FR3096128A1 - 720 ° absolute inclinometer capable of operating in milli-gravity - Google Patents

Abstract

This presentation relates to a device for recognizing the attitude (7) of an object, characterized in that it comprises: a sphere (11) comprising a wall (13) capable of at least partially letting a luminous flux pass; a ball (17), movably mounted inside said sphere (11); an image acquisition device (19), arranged in said attitude recognition device (7) to acquire a two-dimensional image of the ball (17) and of the sphere (11); a luminous flux emitting device (21), arranged in said attitude recognition device (7) to emit a luminous flux towards said sphere (11); an image processing device (23), connected to the image acquisition device (19), driven by an image processing algorithm and designed to drive the device for emitting a light flux (21) , to recover said image acquired by said image acquisition device (19) and to determine from said image (29) a gravity vector () originating from the center (Os) of the sphere (11) and for end the center (OB) of the ball (17). Figure for the abstract: Fig. 2

Description

720° absolute inclinometer that can operate in milli-gravity

The present invention falls within the field of space exploration, and relates to an opto-mechanical sensor suitable for use in a space environment with low gravity, such as that on the surface of an asteroid or a small moon. .

More specifically, the present invention relates to a device for recognizing the attitude of an object, to an object comprising such a device, to a space probe comprising such a device, to a lander comprising such a device and to a method of attitude recognition of an object, implemented by such a device.

In the field of space exploration, the use of automatic exploration probes makes it possible to carry out missions which are not currently possible for manned missions.

These automatic exploration missions are carried out by space probes equipped with scientific instruments that are launched from Earth using rockets and sent to the celestial bodies to be explored to study them more closely.

Exploration space probes tend more and more often to embark one or more small daughter probes dropped by a mother probe towards the surface of the celestial object to be explored, when it arrives close to the latter.

In this context, the daughter probes are automatic space vehicles, also equipped with scientific instruments, which will therefore land on the surface of the celestial bodies to be explored, unlike the mother probes which, for their part, will remotely explore the largest possible surface of the celestial body. This approach makes it possible to supplement the global scientific data measured remotely by the parent probe with local scientific data, measured in situ at the surface by the daughter probe.

The daughter probes intended to land on the surface of celestial bodies are qualified by the generic term "landers" or by the more specific English terms "lander", "hopper" or "rover" depending on the mobility capabilities of those -this.

Thus, the probes designated by the English term “landers” are not provided with displacement capacities. They carry out scientific measurements at the place of their landing.

The probes designated by the English term "hoppers" are equipped with a device allowing them to perform jumps in order to be able to visit sites other than the initial landing site.

The probes designated by the English term "rovers" are generally equipped with wheels in order to be able to roll on the surface of the object to be explored in order to reach target sites defined by the ground teams according to their potential scientific interest and the risks involved in achieving them.

Depending on the profile of the mission, the lander is dropped at an altitude varying between several tens of meters and several kilometers above the surface of the celestial body to be explored. The lander falls towards the surface of the body to be explored, by gravity. Depending on the release altitude, the initial speed and the intensity of the gravity, the duration of the trip to the surface is between a few seconds and several hours.

Due to its impact speed and the low gravity on the surface of small bodies in the solar system, after contacting the body, the lander usually bounces several times on the surface, before settling in one place. stranger with an unfamiliar attitude. The distance traveled during rebounds can be significant.

For landers with mobility capabilities (probes designated in English by "hoppers" and "rovers"), it is necessary to precisely determine the attitude of immobilization on the surface of the body in order to make the right decisions allowing them to recover and deploy solar generators where appropriate. This step of determining the orientation of the lander in contact with the ground of the body to be explored is critical, because erroneous information on the attitude of the lander can have the consequences of its deterioration or even the loss of the mission (for example attempt to deploy the solar generators while the lander is oriented "upside down" and therefore the solar generators are under the lander, in contact with the surface of the celestial body). It must be possible to determine the attitude autonomously because communication with the Earth is not possible if, for example, the attitude of the lander orients its antenna(s) unfavorably to ensure communication.

Several devices are known from the prior art making it possible to identify the orientation of an object, such as accelerometers, gyroscopes or even a set of photodiodes.

Most of the "off-the-shelf" accelerometers are designed to be used in terrestrial gravity (for example the accelerometers integrated in the mobile communication terminals designated by the English term "smartphone"). Their use on the surface of small bodies in the solar system where gravity can be thousands of times weaker than on Earth is problematic because the output signal would be drowned in measurement noise.

The use of gyroscopes has several disadvantages. First of all, the cost of a spatialized gyroscope, i.e. one that can be used in a space environment (vacuum conditions, thermal cycling, radiation, etc.) is particularly high. Then, the gyroscope uses the initial conditions of the probe (attitude of the probe relative to the landing site) to deduce an angular orientation of the lander. The deduction of the angular orientation of the probe relative to the ground is done by integrating over time the values of angles measured between the first attitude, just before the release by the mother probe, and the final attitude when the lander is on the ground, i.e. after bouncing several times before it stabilizes. However, a restart of the on-board computer following an unexpected change of state in the electronics due to an ionizing event ("SEU" for "Single Event Upset" in English) or a crash of the gyroscope electronics following an event " SEU” during the descent phase or during lander bounces could lead to loss of information. Similarly, shocks and collisions of the lander with geological elements that may be present on the surface during bounces constitute non-linearities that are likely to distort the measurements. Finally, the relatively high power consumption of gyroscopes can be very penalizing for missions that are often very constrained from an energy point of view.

Finally, the solution using photodiodes consists of analyzing the output signal from photodiodes placed at several locations on the lander to deduce its attitude. However, depending on the relief of the site where the lander is immobilized, shadows or the proximity of geological structures are likely to mislead the decision algorithm. Moreover, the regolith which generally makes up the surface of the body to be explored can cover the diodes during rebounds, which can also distort the measurements.

The present invention aims to overcome the above drawbacks, and to do this proposes a device for recognizing the attitude of an object, remarkable in that it comprises:

• a sphere comprising a wall able to pass at least partially a luminous flux;
• a ball, movably mounted inside said sphere;
• an image acquisition device, arranged in said attitude recognition device to acquire a two-dimensional image of the ball and the sphere;
• a device for emitting a luminous flux, arranged in said attitude recognition device to emit a luminous flux in the direction of said sphere;
• an image processing device, connected to the image acquisition device, driven by an image processing algorithm and designed to drive the device for emitting a luminous flux, to recover said image acquired by said device image acquisition and to determine from said image a gravity vector having the center of the sphere as its origin and the center of the ball as its end.

Thus, by planning to identify the direction and the direction of the gravity vector at the point of fall of the lander, the software of the latter is enabled to deduce therefrom the attitude of the attitude recognition device and therefore the attitude of the lander and to act accordingly.

The present invention makes it possible to constitute a precise and robust solution compared to the solutions of the prior art and more economical than a solution resulting from the prior art offering similar performance.

Also, by planning to position the ball inside a sphere, the measurement of the gravity vector is obtained over 4 Pi steradians, i.e. over 720°, which allows the attitude recognition device according to the invention to be fully operational regardless of the attitude of the undercarriage inside which the attitude recognition device of the invention is intended to be mounted.

The device therefore performs an absolute measurement, independent of any initial state, robust to any transient failure due to radiation during the descent phase of the lander towards the surface and independent of any stress caused by more or less violent shocks during rebounds. and can operate even in very low gravity if the ball is allowed time to stabilize in the sphere.

This measurement can, moreover, be repeated at any time during the mission on the surface when the lander is immobile and the need to know the attitude of the lander arises. This measurement can therefore be made at any time without having to integrate the movements of the undercarriage during its movements (case of "hopper" and "rover") and therefore without consuming electrical energy during said movements, the device being able to be kept de-energized outside the measurement phases.

According to optional characteristics of the attitude recognition device according to the invention:

• the attitude recognition device of an object comprises a box inside which are mounted:
  • the image acquisition device,
  • the sphere, inside which the ball is mounted, downstream with respect to said image acquisition device,
  • the device for emitting a luminous flux, upstream with respect to said sphere;
• according to one embodiment, the image processing device is mounted inside the housing or is directly integrated into the image acquisition device;
• the device for emitting a luminous flux comprises a set of upstream light-emitting diodes;
• the set of upstream light-emitting diodes is mounted downstream with respect to the image acquisition device;
• according to one embodiment, the attitude recognition device comprises a diaphragm, which extends radially from an outer face of the wall of the sphere, said diaphragm containing a plane of the sphere substantially orthogonal to a longitudinal axis of the housing, said diaphragm defining, on the one hand, an upstream compartment of the casing, inside which are mounted the image acquisition device and the assembly of upstream light-emitting diodes and, on the other hand, a downstream compartment, the device for emission of a luminous flux further comprising a set of downstream light-emitting diodes, mounted in the downstream compartment of said housing;
• the set of downstream light-emitting diodes is mounted downstream of the diaphragm, on the diaphragm or close to the diaphragm;
• the ball has a different color from that of the internal walls of the case;
• in one embodiment, the ball is white in color and the internal walls of the housing are black in color;
• according to one arrangement, the image acquisition device, the device for emitting a light flux and the image processing device are spatialized.

The invention also relates to an object comprising an attitude recognition device according to the invention.

The invention also relates to a space probe comprising an attitude recognition device according to the invention.

The invention also relates to a lander, consisting of a space vehicle designed to land and/or to land and move on the surface of a celestial body to be explored, remarkable in that it includes a recognition device of attitude according to the invention.

The invention further relates to a method for recognizing the attitude of an object implemented by an attitude recognition device according to the invention, noteworthy in that it comprises the following steps aimed at:

• define a three-dimensional direct orthonormal coordinate system (O, X, Y, Z) having the center of the sphere as its origin;
• acquire an image on which the sphere and the ball are visible;
• rectifying said acquired image so as to correct the geometric distortions;
• identifying the ball in said acquired and rectified image;
• define in the image plane from the three-dimensional direct orthonormal reference (O, X, Y, Z) a two-dimensional direct orthonormal reference (O _I , X, Y) originating from the image of the center of the sphere;
• identifying in said acquired and rectified image the coordinates (X _b , Y _b ) of the center of the ball in said two-dimensional direct orthonormal frame (O _I , X, Y);
• calculate, from the two-dimensional coordinates (X _b , Y _b ) and according to parameters of the attitude recognition device, the two sets of three-dimensional coordinates (X _B1 , Y _B1 , Z _B1 ) and (X _B2 , Y _B2 , Z _B2 ) of the ball in the direct orthonormal frame (O, X, Y, Z), corresponding to the two possible physical positions of the ball in the sphere for the position (X _b , Y _b ) of the center of the ball in the image ;
• calculate the value of the apparent surface of the ball in the acquired image;
• comparing said value of the apparent surface of the ball with a value or with a set of calibration values previously obtained during a calibration phase of the attitude recognition device;
• deduce which of the two sets of coordinates (X _B1 , Y _B1 , Z _B1 ) or (X _B2 , Y _B2 , Z _B2 ) is the one that corresponds to the position of the ball in the sphere;
• deduce the gravity vector, whose origin is the center of the sphere and its extremity is given by the coordinates (X _B , Y _B , Z _B ) of the center (O _B ) of the ball in the direct orthonormal frame (O, X , Y, Z).

According to an optional characteristic of the method according to the invention, the step aimed at deducing which of the two sets of coordinates (X _B1 , Y _B1 , Z _B1 ) or (X _B2 , Y _B2 , Z _B2 ) is the one which corresponds to the position of the ball in the sphere comprises a step of image acquisition of the sphere and of the ball by the image acquisition device, said step of image acquisition of the ball and of the sphere comprising a ignition phase of all the downstream light-emitting diodes.

Other characteristics and advantages of the invention will appear on reading the detailed description which follows for the understanding of which reference will be made to the appended drawings in which:

illustrates the undercarriage according to the invention.

illustrates a first embodiment of the attitude recognition device of the invention.

represents the lander according to a first attitude.

shows the attitude recognition device of the lander according to the attitude given in figure 3.

represents the image acquired when the lander is in the attitude of figure 3.

represents the lander according to a second attitude.

shows the attitude recognition device of the lander according to the attitude given in figure 6.

represents the image acquired when the lander is in the attitude of figure 6.

represents the lander according to a third attitude.

shows the attitude recognition device of the lander according to the attitude given in figure 9.

represents the image acquired when the lander is in the attitude of figure 9.

represents a modeling of the attitude recognition device.

shows in transverse view the model of figure 12 with two possible positions of the ball for a given position in the image.

is a view similar to that of figure 13, the ball being positioned in the sphere at a different location from that of figure 12.

illustrates a second embodiment of the attitude recognition device of the invention.

gives an example of use of the attitude recognition device according to the second embodiment.

gives another example of use of the attitude recognition device according to the second embodiment.

In all of the figures, identical or similar references represent identical or similar components or sets of components.

In the description and in the claims, the terms “upstream” and “downstream” must be understood with respect to the attitude recognition device, the upstream being located on the left with reference to FIG. 2 and the downstream being located right with reference to Figure 2.

The lander ("lander" or "rover" or "hopper" in English), is a space exploration daughter probe released from a space exploration mother probe (not shown) with a view to landing on a celestial body at explore. The mother probe is most often placed in orbit around the celestial body to be explored and serves in particular as a communication relay between the lander and the Earth.

Reference is made to FIG. 1 illustrating a lander 1 according to the invention in contact with a ground 3 of a celestial body to be explored.

In all the figures, the lander 1 is a daughter probe, for example a space vehicle designed to move on the ground 3 of the celestial body to be explored. Such a lander is designated by the English term “rover”.

The lander 1 also designates the probes designated by the English term "landers", which are not provided with displacement capabilities, as well as the probes designated by the English term "hoppers", which are equipped with a device allowing them to make jumps in order to be able to visit sites other than the initial landing site.

Given the great distance between the Earth and the celestial bodies likely to be explored, the round trip time of radio transmissions makes it impractical to directly pilot the lander operations in remote mode from the Earth. Therefore, the lander is often designed to have a certain degree of autonomy in carrying out its activities and in making decisions. In the “rover” case, the movement of the undercarriage 1 on the ground 3 is ensured thanks to a set of wheels 5 and an on-board navigation algorithm allows it to move autonomously and in complete safety towards the target designated by teams on Earth.

According to the invention, the lander 1 comprises an attitude recognition device 7 of the lander 1 whose purpose is to allow the piloting system of the lander 1 to determine by itself its attitude relative to the ground 3 after it has stabilized on the surface of the celestial body to be explored or at any time during the mission when the piloting system wants to know the attitude of the lander, for example to quantify the steepness of a slope and therefore its passability (“rover” case).

For reasons of mechanical protection and thermal control, the attitude recognition device 7 is preferably mounted inside a casing 9 of the undercarriage 1.

Reference is made to FIG. 2 illustrating the attitude recognition device 7.

The attitude recognition device 7 includes a hollow sphere 11 . The sphere 11 comprises a wall 13 capable of allowing a light flux to pass at least partially inside its internal volume 15. For this purpose, the wall 13 is therefore not opaque and is preferably transparent. The wall 13 of the sphere has preferentially received an anti-reflection treatment. By way of example, the sphere has a radius of between 2 and 6 cm, preferably about 4 cm. The sphere is fixed relative to the attitude recognition device 7.

The sphere 11 comprises in its internal volume a ball 17, mobile inside the sphere. The movement of the ball in the sphere results from the acceleration undergone by the recognition device 7 when the lander 1 is subjected to a force, in particular its weight under the effect of gravity. In order to prevent it from damaging the sphere due to the vibrations and shocks to which the lander is subjected during the launch, cruise and landing phases, the ball 17 is relatively light, for example 10 grams. The ball 17 is preferably made of a non-electrostatic material so that the electrostatic forces do not hinder its movements inside the sphere 11.

The attitude recognition device 7 also comprises an image acquisition device 19, fixed with respect to the attitude recognition device 7.

In the context of the use of the attitude recognition device 7 for a space mission, the image acquisition device 19 is spatialized. “Spatialised” means the characteristic according to which the device is designed to be used in a space environment. For this purpose, when it is spatialized, the image acquisition device 19 is preferably specifically designed for use in a space environment and undergoes qualification tests making it possible to validate its resistance to the space environment. The image acquisition device 19 is thus qualified as “spatialized” when it has been demonstrated that it is compatible in particular with use in an environment with high levels of radiation, very low pressure (space vacuum), numerous and significant thermal variations (for example between -50° and +70°C), mechanical vibrations and significant shocks.

When it is spatialized, the image acquisition device 19 can be obtained by a digital spatialized camera, or by an analog spatialized camera associated with an image digitizer (“frame grabber” in English) also spatialized. The camera is associated with a spatialized (optical) lens whose characteristics (focal length, aperture, back focus, etc.) are adapted to the need for acquiring sharp images of the entire sphere and ball.

The image acquisition device 19 is arranged in the attitude recognition device 7 to acquire a two-dimensional image of the ball 17 and the sphere 11, as will be seen in the following description illustrating the lander 1 on the floor 3 of the body to explore in different attitudes. For this purpose, the distance between the image acquisition device 19 and the sphere 11 is preferably of the order of a few centimeters, so as to obtain a clear image of the ball, while optimizing the compactness of the recognition device attitude. As an indication, the image acquisition device 19 can be in the form of a parallelepiped of 2 cm x 2 cm x 4 cm.

In order to improve the precision of the measurements, a calibration phase of the image acquisition device 19 according to the state of the art is carried out prior to its operational use. The calibration makes it possible to define the intrinsic and extrinsic parameters of the image acquisition device 19 with a view to rectifying the geometric and alignment distortions of the images produced. This phase can possibly be omitted if the quality of the design and the realization of the image acquisition device allows it (very low geometric distortion, excellent tolerances and opto-mechanical alignments...).

In order to be able to acquire images under constant and controlled lighting conditions, the attitude recognition device 7 further comprises a device for emitting a luminous flux 21.

The device for emitting a luminous flux 21 is arranged in the attitude recognition device 7 so as to emit a luminous flux in the direction of the sphere 11. The luminous power of the device for emitting a luminous flux 21 is calculated and chosen so that it allows the image acquisition device 19 to acquire images of the sphere 11 and of the ball 17 which can be used by an image processing device described later in the description. An image is said to be “usable” when the luminosity of the image of the sphere 11 and of the ball 17 is such that the sphere 11 and the ball 17 are visible in the image that the image acquisition device 19 acquires and that the image is neither underexposed nor overexposed. In the remainder of the description, examples of embodiments of the device 21 for emitting a light flux are given, as well as examples of the arrangement of the device for emitting a luminous flux 21 in the device. attitude recognition 7.

According to the invention, the image acquisition device 19 is connected to an image processing device 23.

The image processing device 23 is driven by an image processing algorithm and is designed to drive the device for emitting a luminous flux 21, to drive the image acquisition device 19, to recover a image acquired by the image acquisition device 19, to rectify said image and to determine from the rectified image the gravity vector having the center O _S of the sphere 11 as its origin and the center O _B of the ball 17 as its end.

The assembly formed by the image acquisition device 19, the sphere 11, inside which the ball 17 is mounted, the device for emitting a luminous flux 21 can be mounted inside a box 25 of the attitude recognition device 7.

In this case, the sphere 11 is mounted in the housing 25 downstream with respect to the image acquisition device 19, and the device for emitting a luminous flux 21 is mounted upstream with respect to the sphere 11.

The device 21 for emitting a luminous flux preferably comprises a set of light-emitting diodes 27, called "upstream", because of their locations in the housing 25. The set of upstream light-emitting diodes 27 is however preferably mounted downstream with respect to the image acquisition device 19. The upstream light-emitting diodes 27 can for example be distributed in the housing so as to form a circle contained in a plane (XY) with reference to the direct trihedron (X, Y, Z) represented in Figure 2. Other light sources can also be considered instead of light-emitting diodes (for example incandescent bulbs, fluorescent sources, etc.), but their robustness, mass and energy efficiency are generally less well suited to a mission spatial.

The housing 25 adopts for example a parallelepiped shape. The ball 17 preferably has a color different from that of the internal walls of the housing 25. Preferably, the ball 17 has a relatively light color, such as white, preferably matte. The internal walls of the box 25 preferably have a dark color such as black, preferably matte, in order to make the acquired images of the sphere 11 and the ball 17 more usable. The ball and the walls can be painted if necessary. to get the required colors.

The image processing device 23 is preferably mounted inside the casing 25 which makes it possible to make the attitude recognition device 7 autonomous. As a variant, the image processing device 23 can be integrated into the image acquisition device 19. As a variant, the image processing device 23 can be physically remote from the attitude recognition device 7 in which case, it is for example integrated into the on-board computer of the lander 1.

The image processing device 23 consists of a choice of a microcontroller, a microprocessor, an array of in situ programmable gates designated by the acronym "FPGA" for "Field-Programmable Gate Array" or an integrated circuit specific to an application designated by the acronym “ASIC” for “Application-Specific Integrated Circuit” as well as the associated electronic components and, where applicable, the software necessary for its proper functioning, for controlling the device of emission of a luminous flux 21, the control of the image acquisition device 19 and the communications with the computer on board the lander 1.

In the context of the use of the attitude recognition device 7 for a space mission, the image processing device 23 is spatialized, in the same way as the image acquisition device 19 and the transmission device of a luminous flux 21.

In the example given in FIGS. 3 to 5, the lander 1 (in its "rover" version) is stabilized on the ground 3 upright, that is to say that the wheels 5 of the undercarriage 1 rest on the ground 3 which itself is flat (the gravity vector is perpendicular to the ground and directed towards the ground).

When the lander 1 has stabilized on the body to be explored, the ball 17 ends up stabilizing in contact with the internal face of the wall 13 of the sphere 11, under the effect of gravity, as visible in the figure 4 which represents the attitude recognition device 7 in a simplified manner, the device for emitting a luminous flux 21 and the image processing device 23 not being represented.

FIG. 5 illustrates the image 29 acquired by the image acquisition device 19 when the lander 1 is in the attitude represented in FIG. 3. It is noted that the ball 17 is at the bottom of the image 29, which reflects the fact that the undercarriage 1 has its back oriented upwards, that the ball is on the Y axis, which reflects the fact that the undercarriage is not inclined to its right or its left and finally that the ball is in contact with the circle representing the image of the sphere 11, which reflects the fact that the undercarriage 1 is not tilted forwards or backwards either. From the image we can deduce that the lander 1 is therefore correctly placed on its wheels and that it is flat.

We refer to figures 6 to 8 which give another example of the attitude of the lander 1, still in its "rover" version. The attitude of the undercarriage 1 with respect to the ground is such that the undercarriage 1 has landed on its back, that is to say it is resting on its upper face 30, the wheels 5 not being at the ground contact. Furthermore, the upper face 30 of the undercarriage 1 is not perpendicular to the gravity vector, either because the ground is locally sloping or because the undercarriage is not placed flat on the ground, for example because of the presence of stones.

FIG. 7 illustrates (in a simplified view, only the sphere 11, the ball 17 and the image acquisition device 19 being represented) the attitude recognition device 7 when the lander 1 is in the attitude represented in Figure 6.

FIG. 8 illustrates image 29 acquired by image acquisition device 19 when lander 1 is in the attitude shown in FIG. 6. Note that ball 17 is at the top of the image 29, which translates the fact that the lander 1 is placed on its back, that the ball is on the Y axis, which translates the fact that the lander is not inclined towards its right or its left, and finally that the ball is not in contact with the circle representing the sphere 11 which reflects the fact that the landing gear 1 is also inclined forwards or backwards.

Reference is now made to Figures 9 to 11 giving another example of the attitude of the lander 1, still in its “rover” version.

The attitude of Lander 1 relative to the ground is such that Lander 1 has landed on its left side, i.e. it is resting on its left side face.

FIG. 10 illustrates (in a simplified view, only the sphere 11, the ball 17 and the image acquisition device 19 being represented) the attitude recognition device 7 when the lander 1 is in the attitude represented in Figure 9.

FIG. 11 illustrates the image 29 acquired by the image acquisition device 19 when the lander 1 is in the attitude represented in FIG. 9. It is noted that the ball 17 is on the left of the image 29, which reflects the fact that the undercarriage 1 is tilted to its left, that the ball is on the X axis, which reflects the fact that the undercarriage is not also tilted forward or l rear, and finally that the ball is in contact with the circle representing the sphere 11, which reflects the fact that the left side of the undercarriage is perpendicular to the gravity vector.

As understood from FIGS. 3 to 11, the identification of the direction of the gravity vector can be obtained over 4 Pi steradians, that is to say over 720°, which therefore allows the attitude recognition device 7 of the invention to determine the direction of the gravity vector whatever the attitude of the lander 1 with respect to the ground 3.

Reference is made to FIGS. 12 and 13 in order to explain the steps of the method for recognizing the attitude of the lander 1 implemented by the attitude recognition device 7 according to the invention.

FIG. 12 illustrates a modeling of the attitude recognition system 7. After calibration of the image acquisition device 19, the latter can be modeled by a projection center O _L and an image plane 31 located at the focal distance f downstream from the center of projection O _L and upstream from the sphere 11.

The ray of light 33 connecting the center of projection O _L to the center of the sphere O _S is the main axis of the system. It crosses the image plane 31 at the main point O _I .

The ray of light 35 connecting the center of projection O _L to the center O _B of the ball 17 intersects the image plane 31 at a point P _B . The point P _B is therefore the point of the image corresponding to the center of the ball 17.

It is specified that the locus described by all of the possible positions of the center of the ball 17 when the latter is in contact with the sphere 11 is a "virtual" sphere 37 of center O _S identical to that of the sphere 11 and of radius R = (R _s - R _B ), with R _s radius of the sphere 11 and R _B radius of the ball 17.

All of the steps below aim to determine the direction of the gravity vector (or acceleration) from the image of the ball 17 acquired by the image acquisition device 19.

The first step of the method for recognizing the attitude of the lander 1 aims to define a three-dimensional direct orthonormal reference (O, X, Y, Z) having as its origin O the center O _s of the sphere 11 and having the director vector of the Z axis which points to the center of projection O _L .

In this coordinate system (O, X, Y, Z): the coordinates (X _L , Y _L , Z _L ) of the center of projection O _L are (0, 0, Z _L ), Z _L being the distance between O _L , the center of projection, and O _S , the center of the sphere 11. The image plane 31 is parallel to the plane (O, X, Y) and its Z coordinate is equal to (Z _L -f), f being the focal distance of the image acquisition device 19. The coordinates of the center O _B of the ball 17 are (X _B , Y _B , Z _B ) and vary according to the position of the ball in the sphere.

According to a second step of the method of the invention, an image 29 (shown in FIGS. 5, 8 and 11) is acquired, thanks to the image acquisition device 19, on which the sphere 11 and the ball 17 are visible. The image acquisition device 19 then communicates to the image processing device 21 the acquired two-dimensional image.

According to a third step of the method of the invention, the image processing device 21 carries out the rectification of the acquired image. This step aims to use the results of the calibration of the image acquisition device 19 to correct the geometric distortions of the image.

According to a fourth step of the method of the invention, the ball 17 is identified in the image 29 acquired and rectified. To do this, in the case where the ball 17 is white in color and the internal walls of the housing 25 are black in color, the image processing algorithm of the image processing device 23 can advantageously be designed to search for the group of brightest connected pixels in the image, corresponding to ball 17.

According to a fifth step of the method of the invention, the image processing algorithm of the image processing device 23 identifies the image coordinates (u _b , v _b ), in pixels and fractions of pixels, of the center of the ball.

To do this, the image processing algorithm of the image processing device 23 can advantageously be designed to calculate the coordinates of the barycenter of the pixels identified as belonging to the ball 17 in the fourth step.

According to a sixth step of the method of the invention, a two-dimensional direct orthonormal reference (O _I , X, Y) is associated with the acquired image 29 from the three-dimensional direct orthonormal reference (O, X, Y, Z). The marker (O _I , X, Y) is defined so as to form a direct orthonormal marker seen by the image acquisition device 19 whose center O _I of the marker (O _I , X, Y) corresponds to the center O of the reference (O, X, Y, Z), therefore at the center O _s of the sphere 11. The points O _L , O _I and O are therefore all three located on the main axis defined by the ray of light 33.

According to a seventh step of the method of the invention, the coordinates (X _b , Y _b ) of the center O _B of the ball 17 in the two-dimensional direct orthonormal frame (O _I , X, Y) from the image coordinates (u _b , v _b ), the physical dimension of the pixels of the camera and the image coordinates (u _o , v _o ) of the point O _I .

Reference is made to FIG. 13 which illustrates, on a longitudinal section of the model of the attitude recognition device 7, the fact that for a given position of the ball 17 in the image 29, in general two positions of the ball 17 in the sphere 11 are possible. Note that there is however a set of positions of ball 17 in image 29 for which only one position of ball 17 in sphere 11 is possible.

According to an eighth step of the method of the invention, the coordinates (X _{B1 , Y B1} , Z _B1 ) and (X _B2 , Y _B2 , Z _B2 ) of _the two possible positions of the ball corresponding to the coordinates (X _b , Y _b ) of the center O _B of the ball 17 in the two-dimensional direct orthonormal frame (O _I , X, Y).

The calculation of the coordinates (X _B1 , Y _B1 , Z _B1 ) is for example obtained by applying the following formulas:

The calculation of the coordinates (X _B2 , Y _B2 , Z _B2 ) is for example obtained by applying the following formulas:

f being the focal length of the image acquisition system 19,

Z _L the distance between O _S the center of the sphere 11 and the center of projection O _L of the image acquisition system 19 and

R = (R _S – R _B ), difference between the radius of the sphere 11 and the radius of the ball 17.

The method of the invention comprises a ninth and a tenth step according to which it is determined which of the two sets of three-dimensional coordinates (X _B1 , Y _B1 , Z _B1 ) and (X _B2 , Y _B2 , Z _B2 ) is the one which corresponds to the actual position of ball 17 in sphere 11.

To do this, according to the ninth step of the method of the invention, the processing algorithm calculates the value of the apparent surface of the ball 17 in the acquired image 29. According to the tenth step of the method of the invention, the processing algorithm compares the value of the apparent surface of the ball 17 with a value or a set of calibration values previously obtained during a calibration phase of the attitude recognition device 7.

The two possible three-dimensional positions of the ball 17 being on the ray of light 35 passing through the center of projection O _L and the point P _B , intersection between the ray of light 35 and the image plane 31, a position O _B1 is obtained for which the ball 17 is closer to the center of projection O _L and a position O _B2 for which the ball 17 is farther from O _L . Due to the perspective projection, when the ball 17 is closer to the center of projection O _L , the apparent surface of the ball 17 in the image 29 is greater than the apparent surface of the ball 17 in the image 29 than the is obtained when the ball 17 is further from the center of projection O _L .

Thus, the step according to which the value of the apparent surface of the ball 17 is compared to a value or to a set of calibration values previously obtained during the calibration phase of the attitude recognition device 7 allows to determine whether the ball 17 is in the position closer or further from the center of projection O _L and thus to determine which of the two sets of three-dimensional coordinates calculated in the eighth step is the correct one (tenth step of the method according to the invention) .

In the particular case mentioned previously where only one position of the ball corresponds to the coordinates (X _b , Y _b ) of the center O _B of the ball 17 in the two-dimensional direct orthonormal frame (O _I , X, Y), we have X _B1 = X _B2 , Y _B1 = Y _B2 and Z _B1 = Z _B2 and steps 9 and 10 can be omitted.

According to an eleventh step of the method of the invention, the algorithm deduces the gravity vector (or acceleration), whose origin is given by the center O _S of the sphere 11 and its end is given by the coordinates (X _B , Y _B , Z _B ) of the center O _B of the ball 17 in the direct orthonormal frame (O, X, Y, Z), determined at the end of the tenth step of the method or at the end of the eighth step of the method in the particular case where only a three-dimensional position of the ball 17 in the sphere 11 corresponds to the two-dimensional position of ball 17 in image 29.

Reference is now made to FIG. 14 which illustrates the fact that for certain attitudes of the attitude recognition device 7 the two positions of the ball 17 in the sphere 11 corresponding to the position of the ball 17 in the image 29 can be very close to each other, thus making it potentially difficult to choose between the two sets of possible three-dimensional coordinates by analyzing the apparent surface of the ball in the image. For most missions, this case can advantageously be handled by calculating the barycenter of the two positions. However, for missions for which it is desired to improve the precision of the measurements when a situation such as that illustrated in FIG. 14 arises, the attitude recognition device 7 can be obtained according to a second embodiment illustrated in FIG. 15 to which reference is now made, showing the attitude recognition device 7 in longitudinal section.

In this second embodiment, the attitude recognition device 7 further comprises a diaphragm 39, fixed with respect to the attitude recognition device 7.

The diaphragm 39 extends radially from an outer face of the wall 13 of the sphere 11. More specifically, the diaphragm 39 contains a plane (XY) of the sphere 11, which extends substantially orthogonal to a longitudinal axis 41 of the housing, so that the diaphragm 39 defines an upstream compartment 43 of the housing and a downstream compartment 45 of the housing 25.

The diaphragm is preferably mounted at a distance Z _D from the center O _S of the sphere such that:

Z _L being the distance between O _S, the center of the sphere 11, and O _L , the center of projection of the image acquisition device 19 and

R = (R _S – R _B ), difference between the radius of the sphere 11 and the radius of the ball 17.

Inside the upstream compartment 43 of the housing 25 are mounted the image acquisition device 19 and the assembly of upstream light-emitting diodes 27 of the device for emitting a luminous flux 21.

According to this second embodiment, the device 21 for emitting a luminous flux comprises a second set of light-emitting diodes, called downstream light-emitting diodes 47, mounted in the downstream compartment 45 of the housing 25. The set of downstream light-emitting diodes 47 is mounted close to the diaphragm 39 or on the diaphragm 39.

As for the set of upstream light-emitting diodes 27, the set of downstream light-emitting diodes 47 is arranged in the attitude recognition device 7 to emit a luminous flux in the direction of the sphere 11.

In order to determine which of the two sets of three-dimensional coordinates (X _B1 , Y _B1 , Z _B1 ) or (X _B2 , Y _B2 , Z _B2 ) is the one which corresponds to the real position of the ball 17 in the sphere 11, the method implemented by the attitude recognition device 7 obtained according to the second embodiment of the invention provides for an additional step comprising the extinction of all the upstream light-emitting diodes 27 and the lighting of the set of diodes downstream light-emitting diodes 47, mounted in the downstream compartment of the housing 25, then an image acquisition step of the sphere 11 and the ball 17, the lighting this time being provided solely by the downstream light-emitting diodes 47.

Thus, as shown in Figure 16, when the ball 17 is upstream of the diaphragm 39, the ball will appear illuminated from the front in the image acquired in upstream lighting and illuminated from behind in the image acquired in downstream lighting .

In the same way, as represented in FIG. 17, when the ball 17 is downstream of the diaphragm 39, the ball will appear illuminated from the front in the image acquired in upstream illumination and also illuminated from the front in the image acquired in downstream illumination.

Consequently, according to the appearance of the ball in the image as a function of the illumination from upstream or from downstream, the method according to the invention determines whether the ball 17 is located upstream or downstream of the diaphragm 39 and deduces which is the correct set of three-dimensional coordinates of the ball 17 in the direct orthonormal frame (O, X, Y, Z).

When the coordinates of the ball 17 are determined, the method of the invention deduces the gravity vector, whose origin is the center O _S of the sphere 11 and its end is given by the coordinates (X _B , Y _B , Z _B ) of the center of the ball in the direct orthonormal frame (O, X, Y, Z).

The switching on of the sets of upstream and downstream light-emitting diodes at the right time and the sequence of steps is controlled by the image processing algorithm.

It goes without saying that the present invention is not limited to the sole embodiments of this attitude recognition device, of this object comprising the attitude recognition device of the invention, of this lander and of this method of attitude recognition, described above only by way of illustrative examples, but it embraces on the contrary all the variants involving the technical equivalents of the means described as well as their combinations if these fall within the scope of the invention .

Thus, the present invention can be used on board any space vehicle needing to determine its attitude with respect to an acceleration vector provided that this vector does not vary rapidly over time, in order to allow time for the ball to stabilize. .

Similarly, the present invention can also function perfectly on Earth, in terrestrial gravity, in any application requiring an absolute attitude measurement over 720°, provided that the attitude does not vary too quickly over time in order to allow time the ball to stabilize. In this case, the various constituent elements of the invention obviously do not need to be spatialized. Thus, the attitude recognition device may be intended to be integrated into an object designed to operate in a terrestrial environment. Such an object can be provided with displacement capabilities, such as an aircraft, or devoid of its own displacement capabilities but of which one wishes for example to know the evolution of the attitude during its handling or its transport, such as for example a crate or a container.

Claims (15)

Attitude recognition device (7) of an object, characterized in that it comprises:
- a sphere (11) comprising a wall (13) suitable for at least partially passing a luminous flux;
- a ball (17), movably mounted inside said sphere (11);
- an image acquisition device (19), arranged in said attitude recognition device (7) to acquire a two-dimensional image (29) of the ball (17) and of the sphere (11);
- a device for emitting a luminous flux (21), arranged in said attitude recognition device (7) to emit a luminous flux in the direction of said sphere (11);
- an image processing device (23), connected to the image acquisition device (19), driven by an image processing algorithm and designed to drive the device for emitting a luminous flux (21 ), to recover said image (29) acquired by said image acquisition device (19) and to determine from said image (29) a gravity vector ( ) having as its origin the center (O _s ) of the sphere (11) and as its end the center (O _B ) of the ball (17). Attitude recognition device (7) of an object according to claim 1, characterized in that it comprises a casing (25) inside which are mounted:
- the image acquisition device (19),
- the sphere (11), inside which the ball (17) is mounted, downstream from said image acquisition device (19),
- the device for emitting a luminous flux (21), upstream with respect to said sphere (11). Attitude recognition device (7) of an object according to claim 2, characterized in that the image processing device (23) is mounted inside the housing (25) or is directly integrated into the device for acquisition of images (19). Attitude recognition device (7) of an object according to one of Claims 2 or 3, characterized in that the device for emitting a luminous flux (21) comprises a set of upstream light-emitting diodes (27) . Attitude recognition device (7) of an object according to claim 4, characterized in that the set of upstream light-emitting diodes (27) is mounted downstream with respect to the image acquisition device (19). Attitude recognition device (7) of an object according to one of Claims 4 or 5, characterized in that it comprises a diaphragm (39), which extends radially from an outer face of the wall (13 ) of the sphere (11), said diaphragm (39) containing a plane (XY) of the sphere (11) substantially orthogonal to a longitudinal axis (41) of the housing (25), said diaphragm (39) defining, on a hand, an upstream compartment (43) of the casing, inside which are mounted the image acquisition device (19) and the assembly of upstream light-emitting diodes (27) and, on the other hand, a downstream compartment (45), said attitude recognition device (7) being further characterized in that the device for emitting a luminous flux (21) comprises a set of downstream light-emitting diodes (47), mounted in the downstream compartment (45) of said housing (25). Attitude recognition device (7) of an object according to claim 6, characterized in that the assembly of downstream light-emitting diodes (47) is mounted downstream of the diaphragm, on the diaphragm (39) or close to the diaphragm (39). Attitude recognition device (7) of an object according to any one of Claims 2 to 7, characterized in that the ball (17) has a different color from that of the internal walls of the casing (25). Attitude recognition device (7) of an object according to claim 8, characterized in that the ball (17) is white in color and in that the internal walls of the casing (25) are black in color. Attitude recognition device (7) of an object according to any one of Claims 1 to 9, characterized in that the image acquisition device (19), the device for emitting a luminous flux (21) and the image processing device (23) are spatialized. Object comprising an attitude recognition device (7) according to any one of Claims 1 to 10. Space probe comprising an attitude recognition device (7) according to claim 10. Lander (1), consisting of a space vehicle designed to land and/or to land and move on the surface of a celestial body to be explored, characterized in that it comprises an attitude recognition device (7 ) according to claim 10. Method of attitude recognition of an object implemented by an attitude recognition device (7) according to any one of Claims 1 to 10, characterized in that it comprises the following steps aimed at:
- define a three-dimensional direct orthonormal reference (O, X, Y, Z) originating from the center (O _s ) of the sphere (11);
- acquiring an image (29) on which the sphere (11) and the ball (17) are visible;
- rectifying said acquired image (29) so as to correct the geometric distortions;
- identifying the ball (17) in said acquired and rectified image (29);
- define in the image plane (31) from the three-dimensional direct orthonormal reference (O, X, Y, Z) a two-dimensional direct orthonormal reference (O _I , X, Y) having as its origin (O _I ) the image of the center (O _s ) of the sphere (11);
- identifying in said acquired and rectified image (29) the coordinates (X _b , Y _b ) of the center (O _B ) of the ball (17) in said two-dimensional direct orthonormal frame (O _I , X, Y);
- calculate, from the two-dimensional coordinates (X _b , Y _b ) and according to parameters of the attitude recognition device, the two sets of three-dimensional coordinates (X _B1 , Y _B1 , Z _B1 ) and (X _B2 , Y _B2 , Z _B2 ) of the ball (17) in the direct orthonormal frame (O, X, Y, Z), corresponding to the two physical positions of the ball (17) in the sphere (11) possible for the position (X _b , Y _b ) of the center of the ball (17) in the image (29);
- calculating the value of the apparent surface of the ball (17) in the image (29) acquired;
- comparing said value of the apparent surface of the ball (17) with a value or with a set of calibration values previously obtained during a calibration phase of the attitude recognition device (7);
- deducing which of the two sets of coordinates (X _B1 , Y _B1 , Z _B1 ) or (X _B2 , Y _B2 , Z _B2 ) is the one which corresponds to the position of the ball (17) in the sphere (11);
- deduce the gravity vector ( ), whose origin is the center (O _S ) of the sphere (11) and its end is given by the coordinates (X _B , Y _B , Z _B ) of the center (O _B ) of the ball (17) in the direct orthonormal reference (O, X, Y, Z). Method for recognizing the attitude of an object according to Claim 14, implemented by an attitude recognition device (7) according to one of Claims 6 or 7, characterized in that the step aimed at deducing which of the two sets of coordinates (X _B1 , Y _B1 , Z _B1 ) or (X _B2 , Y _B2 , Z _B2 ) is the one which corresponds to the position of the ball (17) in the sphere (11) comprises a step of image acquisition of the sphere (11) and the ball (17) by the image acquisition device (19), said step of acquiring an image of the ball (17) and the sphere (11 ) comprising an ignition phase of all the downstream light-emitting diodes (47).