift are known to those skilled in the art.In particular, a method for correcting this offset is for example presented by the document written by QIN et al. High-Precision Motion Compensation for LiDAR Based on LiDAR Odometry. .

[0036] It is further understood that this correction is not necessary in the second examples in which the lidar sensor 10 is a flash type lidar sensor insofar as the respective points of each of the matrices are acquired simultaneously. Indeed, the acquisition of two matrices of points at two different instants of time by a flash type lidar makes it possible to directly obtain two matrices of points associated with two different moments of time.

[0037] As illustrated in [Fig. 3], the method 100 then comprises an operation 120 of projection onto a two-dimensional projection plane of the two point matrices so as to obtain, for each of the matrices, an intensity image and a depth image. An intensity image is an image comprising a plurality of points in a two-dimensional space, each point being associated with an intensity value. As regards the depth image, it is an image comprising a plurality of points in a two-dimensional space, each point being this time associated with a depth value. When the points of the matrices are projected into a two-dimensional space while retaining an intensity value for the intensity image and a depth value for the depth image, these points can also be called pixels of the intensity image and pixels of the depth image.

[0038] As illustrated in [Fig. 3], the method 100 then comprises an operation 130, implemented on each of the first and second matrices, of determining at least one characteristic belonging to an element of the environment of the lidar sensor on the matrix of points so as to associate the characteristic with a point of the matrix. The determination of at least one characteristic is carried out from the three-dimensional coordinates and the intensity values ​​of the points.

[0039] The term characteristic must be interpreted in the present disclosure in the sense that it takes in the field of image processing, the image here being the matrix of points considered. In this same field, the French term “characteristic” is generally designated by the English term “feature”. In this case, in the field of image processing, the term “characteristic” can refer to a visual attribute or a distinctive property of an image, such as contours, textures, patterns, colors, shapes, angles, corners, etc.

[0040] In examples, the operation 130 of determining at least one characteristic belonging to an element of the environment of the lidar sensor on a considered point matrix may comprise the determination of at least one of an outline, a texture, a pattern, a color, a shape, an angle, or a corner of an element of the environment of the lidar sensor.

[0041] In examples, the operation 130 of determining at least one characteristic belonging to an element of the environment of the lidar sensor on the dot matrix may comprise an angle of a vehicle.

[0042] The characteristic determined on the point matrix is ​​associated with a point of the point matrix considered. In examples in which the characteristic determined is included in several points of the point matrix, the point of the point matrix associated with the characteristic is chosen from the points comprising a subsection of the determined feature. In examples in which the determined feature is included in a single point of the dot matrix, the point of the dot matrix associated with the feature corresponds to the point comprising the feature.

[0043] In examples, a characteristic of an element of the environment of the lidar sensor 10 is determined on a matrix of points from an intensity gradient.

[0044] As illustrated in [Fig. 3], the method 100 then comprises an operation 140, implemented for at least one characteristic determined on each matrix. Advantageously, the operation 140 is implemented on each characteristic determined on the first and on the second matrix during the operation 130.

[0045] Operation 140 is performed in the intensity image of the matrix. It corresponds to a determination of a first window of neighboring points linked to a considered characteristic of the matrix, and comprising the point associated with the characteristic in the intensity image. The determination of the first window of neighboring points is performed from the two-dimensional coordinates of the points of the matrix and their intensity value in the intensity image. In this case, when a characteristic is determined on a matrix during operation 130 and it is associated with a given point of the matrix, it is entirely possible to follow this projected point in the intensity and / or depth image. Thus, a first window of neighboring points can be determined, in the intensity image, comprising this point associated with the characteristic.A neighbor point window, as its name suggests, comprises a plurality of points positioned next to each other in the two-dimensional space of the image on which it is determined.

[0046] In examples, the first window of neighboring points linked to a specific feature may include the point associated with the specific feature and the neighboring points of this point, i.e., the points having a distance less than a predetermined distance threshold from the point associated with the specific feature. In these examples, the point associated with the specific feature may correspond to a central point of the first window of neighboring points.

[0047] As illustrated in [Fig. 3], the method 100 then comprises an operation 150 of determining a second window of neighboring points linked to the characteristic. The neighboring points of the second window of neighboring points correspond to the neighboring points of the first window of neighboring points which have a distance with the point representing the characteristic less than a predetermined threshold in the three-dimensional space or in the depth image. This operation involves removing the points of the first window of neighboring points which are too far from the point associated with the characteristic in the three-dimensional space or in the depth image to determine the second window of neighboring points. Indeed, these points, as soon as they are too far from the point representing the characteristic in a space associated with a depth value, can potentially belong to an element other than the element comprising the characteristic determined in the space. However, and as will be described below, the method 100 makes it possible to follow the movement of a characteristic between two instants of time on the basis of the second windows of points associated with this characteristic during these two instants of time, i.e. the first instant of time being represented by the first matrix while the second instant of time being represented by the second matrix.

[0048] As illustrated in [Fig. 3], the method 100 then comprises an operation 160 of comparing a second window of neighboring points belonging to the first matrix of points (at the first instant of time) to a second window of neighboring points belonging to the second matrix of points (at the second instant of time). The comparison of the second windows of points is carried out in the intensity image. Advantageously, each second window of neighboring points of the first matrix of points is respectively compared to each second window of neighboring points of the second matrix of points.

[0049] In examples, the operation 160 of comparing a second window of neighboring points belonging to the first matrix of points to a second window of neighboring points belonging to the second matrix of points comprises a determination of a distance between the windows of neighboring points compared and a comparison of the distance between the windows of neighboring points compared. The distance used during this operation may in particular correspond to a Hamming distance.

[0050] As illustrated in [Fig. 3], the method 100 comprises an operation 170 of association between a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points from the comparison. Thus, during the operation 170, a correspondence is established between a characteristic of the first matrix of points and a characteristic of the second matrix of points from their respective second window of neighboring points.

[0051] In examples, a feature related to a second window of neighboring points of the first point matrix is ​​associated with a feature related to a second window of neighboring points of the second point matrix when the distance between their respective window of points is less than a predetermined threshold.

[0052] The method 100 according to the present disclosure therefore makes it possible, from information acquired from a lidar sensor 10, to follow a characteristic of an element of the environment of the lidar sensor between two point matrices. It is further understood that the method is presented for two point matrices but can be implemented works on more than two point matrices, in particular when they are acquired successively, so as to trace a characteristic of the element of the environment of the lidar sensor 10 over time. Thus, in applications in which the lidar sensor 10 is embedded in a motor vehicle, the traced characteristic may for example belong to another motor vehicle so that motor vehicles traveling near the vehicle carrying the lidar sensor 10 can be tracked, for example to estimate the speed of these motor vehicles or their respective trajectory. Of course, many other applications can be envisaged and the present disclosure is not limited to automotive applications alone.

[0053] The tracking of the characteristic is carried out, in the present disclosure, from information acquired by a lidar sensor 10. The lidar sensor 10 has the advantage, in particular compared to a camera, of acquiring points having coordinates in a three-dimensional space. Consequently, tracking of characteristics carried out on the basis of three-dimensional coordinates is more precise than tracking carried out on the basis of two-dimensional coordinates.

[0054] The method 100, via the acquisition of data from the lidar sensor 10, makes it possible, for example, to carry out tracking of characteristics even in the event of occlusion situations. Occlusion situations designate situations in which a characteristic to be tracked, belonging to a first element of the sensor's environment, is partly masked or hidden by a second element of the sensor's environment on a matrix of points. In these situations, defining a window of neighboring points linked to the characteristic of the first element of the sensor's environment in a matrix of points without considering depth information of these points can make tracking this characteristic difficult to the extent that such a window could comprise points belonging to the second element, distinct from the first element, in the sensor's environment, this second element being able to move relative to the first element.Thus, the association of corresponding characteristics of two point matrices which is based on a comparison of the windows of neighboring points linked to these characteristics is impacted since a window of neighboring points of a first point matrix representing the environment of the sensor at a time t1 may include points belonging to the second element whereas a window of neighboring points of a second matrix representing the environment of the sensor at a time t2 will no longer include any if the second element has moved relative to the first element between times t1 and t2. The method according to the present disclosure makes it possible to manage this type of situation by considering in particular that the points determined in the second window of neighboring points correspond to points close to the characteristic in the three-dimensional space or in the depth image (i.e. in . (considering depth information). Consequently, the probability that some points composing the second window of neighboring points belong to an element other than the element comprising the characteristic to be followed over time is reduced.

[0055] An example of an occlusion situation is notably illustrated schematically in [Fig.4], in order to facilitate the understanding of this type of situation. In this figure, a first matrix of points M1 representing the environment of the camera at a time instant t1 and a second matrix of points M2 representing the environment of the camera at a time instant t2 are represented. On the first matrix of points M1, a first window of neighboring points fl linked to a characteristic C is schematically represented and comprises points belonging to a first element El and to a second element E2, partially masking the first element El. The first window of neighboring points fl is linked to a characteristic C belonging to the element El.On the second matrix of points M2, a second window of neighboring points f2 also linked to the characteristic C of the element El is schematically represented and comprises points belonging to the first element El but no longer comprises points belonging to the second element E2 insofar as this element E2 has moved relative to the first element El between the acquisition of the first matrix M1 and the acquisition of the second matrix M2. It is understood here that a comparison of the first window fl with the second window f2 of neighboring points, when these neighboring points are considered in a two-dimensional coordinate space, will not make it possible to determine that the characteristics C associated with the windows fl and f2 are in reality corresponding characteristics C of the element El since the points of the windows of neighboring points compared are relatively different.However, the method 100 of the present disclosure makes it possible to avoid this type of situation insofar as the second windows of neighboring points compared during operation 160 should no longer or almost no longer include points belonging to an element other than the element comprising the characteristic to be followed, insofar as depth information is used to select the points of the second depth windows.

[0056] Other operations may optionally be integrated into the method 100 and are presented in the remainder of this disclosure. These operations may be integrated into the method 100 in combination with each other except when the contrary is expressly explained in this disclosure.

[0057] In examples, prior to the comparison operation 160, the method 100 may further comprise an operation 155 of determining, for each of the second windows of neighboring points compared during the operation 160, a descriptor corresponding to a characteristic value of the second window of points. neighbors. An example of calculating a descriptor for a window of neighboring points considered is notably described in the document written by CALONDER et al. BRIEF: Binary Robust Independent Elementary Features. The BRIEF document presents in particular a method known under the English name "Binary Robust Independent Elementary Features method" (BRIEF method), which will be referred to in this document as "method of binary robust and independent elementary characteristics".

[0058] In examples, a descriptor of a second neighboring point window may be determined from the two-dimensional coordinates and intensity values ​​of the points of the second neighboring point window in the intensity image.

[0059] In examples in which a descriptor is determined for the neighboring point windows compared during operation 160, operation 170 of association between a characteristic linked to a second neighboring point window of the first point matrix and a corresponding characteristic linked to a second neighboring point window of the second point matrix may be performed from a distance between the descriptor associated with the second neighboring point window of the first matrix and the descriptor associated with the second neighboring point window of the second matrix; for example when a distance between the descriptor associated with the second neighboring point window of the first matrix and the descriptor associated with the second neighboring point window of the second matrix is ​​less than a predetermined distance threshold. The distance between the descriptors in question here may for example correspond to a Haming distance.

[0060] In these examples, the operation 155 of determining a descriptor of a window of neighboring points and the operation 170 of associating a characteristic linked to a window of neighboring points of the first matrix of points and a corresponding characteristic linked to a window of neighboring points of the second matrix of points from a distance between the descriptors of these windows can be implemented from a “binary robust and independent elementary features method” (i.e. Binary Robust Independent Elementary Features method) or from a Lucas-Kanade method.

[0061] In examples, the method 100 may comprise an operation 180 of determining a displacement of the characteristic belonging to the element between the first instant of time and the second instant of time, from the two windows of neighboring points linked to the associated characteristics. In particular, by comparing the position of the second window of neighboring points on the second matrix of points with respect to its position on the first matrix of points and by knowing the instant of time associated with each of the first and second matrices, it is possible to determine a displacement of the window of neighboring points between the first and the second point matrix, which corresponds to a displacement of the characteristic between the first instant of time and the second instant of time.

[0062] In examples, the operation 180 of determining a displacement of the characteristic belonging to the element of the environment of the lidar sensor between the first instant of time and the second instant of time comprises: - a determination of a two-dimensional displacement of the characteristic between the first instant of time and the second instant of time, in the intensity image, from the two-dimensional coordinates and the intensity values ​​of the points of the two second windows of points associated with the characteristic in the intensity image; and - a determination of a three-dimensional displacement of the characteristic between the first instant of time and the second instant of time, from the two-dimensional displacement in the intensity image, and from the two-dimensional coordinates and the depth values ​​of the points of the two second windows of points associated with the characteristic in the depth image.

[0063] In examples including: - the operations of determining 155 a descriptor and of associating 170 from a distance between the descriptors implemented from a Lucas-Kanade method; and - the operation 160 of determining a displacement of the characteristic belonging to the element between the first instant of time and the second instant of time, from the three-dimensional coordinates of the two windows of neighboring points linked to the corresponding associated characteristics, the displacement of the feature of the element between the first instant of time and the second instant of time can be determined from a minimization of the distance between the descriptors of the two windows of neighboring points linked to the corresponding associated features.

[0064] These examples make it possible to determine a displacement of the characteristic between the two matrices at a scale smaller than that of a point of the point matrices obtained from the acquisitions of the lidar sensor 10, so that the determined displacement is obtained extremely precisely.

[0065] The method 100 according to the present disclosure therefore makes it possible, from information acquired from a lidar sensor 10, to track a characteristic of an element of the environment of the lidar sensor between two point matrices, and possibly to determine a displacement of this characteristic between the two matrices. The method 100 is presented for two point matrices but can be implemented on more than two point matrices so as to track a characteristic of the element of the environment of the lidar sensor 10 over time, and possibly to determine the displacement of this characteristic over time in an extremely precise manner.

Claims

1. Claims Method (100), implemented by a computer (11), for processing data acquired by a lidar sensor (10), the method (100) comprising: - obtaining (110) a first and a second matrix of points from acquisitions by a lidar sensor, the points of the first and second matrices respectively representing the environment of the lidar sensor (10) at a first and a second instant of time; each point of the matrix of points being associated with coordinates in a three-dimensional space and with an intensity value; - a projection (120) onto a two-dimensional projection plane of the point matrices so as to obtain, for each of the matrices, an intensity image and a depth image; for each of the first and second point matrices: * a determination (130) of at least one characteristic belonging to an element of the environment of the lidar sensor (10) on the point matrix, from the three-dimensional coordinates and the intensity values ​​of the points of the point matrix, a characteristic thus being associated with a point of the matrix; then for at least one characteristic determined on each matrix: * a determination (140), in the intensity image, of a first window of neighboring points linked to the characteristic and comprising the point associated with the characteristic, from the two-dimensional coordinates of the points of the matrix and their intensity value in the intensity image; * a determination (150) of a second window of neighboring points linked to the characteristic, the neighboring points of the second window of neighboring points corresponding to the neighboring points of the first neighboring window which have a distance with the point representing the characteristic less than a predetermined threshold in the three-dimensional space or in the depth image; then - a comparison (160) of a second window of neighboring points of the first matrix of points to a second window of neighboring points of the second matrix of points; then - an association (170) between a characteristic linked to a second window of neighboring points of the first matrix of points and a character- corresponding characteristic linked to a second window of neighboring points of the second matrix of points from the comparison.

2. Method according to the preceding claim, further comprising: - a determination (180) of a displacement of the characteristic belonging to the element of the environment of the lidar sensor (10) between the first instant of time and the second instant of time, from the two second windows of neighboring points linked to the corresponding associated characteristics.

3. Method according to the preceding claim, in which the determination of a displacement of the characteristic belonging to the element of the environment of the lidar sensor (10) between the first instant of time and the second instant of time comprises: - a determination of a two-dimensional displacement of the characteristic between the first instant of time and the second instant of time, in the intensity image, from the two-dimensional coordinates and the intensity values ​​of the points of the two second windows of points associated with the characteristic in the intensity image;and - a determination of a three-dimensional displacement of the characteristic between the first instant of time and the second instant of time, from the two-dimensional displacement in the intensity image, and from the two-dimensional coordinates and depth values ​​of the points of the two second windows of points associated with the characteristic in the depth image.;

4. Method according to any one of the preceding claims, further comprising, for a second window of neighboring points linked to a characteristic, a determination (155) of a descriptor of the second window of neighboring points, a descriptor corresponding to a characteristic value of the window of neighboring points determined from the coordinates and the intensity values ​​of the points of the second window of points in the intensity image; and in which the association between a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points is carried out from a distance between the descriptor associated with the second window of neighboring points of the first matrix and the descriptor associated with the second window of neighboring points of the second matrix.

5. Method according to the preceding claim, in which the determination of a descriptor of a second window of neighboring points and the association between a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points from a distance between the descriptors of these windows are implemented from a method of robust and independent binary elementary characteristics.

6. Method according to claim 4 combined with claim 2, wherein the determination of a descriptor of a second window of neighboring points and the association between a characteristic linked to a second window of neighboring points of the first matrix of points and a corresponding characteristic linked to a second window of neighboring points of the second matrix of points are implemented from a Lucas-Kanade method; and wherein the displacement of the characteristic of the element between the first instant of time and the second instant of time is determined from a minimization of the distance between the descriptors of the two windows of neighboring points linked to the corresponding associated characteristics.

7. Computer program product comprising instructions for implementing any one of the methods according to claims 1 to 6 when this program is executed by a processor.

8. A non-transitory computer-readable recording medium having recorded thereon a program for carrying out any one of the methods according to claims 1 to 6.

9. Computer (11) configured to implement a method according to any one of claims 1 to 6.

10. Motor vehicle (2) comprising a computer (11) according to the preceding claim.

Images