WO2018069059A1

WO2018069059A1 - Device and method for locating a vehicle

Info

Publication number: WO2018069059A1
Application number: PCT/EP2017/074713
Authority: WO
Inventors: Kichun JO; Paulo Resende; Benazouz Bradai; Thomas Heitzmann; Alexandre GARNAULT
Original assignee: Valeo Schalter Und Sensoren Gmbh
Priority date: 2016-10-13
Filing date: 2017-09-28
Publication date: 2018-04-19
Also published as: FR3058515A1

Abstract

The invention relates to a location device designed to be installed in a vehicle, the location device comprising a sensor (11) and an electronic processing unit programmed for: - matching data (DT1), obtained by means of said sensor and representing a road environment of the vehicle, with data (D1) relating to the position of at least one significant element (ER1, ER2, ER4) located in said road environment and that is pre-recorded in a storage unit (13) of the electronic processing unit; - determining a reliability level (NF1) of said significant element as a function of the result of said matching operation; and - determining a position (POSV) of the vehicle using the result of said matching operation only if the reliability level (NF1) of the corresponding significant element is greater than a given reliability threshold (SF1). The invention further relates to an associated location method.

Description

Device and method for locating a vehicle

TECHNICAL FIELD TO WHICH THE INVENTION RELATES The present invention relates to a locating device adapted to be installed in a vehicle.

It relates more particularly to a locating device for determining a position of the vehicle along a network of traffic lanes.

The invention also relates to a localization method implemented in such a device.

It applies particularly interestingly in a motor vehicle.

ARRI TECHNOLOGICAL BACKGROUND

More and more motor vehicles are equipped with GPS and / or GMS locating device (according to the acronyms of the "Global Positioning System" and "Global System for Mobile Communication"), making it possible to determine a position of the vehicle along a network of taxiways.

It is also known, in order to determine such a position, to map:

- data, pre-recorded in a vehicle storage unit, relating to the positioning of remarkable elements, such as traffic lights or road signs, located on this network of traffic lanes, with

an image captured by a front camera equipping the vehicle, and then determining the position of the vehicle according to this mapping.

However, when one of the remarkable elements of the road environment of the vehicle has been displaced, for example during road works, said data representative of its positioning, stored in the storage unit of the locating device, can no longer accurately match the actual position of this remarkable item. The mapping mentioned above can then lead to an erroneous, or at least imprecise, determination of the position of the vehicle. Moreover, one of the remarkable elements located in the vehicle's road environment may be partially masked, for example by plants, bystanders, or other vehicles. This element lends itself poorly to such a matching, which it may degrade accuracy.

OBJECT OF THE INVENTION

In this context, the invention proposes a locating device adapted to be installed in a vehicle, the locating device comprising:

a sensor, and

an electronic processing unit comprising a storage unit in which data relating to the positioning of remarkable elements located on or near a network of traffic lanes are stored,

characterized in that the processing unit is programmed to:

E1) for at least one of said remarkable elements located in a road environment of the vehicle:

mapping the data relating to the positioning of this remarkable element, provided by the storage unit, with data representative of the road environment of the vehicle obtained by means of said sensor, and

determining a level of reliability of this remarkable element, as a function of the result of said mapping, and for

E2) determining a position of the vehicle using the result of said matching only if said reliability level is above a given reliability threshold.

Thanks to the invention, the position of the vehicle is determined based on those of the remarkable elements actually making it possible to perform such a matching in a reliable and accurate manner.

So, one of the remarkable elements:

for which the data stored in the storage unit of the location device would not be reliable, for example because this item has been moved, or

- for which the captured data do not allow a reliable mapping, for example because this element is partially masked, will not be taken into account in the vehicle location process.

The position of the vehicle is thus advantageously robustly determined with respect to any errors or defects in updating said cartographic data, as well as with regard to possible defects in the captured data caused for example by noise. or measurement errors.

The accuracy with which this position is determined is also improved.

These properties are particularly interesting when the vehicle is of the autonomous type, that is to say when it is adapted to perform maneuvers or movements without intervention of his driver, or at least with a reduced intervention of the latter. Indeed, during such an autonomous operation of the vehicle, it must be able to determine its position with high accuracy and reliability.

Other nonlimiting and advantageous features of the locating device according to the invention are the following:

said storage unit contains, for each of said remarkable elements, an indication of a category of remarkable elements to which this remarkable element belongs, and, for at least one of the categories of remarkable elements, descriptive, representative data. common characteristics specific to the remarkable elements belonging to this category;

the processing unit is programmed, in step E1), to carry out said mapping as a function, moreover, of the descriptive data associated with the category to which said remarkable element mapped corresponds;

- the processing unit is further programmed for, the mapping of the step E1) having been performed for several of said remarkable elements located in the road environment of the vehicle, determine that the category, to which belongs one at least one of said matched remarkable elements is reliable, as a function at least of the reliability level of this remarkable element, and for determining the position of the vehicle using, for each of said mappings, the corresponding result, only if the remarkable item associated belongs to one of the reliable categories remarkable elements;

the processing unit is programmed to determine the level of reliability of the remarkable element matched by testing to what extent the data relating to the positioning of this remarkable element, stored in the storage unit of the localization device, allow to report data representative of the road environment of the vehicle, obtained by means of said sensor;

the processing unit is programmed to test to what extent the data relating to the positioning of the remarkable element matched make it possible to account for said data representative of the road environment of the vehicle, taking account of a position previous vehicle and a model of vehicle movement;

the processing unit is programmed to determine a plurality of potential positions that could occupy the vehicle, distributed around an average potential position, this average potential position being determined according to the previous position of the vehicle and a speed of vehicle movement;

- The processing unit is programmed to carry out the matching of said data representative of the road environment of the vehicle, with the data relating to the positioning of said remarkable element, on the basis of said potential positions that could occupy the vehicle.

It can also be provided that the processing unit is programmed to perform said mapping, by means of the following steps, performed for each of said potential positions:

determining, as a function of this potential position and the data relating to the positioning of the remarkable element mapped, stored in the storage unit, an expected datum representative of a position occupied by said remarkable element with respect to the vehicle if the vehicle was positioned at this potential position,

comparing said expected data with said data representative of the road environment of the vehicle, and

determining, for said remarkable element, a level of likelihood associated with this potential position of the vehicle, as a function of the result of said comparison.

Other optional features of the localization device according to the invention are the following:

the processing unit is programmed to determine the level of reliability of said remarkable element matched as a function of the likelihood levels associated respectively with the potential positions of the vehicle, and which have been determined for this remarkable element;

the processing unit is programmed to determine, for each of said remarkable elements, the corresponding level of reliability, by summing the likelihood levels which are associated with the potential positions of the vehicle and which have been determined for this remarkable element;

the processing unit is programmed, in step E2, to determine the position of the vehicle as a function of the said potential positions, and for each of the said potential positions, using the corresponding level of likelihood only if the level of reliability of the remarkable element, for which the level of likelihood of this potential position has been determined, is greater than said reliability threshold.

the processing unit is programmed to determine, for each of said potential positions, a global level of likelihood equal to the sum of the set of likelihood levels associated with this potential position, which have been obtained respectively for each of the elements the level of reliability is higher than the corresponding reliability threshold, and to determine the position of the vehicle according to the overall likelihood levels respectively associated with each of said potential positions.

the processing unit is programmed to determine that the position of the vehicle is equal to that of said potential positions which, among said potential positions, has the largest overall level of likelihood;

the processing unit is programmed to determine that the position of the vehicle is equal to a weighted average of said potential positions each assigned a weighting coefficient which is determined according to the overall level of likelihood of this potential position;

the processing unit is furthermore programmed, in step E2), to determine an orientation of the vehicle according to said setting correspondence.

The invention also proposes a method for locating a vehicle, implemented in a locating device which equips the vehicle and which comprises:

a sensor adapted to capture data representative of a road environment of said vehicle, and

the method comprising the following steps, performed by the electronic processing unit:

mapping of the data relating to the positioning of this remarkable element, provided by the storage unit, with data representative of the road environment of the vehicle obtained by means of said sensor, and

determination of a level of reliability of this remarkable element, as a function of the result of said mapping,

E2) determining a position of the vehicle using the result of said matching only if said reliability level is greater than a given reliability threshold.

The optional features presented above in terms of device can also be applied to the method just described.

DETAILED DESCRIPTION OF AN EXEMPLARY EMBODIMENT The following description with reference to the accompanying drawings, given by way of non-limiting examples, will make it clear what the invention consists of and how it can be implemented.

In the accompanying drawings:

FIG. 1 schematically shows, seen from the side, a vehicle equipped with a locating device according to the teachings of the invention,

FIG. 2 schematically represents an image of the road environment of the vehicle of FIG. 1, captured by means of an image sensor of FIG. locating device that equips this vehicle,

FIGS. 3A and 3B each schematically represent an image, respectively of a first type of remarkable element and of a second type of remarkable element capable of being located in the road environment of the vehicle,

FIG. 4 schematically represents a localization method implemented in the location device of FIG. 1, and

- Figure 5 schematically shows a previous position of the vehicle of Figure 1, and a plurality of potential positions that could occupy the vehicle.

FIG. 1 schematically represents a vehicle 1, here a motor vehicle such as a car, a truck, or a bus, located on a taxiway 3. The vehicle 1 may be a vehicle whose movements are manually controlled by a driver, or an autonomous vehicle.

The vehicle 1 comprises a locating device 10 which comprises a sensor 11, making it possible to capture data relating to a road 2 environment of the vehicle 1, in particular data relating to the positions of elements of this environment with respect to the vehicle 1.

The sensor 1 1 can be achieved by means, in particular:

an image sensor, such as a front, rear or side camera,

- a radar,

- an ultrasonic detector,

- or a lidar (according to the acronym for "Light Detection and Ranging"), such as a laser rangefinder (or laser scanner).

The road environment 2 of the vehicle designates all the objects, terrains, roads, reliefs, plants, and constructions located near the vehicle 1, as well as those possibly separated from the vehicle 1 but visible or at least detectable from it such as a particularly high distant building.

The sensor 1 1 is adapted more precisely to capture data representative of a portion of the road environment of the vehicle located in a detection field of the sensor.

In the embodiment described below, the sensor 1 1 is produced by means of an image sensor, in this case a video camera. The detection field of the sensor 1 1 corresponds in this case to the field of view CH of the image sensor.

The image sensor 1 1 is here adapted to be disposed in the vehicle 1 so that its field of vision CH includes part of the road environment 2 facing the vehicle 1.

Optionally, the location device may comprise one or more additional sensors as described above. Thus, the locating device may for example comprise the above-mentioned image sensor 11 and a lidar.

The locating device 10 also comprises an electronic processing unit 12 including in particular a processor performing logical operations, for example a microprocessor, and a storage unit 13, made for example by means of a hard disk or a non-memory device. -volatile rewritable.

The electronic processing unit 12 is connected to the sensor 1 1, and is adapted to receive the data captured by this sensor January 1, or at least representative data thereof.

Map data, representative of a network of traffic lanes on which the vehicle 1 is traveling, are stored in the storage unit 13.

These cartographic data include data relating to the positioning of remarkable elements, detectable by means of said sensor, located on this network of traffic lanes or in the vicinity of this network of traffic lanes.

The term "remarkable element" designates an element that can be identified and effectively identified by means of the data captured by the sensor 11, such as for example a road sign element such as a traffic sign, a traffic light, including a traffic light, or a marking line on the ground. Such a remarkable element may also correspond to a barrier or a parapet lining a traffic lane, a building facade, a piece of street furniture, or a relief having a specific shape.

Here, said map data further comprises, for each of said remarkable elements, an indication of a category of elements remarkable to which belongs this remarkable element.

Such a category corresponds, for example, to the remarkable elements of the "tricolor traffic light" type, or of the "speed limit sign" type, or to the "right-turn traffic sign" type.

The storage unit further comprises, for at least one of these categories of remarkable elements, here for each of them, descriptive data, representative of common characteristics specific to the remarkable elements belonging to this category.

These descriptive data are representative of characteristics specific to this category, and common to the various remarkable elements that are part of it. They can be representative of a shape or texture common to these elements, or include a typical image of an element belonging to this category.

FIGS. 3A and 3B show two such images: one, IM1, representative of a remarkable element of the "tricolor" type (FIG. 3A), and the other, IM2, representative of a remarkable element of the "panel" type speed limitation indication ".

These descriptive data may also include data representative of the positioning, with respect to each other, of characteristic points, such as angular points, or points having a high brightness or contrast, of a remarkable element typical of this category.

FIG. 2 diagrammatically represents, for an exemplary road configuration, an IMG image of the road 2 environment of the vehicle 1 captured by the image sensor.

Several remarkable elements located in this road environment are visible in this IMG image, in particular an ER1 traffic light, an ER2 road sign, an ER3 road marking line, and an ER4 building facade.

According to a particularly remarkable characteristic of the locating device 1 0, the processing unit 12 is programmed to perform the following steps of the locating method shown schematically in FIG. 4:

E1) for at least one of said outstanding elements ER1, ERi, ERN located in the 2 road environment of vehicle 1:

- Mapping (operations MC1, MCi, MCN of Figure 4) the data relating to the positioning of this remarkable element ER1, ERi, ERN, provided by the storage unit 1 3, with data representative of the road environment 2 of the vehicle 1 obtained by means of said sensor 1 1, and

determining a reliability level NF1, NFi, NFN of this remarkable element ER1, ERi, ERN, as a function of the result of said mapping, and for

E2) determining a POSV position of the vehicle 1 using the result of said matching only if said reliability level NF1, NFi, NFN is greater than a given reliability threshold SF1, SFi, SFN.

With this arrangement, the POSV position of the vehicle 1 is determined based on that, or those of the outstanding elements to actually achieve such a matching reliably and accurately.

So, one of the remarkable elements:

for which the data stored in the storage unit of the location device are unreliable, for example because this item has been moved, or

- for which the captured data do not allow a reliable mapping, for example because this element is partially masked,

is not taken into account when determining the position of the vehicle. The accuracy and reliability with which the POSV position of the vehicle is determined are then advantageously improved.

It will be noted that the data relating to the positioning of the remarkable elements ER1, ERi, ERN mentioned above are present in the storage unit 13 before the execution of the steps E1) and E2), for example because they were recorded there before the commissioning of the location device 10.

Here, the processing unit 12 is programmed more precisely for, in step E1), performing said mapping, and determining the level of reliability NF1, NFi, NFN associated, for each of the remarkable elements ER1, ERi, ..., ERNs located in the road 2 of the vehicle 1.

The number of remarkable elements ER1, ERi, ERN which are located in the 2 road environment of the vehicle 1 and which are included in the detection field of the sensor, is denoted N. These remarkable elements are referenced here by the index i between 1 and N.

The locating method, the main steps of which are shown in FIG. 4, comprises the steps E1) and E2) mentioned above. This method is implemented in the locating device 10 described above.

Steps E1) and E2) are now described in more detail.

Step E1)

In the embodiment of the localization method described here, the step E1) comprises a processing step E10) of captured data DCAPT by the sensor 1 1, prior to the mapping E1 1) itself, as this is shown schematically in FIG.

The processing step E1 0) of the captured data DCAPT comprises:

a detection of the remarkable elements present in the detection field of the sensor 1 1, comprising an identification of the category to which each of the detected remarkable elements belongs, and

a determination, for each remarkable element ER1, ERi, ERN detected, of processed data DT1, DTi, DTN concisely representing this remarkable element.

It will be recalled that the detection field of the sensor corresponds here to the field of view CH of the image sensor 11, and that the captured data DCAPT are representative of an image of the road environment 2 of the vehicle captured by it.

The remarkable elements mentioned above are detected by analysis of this image, by means of a shape recognition algorithm, for example.

The processed data DT1, DTi, DTN representing one of these remarkable elements can comprise in particular:

a position of the remarkable element, in the captured image,

a position of the remarkable element with respect to the vehicle, deduced from the captured image,

- positions of characteristic points of this remarkable element (angular point, very bright point and / or contrast, point having particular properties vis-à-vis a change of scale of the image),

a part of the captured image (that is to say a sub-image), in which this remarkable element is visible.

When the locating device comprises several sensors (for example several cameras), it is possible, during the processing step E10), to

- process the data captured by each of these sensors, then

determining said processed data DT1, DTi, DTN by merging the results obtained during the processing of the data respectively from the different sensors.

The data provided by the storage unit 13 of the locating device 10 here comprises, for each remarkable element ER 1, ER 1, ERN situated in the road 2 of the vehicle 1, reference data D 1, Di, DN relating to this remarkable element.

The reference data D1, Di, DN relating to one of these remarkable elements include:

- the data relating to the positioning of this remarkable element, and here,

the descriptive data (which have been described above), corresponding to the category to which this remarkable element belongs (for example a characteristic image of a traffic light, if the remarkable element is a traffic light).

During the following step E1 1), for each remarkable element ER1, ERi, ERN located in the road 2 of the vehicle 1, the reference data D1, Di, DN relating to this element are mapped MC1, MCi, MCN, individually, with the processed data DT1, DTi, DTN corresponding to this element.

As indicated above, during the mapping step E1 1), a reliability level NF1, NFi, NFN is determined for each of the remarkable elements.

This level of reliability is determined here by testing to what extent the reference data D1, Di, DN relating to this remarkable element can account for the captured data. This test can be performed using a known method of statistical test comparison or matching between the captured data and the reference data.

The level of reliability is determined more precisely here by testing to what extent the data relating to the positioning of this remarkable element (included in said reference data D1, D1, DN) allow, given a previous POSP position of the vehicle. 1 (Figure 5) and a vehicle displacement model, report captured data DCAPT.

The method comprises for this purpose a determination of a plurality of potential positions POS1, POSj, ... POSM, which could occupy the vehicle 1 (Figure 5).

These potential positions POS1, POSj, ... POSM are distributed around an average potential position <POSj> which is determined according to the previous POSP position of the vehicle 1 and the vehicle speed vector 1.

The average potential position <POSj> of the vehicle is determined according to the vehicle displacement model, for example by integrating over time, from a time corresponding to the previous position POSP, to a subsequent time corresponding to this average position < POSj>, the velocity vector of the vehicle. The model of displacement of the vehicle can correspond for example to a kinematic model of displacement in which the vehicle is assimilated in a simplified manner to a two-wheeled vehicle, generally called "bicycle model", or to a kinematic model of displacement taking into account the geometry of the front end of the vehicle, such as the kinematic model called "Ackermann model".

The number of potential positions POS1, POSj, ... POSM thus determined is denoted M. These potential positions are referenced here by the index j, between 1 and M.

For each of the remarkable elements ER1, ERi, ERN, the associated mapping operation MC1, MCi,..., MCN is then performed by means of the following steps, executed for each of said potential positions POS1, POSj,. POSM:

a) determination, according to the said potential position and data relating to the positioning of the remarkable element stored in the storage unit 1 3, an expected data representative of a position that would occupy said remarkable element with respect to the vehicle if the vehicle was positioned at this level; potential position POS1, POSj, ... POSM,

b) comparing said expected data with said captured DCAPT data, and

c) determining, for said remarkable element ER1, ERi, ... ERN, a likelihood level NVij associated with this potential position POS1, POSj,... POSM of the vehicle, as a function of the result of said comparison.

The level of likelihood bearing the reference NVij is that associated with the potential position POSj, and is determined by comparing, for the remarkable element ERi, the data captured with the pre-recorded data in the storage unit 1 3.

The expected data determined during step a) can, as here, include expected perceptual data, of the same type as the processed data DT1, DTi, ... DTN relating to this remarkable element.

These expected perceptual data correspond to an estimate of the processed data that would be obtained if the vehicle was positioned at this potential position POS1, POSj, ... POSM.

During step b), the expected data determined in step a) are more particularly compared, here, with the processed data DT1, DTi, ..., DTN relating to the remarkable element considered.

When the processed data DT1, DTi,..., DTN comprise the positions of characteristic points of the remarkable element, this comparison is for example carried out in:

calculating, for each characteristic point, a difference between the position of this point as indicated by the data processed, and the expected position of this point, as indicated by the expected data mentioned above, then by

- summing the differences thus determined to obtain an overall difference between predictions and observation, for this remarkable element.

The comparison made in step b) can also be performed, more simply, by calculating a difference between:

- on the one hand, the position that would occupy the remarkable element considered, relative to the vehicle, if the vehicle were positioned at one of the potential positions, determined according to the data relating to the positioning of this remarkable element stored in the storage unit 13 of the device, and,

on the other hand, a position of this remarkable element with respect to the vehicle, deduced from the captured data.

Here, in step c), the likelihood level NVij, associated respectively with each of the potential positions of the vehicle, is determined so as to be even larger than the aforementioned global difference is small. Otherwise formulated, the level of likelihood is all the greater that the expected data, determined for this potential position, are close to the processed data from the captured data. By way of example, the level of likelihood NVij may for example be inversely proportional to the overall data gap mentioned above.

In the embodiment of the locating method described here, the reliability level NFi associated with the remarkable element ERi is determined as a function of the likelihood levels NVi1, NVij, NViM, associated respectively with POS1 potential positions POSI POSM of the vehicle, and which were determined during the MCi matching based on this remarkable element ERi.

This level of reliability NFi is determined more precisely here by summing the likelihood levels determined for this remarkable element ERj, for all POS1 POSI potential positions, posi POSM of the vehicle, according to the following formula F1:

NFi = Σj 5 * NVij (Fl).

As a variant, this level of reliability NFi could, for example, be equal to the level of likelihood NFij, which, among the levels of likelihood determined for this remarkable element ERi, is the largest.

Step E2)

In step E2), the processing unit 12 tests here, for each of the remarkable elements ER1, ERi,..., ERN which has been the subject of a mapping, if the level of reliability NF1, NFi , NFN of this remarkable element is greater than the reliability threshold SF1, ..., SFi, ..., SFN. When the reliability level NF1, NFi, NFN of this remarkable element is greater than the reliability threshold SF1, SFi, SFN corresponding, this remarkable element is considered reliable. Otherwise, it is considered unreliable.

Among the N remarkable elements that have been mapped, a number N '(less than or equal to N) of reliable remarkable elements, ERF1, ... ERFi', ... ERFN 'are thus selected . These remarkable reliable elements are referenced here by the index i ', between 1 and N'.

Here, the value of the reliability threshold to be used is determined directly by the category to which belongs the remarkable element that is the subject of this reliability test.

Thus, if the remarkable element is a first tricolor traffic light of the road environment of the vehicle, or a second tricolor traffic light of this environment, the value of the reliability threshold used for this element will be that associated with the category d 'remarkable elements' tricolor traffic light'.

The value of the reliability threshold to be used for a given category of remarkable elements can for example be determined during tests preliminary to the putting into service of the tracking device 10, by comparing the levels of reliability obtained in road situations for which it has been shown that a remarkable element of this category is reliable, and those obtained in other configurations for which the remarkable elements of this category are manifestly unreliable.

The value of the reliability threshold to be used for a given category of outstanding elements may also be determined according to values usually recommended for a statistical fit test (such as the reliability test performed here), for example those recommended for suitability test called "chi-square" (also called "chi-square", or "^2" ).

In alternative embodiments (not shown), instead of testing the reliability of said remarkable elements by considering them individually, isolated from each other (as described above), one could test their reliability by considering all the elements of the same class, located in the road environment of the vehicle.

For example, in this case, for each of the categories, of which at least one remarkable element is located in the vehicle's road environment, to determine whether this category is reliable, or unreliable, according to the reliability levels of the remarkable elements belonging to this category.

The processing unit may for example be programmed to determine that such a category is reliable provided that the reliability level of each of its elements is greater than the reliability threshold associated with this category.

As a variant, the processing unit could also be programmed to determine that such a category is reliable, provided that an average of the reliability levels of its remarkable elements is greater than the reliability threshold associated with this category.

In the above-mentioned embodiment, for which reliability levels are determined directly category-by-category, and not item-by-element, a category is considered reliable if a reliability level is higher than the corresponding reliability threshold.

In the context of these alternative embodiments, a remarkable element is finally considered reliable when it belongs to a reliable category of remarkable elements.

Whatever the embodiment considered, the POSV position of the vehicle 1 is then determined, according to the results of the matching carried out for the remarkable reliable elements ERF1, ... ERFi ', ... ERFN'.

For this purpose, the processing unit 12 here firstly determines, for each potential position POSj, a global level of likelihood NGVj, as a function of the likelihood levels NVi'j associated with this potential position POSj, and which have been obtained for each of the remarkable elements reliable ERFi '.

The overall likelihood level NGVj of the potential position POSj is, for example, determined in accordance with the formula F2 below, that is to say by summing, for all the reliable reliable elements, the likelihood levels NVi'j associated with this POSj potential position:

The processing unit 1 2 then determines the position POSV of the vehicle 1 as a function of the potential positions POS1, POSj, POSM thereof and the corresponding global likelihood levels NGV1, NGVj, ... NGVM. The POSV position of the vehicle 1 is determined more particularly by calculating a weighted average of said potential positions POS1, POSj, POSM, each assigned a weighting coefficient which is determined according to the overall level of likelihood NGV1, NGVj, ... NGVM of this potential position.

The weighting coefficients associated respectively with each of the potential positions of the vehicle may in particular be equal to each of the overall level of likelihood of this potential position. In this case, the position of the vehicle is determined in accordance with the following formula F3:

POSV = [Zj! J POSj x NGVj] / [Σj ₌ ? NGVj] (F3).

As a variant, instead of being determined by a weighted average calculation, the position POSV of the vehicle could for example be determined to be equal to that of the said potential positions POS1, POSj, POSM which (among these potential positions) presents the overall level likelihood the greatest.

Optionally, a final reliability level, associated with the POSV position of the vehicle 1 thus determined, is calculated in step E2), as a function of the reliability levels of the remarkable elements present in the road environment 2 of the vehicle 1.

The final level of reliability NFPOS is for example equal to the sum of the levels of reliability of the remarkable reliable elements:

NFPOS

NFi '(F4).

According to another optional feature, the step E2) may comprise the calculation of an uncertainty of determination, related to the precision with which the POSV position of the vehicle 1 is obtained. This uncertainty of determination may, for example, be equal to the standard deviation of all the potential positions of the vehicle, each of which is assigned a weighting coefficient equal, for example, to the overall level of likelihood of this potential position.

According to a particularly remarkable optional feature, the locating method may, as here, comprise a determination of an orientation of the vehicle 1 with respect to its road environment 2, in addition to the determination of its position POSV described above.

The data captured DCAPT by the sensor 1 1 are indeed sensitive to the orientation of the vehicle 1. Here, for example, the position of a remarkable element in the image captured by the image sensor clearly depends on the orientation of the vehicle 1 with respect to its road environment 2. This dependence of the data captured DCAPT with the orientation of the vehicle 1 allows a determination of this orientation.

For this purpose, it is possible, for example, to determine, in step E1), not only potential positions of the vehicle 1, but also potential directions OR 1, ... ORk, ORP thereof.

In a manner comparable to that described above, a level of likelihood NVijk is then calculated for each pair formed of one of the POSj positions and of one of the potential ORk orientations, during the mapping based on the remarkable element ERi.

The level of reliability of the remarkable element ERi can then be determined according to these likelihood levels, for example by summing these likelihood levels for all the positions and potential orientations of the vehicle.

The optional characteristics of the step E2) relating to the determination of the POSV position of the vehicle, which have been described above, can then also apply to the determination of its orientation.

In other embodiments, the reliability levels of the outstanding elements located in the road environment 2 of the vehicle 1 could be determined differently from what has been described above.

For example, in an alternative embodiment, it can be provided that the mapping of the captured data DCAPT, with the reference data provided by the storage unit 13, is performed by calculating the values of the correlation function of the captured image, with an image representative of the relevant element concerned, provided by the storage unit 13. The reliability level of this remarkable element can then be equal, for example, to the maximum value of this correlation function, while the shift between images, for which this correlation function is maximum, provides information on the position of the vehicle.

Claims

1. Locating device (10) adapted to be installed in a vehicle (1), the locating device comprising:

a sensor (1 1), and

an electronic processing unit (12) comprising a storage unit (13) in which data relating to the positioning of remarkable elements located on or near a network of traffic lanes are stored,

characterized in that the processing unit (12) is programmed for: E1) for at least one of said outstanding elements (ER1, ERi, ERN) located in a road environment (2) of the vehicle (1):

- Mapping (MC1, MCi, MCN) the data relating to the positioning of this remarkable element (ER1, ERi, ERN), provided by the storage unit (1 3), with data representative of the road environment ( 2) of the vehicle (1) obtained by means of said sensor (1 1), and

determining a reliability level (NF1, NFi, NFN) of this remarkable element (ER1, ERi, ERN), as a function of the result of said mapping, and for

E2) determining a position (POSV) of the vehicle (1) using the result of said matching only if said reliability level (NF1, NFi, NFN) is greater than a given reliability threshold (SF1, SFi, SFN) .

2. Locating device (1 0) according to claim 1, wherein the storage unit (13) contains:

for each of said remarkable elements (ER1, ERi, ERN), an indication of a category of remarkable elements to which this remarkable element belongs, and

for at least one of the categories of remarkable elements, descriptive data representative of common characteristics of the remarkable elements belonging to this category.

A locating device (10) according to claim 2, wherein the processing unit (12) is programmed to, in step E1), performing said mapping in further function of the descriptive data associated with the category to which belongs the remarkable element matched.

4. Location device (1 0) according to one of claims 2 and 3, wherein the processing unit (12) is further programmed for, the mapping of the step E1) having been performed for several said remarkable elements located in the road environment of the vehicle:

determining that the category, to which belongs at least one of said remarkable elements matched, is reliable, depending at least on the level of reliability of this remarkable element, and for

determining the position of the vehicle using, for each of said mappings, the corresponding result, only if the associated remarkable element belongs to one of the reliable categories of remarkable elements.

5. Locating device (10) according to one of claims 1 to 4, wherein the processing unit (12) is programmed to determine the level of reliability (NF1, NFi, NFN) of the remarkable element (ER1 , ERi, ERN) mapped by testing to what extent the data relating to the positioning of this remarkable element, stored in the storage unit (13) of the localization device, make it possible to account for the data representative of the road environment (2) of the vehicle (1), obtained by means of said sensor (1 1).

6. locating device (10) according to claim 5, wherein the processing unit (1 2) is programmed to test to what extent the data relating to the positioning of the remarkable element (ER1, ERi, ERN) set in correspondence, make it possible to account for said data representative of the road environment (2) of the vehicle (1), taking into account a previous position (POSP) of the vehicle (1) and a model of vehicle displacement ( 1).

The locating device (10) according to claim 6, wherein the processing unit (1 2) is programmed to:

determining a plurality of potential positions (POS1, POSj, POSM) that could be occupied by the vehicle (1), distributed around an average potential position (<POSj>), this average potential position (<POSj>) being determined according to position (POSP) of the vehicle (1) and a vehicle traveling speed (1), and for

performing the mapping of said data representative of the road environment (2) of the vehicle (1) with the data relating to the positioning of said remarkable element (ER1, ERi, ERN), on the basis of said potential positions that could be occupied by the vehicle (1).

8. Locating device (10) according to claim 7, wherein the processing unit (12) is programmed to perform said matching, by means of the following steps, performed for each of said potential positions (POS1, POSj, POSM ):

determining, as a function of this potential position (POSj) and the data relating to the positioning of the remarkable element (ERi) mapped, stored in the storage unit (1 3), an expected data representative of a position that would occupy said remarkable element (ERi) relative to the vehicle (1) if the vehicle (1) was positioned at this potential position (POSj),

comparing said expected data with said data representative of the road environment (2) of the vehicle (1), and

determining, for said remarkable element (ERi), a level of likelihood (NVy) associated with this potential position (POSj) of the vehicle (1), as a function of the result of said comparison.

The locating device (10) according to claim 8, wherein the processing unit (12) is programmed to determine the reliability level (NFi) of said remarkable element (ER1, ERi, ERN) mapped as a function of the likelihood levels (NV, NVY, NV _iM) respectively associated with the potential position (POS1, POSj, MOPP) of vehicle, and which have been determined for this remarkable element (ERi).

1 0. Locating device (10) according to one of claims 8 and 9, wherein the processing unit (12) is programmed for, in step E2), determine the position (POSV) of the vehicle (1 ) according to said potential positions (POS1, POSj, POSM), and for each of said potential positions, using the level of likelihood (NV-n, NV, NV _N ) corresponding only if the reliability level of the remarkable element , for which the level of likelihood of this potential position has been determined, is above said reliability threshold.

1 1. A locating device (10) according to claim 1 0, wherein the processing unit (12) is programmed for, the mapping of step E1) having been performed for a plurality of said remarkable elements located in the road environment of the vehicle :

determining, for each of said potential positions (POS1, POSj, POSM), a global level of likelihood (NGV _j ) equal to the sum of the set of likelihood levels (NV _{1j 5} NV _rj , NV _Nj ) associated with this potential position (POSj), which were respectively obtained for each of the remarkable elements (ERF1, ERFi ', ERFN') whose reliability level is higher than the corresponding reliability threshold, and for

determining the position (POSV) of the vehicle (1) as a function of the overall likelihood levels (NGV _j ) respectively associated with each of said potential positions (POS1, POSj, POSM).

The locating device (10) according to claim 11, wherein the processing unit (12) is programmed to determine that the position (POSV) of the vehicle (1) is equal to that of said potential positions (POS1, POSj,

POSM) which, among the said potential positions, presents the overall level of likelihood (NGV _1, NGV _j , NGV _M ).

The locating device (10) according to claim 11, wherein the processing unit (12) is programmed to determine that the position (POSV) of the vehicle (1) is equal to a weighted average of said potential positions. (POS1, POSj, POSM) each assigned a weighting coefficient which is determined according to the overall level of likelihood (NGV _j ) of this potential position (POSj).

14. Locating device (10) according to one of claims 1 to 13, wherein the processing unit (12) is further programmed for, in step E2), determine an orientation of the vehicle (1) in according to said mapping.

A method of locating (1 0) a vehicle (1), implemented in a locating device (10) which equips the vehicle (1) and which comprises:

a sensor (1 1) adapted to capture data representative of a road environment (2) of said vehicle (1), and

the method comprising the following steps, performed by the electronic processing unit (12):

E1) for at least one of said remarkable elements (ER1, ERi,

ERN) located in a road environment (2) of the vehicle (1):

- mapping (MC1, MCi, MCN) data relating to the positioning of this remarkable element (ER1, ERi, ERN), provided by the storage unit (1 3), with data representative of the road environment ( 2) of the vehicle (1) obtained by means of said sensor (1 1), and

determination of a level of reliability (NF1, NFi, NFN) of this remarkable element (ER1, ERi, ERN), as a function of the result of said mapping, and for

E2) determining a position (POSV) of the vehicle (1) using the result of said matching only if said reliability level (NF1, NFi, NFN) is greater than a reliability threshold (SF1, SFi, SFN ) given.