Method for estimating the position of a motor vehicle on a road

Abstract

The invention relates to a method for estimating the position of a motor vehicle (1) on a road (5) comprising at least one traffic lane (6) delimited by a first line (7) positioned in a first direction and a second line (8) positioned in a second direction opposite the first direction, by means of at least one camera of the motor vehicle (1). The method comprises a step of estimating the number of lanes, which comprises a plurality of sub-steps, and a step of estimating the number of the lane under consideration (N°V) by determining a main direction chosen between the first direction and the second direction, the number of the lane located at the edge of the road (5) in the main direction being equal to 1, and the number of the lane located at the edge of the road (5) in the direction opposite to the main direction being equal to the number of lanes.

Description

Description Title of the invention: METHOD FOR ESTIMATING THE POSITION OF A MOTOR VEHICLE ON A ROAD technical field

[0001] The present invention relates generally to the adaptive speed control of a motor vehicle and in particular to the estimation of the position of said motor vehicle on the road.

[0002] Adaptive speed control is a function that allows the vehicle to anticipate changes in speed in certain situations.

[0003] In town for example, a roundabout on a road segment limited to 50 kilometers per hour cannot actually be crossed at 50 kilometers per hour by the motor vehicle but at a lower speed, determined according to several parameters.

[0004] Another example is when a motor vehicle enters a motorway exit lane. Adaptive speed control is activated in this situation because the motorway exit lane is detected and the vehicle's exit signal is triggered. Since the vehicle is leaving the motorway, it is therefore important that its speed be reduced.

[0005] A similar situation occurs when a vehicle moves into the rightmost lane of the road without using the highway exit lane. However, reducing the vehicle's speed can be dangerous in this case. In another similar situation, the vehicle may also be overtaking another vehicle and then attempt to move back into the rightmost lane of the road.

[0006] In such a situation, activating the vehicle's adaptive cruise control can cause panic if confusion arises regarding the vehicle's exact location. Therefore, adaptive cruise control can be dangerous. Previous techniques

[0007] To reduce risks, solutions now exist aimed at determining the exact position of the motor vehicle on the road and in particular at determining which lane the motor vehicle is in.

[0008] A first approach to determining the position of the motor vehicle on the road suggests the use of high-definition mapping and a global positioning system (GPS, "Global Positioning System").

[0009] A second approach involves the use of an SD memory card (“Secure Digital”), a global positioning system and a camera in the motor vehicle.

[0010] However, these solutions are not optimal because they require several vehicle components, sometimes used exclusively for this purpose, thus generating additional installation time, costs and maintenance operations. Description of the invention

[0011] The present invention aims to provide a solution that addresses the aforementioned problems.

[0012] The invention therefore relates to a method for estimating the position of a motor vehicle on a road comprising at least one lane, each lane being delimited by a first line and a second line located on either side of the lane, the method being carried out by means of at least one camera on the motor vehicle. The method comprises: - a step to estimate the number of lanes, comprising the following sub-steps: • Initializing the number of lanes index to 1, • for the lane in question corresponding to the lane in which the motor vehicle is located, detection of the nature of the first and second lines of said lane, the nature of each line that can be crossed or not crossed by the motor vehicle, a crossable line separating the lane in question from an adjacent lane delimited by said crossable line and another line, • when a crossing line is detected, the nature of the other line of the track adjacent to the track in question is detected, • Increment the lane number index by 1 for each traversable line detected,

[0013] the value of the lane count index corresponding to the estimated number of lanes, - a step of estimating the number of the road considered based on the estimated number of roads and at least one impassable line detected.

[0014] According to one embodiment, the first line of the road in question is the left line of the road in question and the second line of the road in question is the right line of the road in question.

[0015] Advantageously, the step of estimating the number of lanes includes a substep of estimating the distance between the first traversable line of the lane in question and the other line of the adjacent lane delimited by said first line and / or the distance between the second traversable line of the lane in question and the line of the adjacent lane delimited by said second line, the substep of incrementing the index of the number of lanes being carried out only if said distance is less than or equal to a predetermined threshold.

[0016] According to one embodiment, the number of the road in question is determined by analyzing the following scenarios: - The number of the road in question is equal to 1 if the first line of the road in question is impassable. - the number of the road in question is equal to 2 if the other line of the road adjacent to the road in question, delimited by the first line of the road in question, is impassable. - The number of the road in question is equal to the number of lanes if the second line of the road in question is impassable. - the number of the road in question is equal to the subtraction of the number of roads and 1 if the other line of the road adjacent to the road in question, delimited by the second line of the road in question, is impassable, - the number of the road in question is equal to 3 if none of the previous scenarios are verified.

[0017] Advantageously, the process includes a step of determining the leftmost lane of the road.

[0018] Advantageously, the process includes a step of determining the rightmost lane of the road.

[0019] According to one embodiment, the process is carried out using images from the front camera of the motor vehicle.

[0020] Advantageously, the process includes a step of robustifying the estimates in which detected strengthening signals lead to validation of the estimation action, invalidation of the estimation action, or delay in the validation or invalidation of the estimation action.

[0021] According to one embodiment, reinforcement signals include information relating to the motor vehicle and / or information relating to road signs.

[0022] The invention also relates to a motor vehicle comprising a computer capable of performing the process as defined above.

[0023] The invention also relates to a computer program comprising instructions which, when the program is executed by a computer in the motor vehicle, lead the latter to implement the process as defined above. Brief description of the drawings

[0024] Other objects, features and advantages of the invention will become apparent from the following description, given solely by way of non-limiting example and made with reference to the accompanying drawings in which:

[0025] - Fig. 1 represents an example of a motor vehicle;

[0026] - Fig. 2 is a schematic view of a vehicle on a four-lane road;

[0027] - Fig. 3 schematically represents an example of estimating the number of lanes in a road;

[0028] - Fig. 4 is a schematic view of a vehicle on a two-lane road;

[0029] - Fig. 5 is a schematic view of a vehicle on a three-lane road;

[0030] - Fig. 6 is a schematic view of a road including a merge lane;

[0031] - Fig. 7 is a schematic view of a road including an exit lane;

[0032] - Fig. 8 is a schematic view of a road including a no right turn sign;

[0033] - Figure 9 is a schematic view of a road including directional signs; and

[0034] - Fig. 10 is a schematic view of a road including road signs relating to the road topology. Detailed description

[0035] We refer to [Fig.1] which represents a motor vehicle 1 in an orthogonal coordinate system X, Y, Z. The X axis is parallel to the longitudinal direction of the vehicle 1 and oriented towards the rear of the vehicle 1 in forward motion, the Y axis is perpendicular to the X axis and oriented towards the left of the vehicle 1 viewed from the front, and the Z axis is perpendicular to the plane defined by the X and Y axes and oriented towards the roof of the vehicle 1. The direction of travel of the motor vehicle is thus oriented along an axis opposite to the X axis.

[0036] The motor vehicle 1 includes a camera 2 and a computer 3 capable of implementing a method for estimating the position of the motor vehicle on a road. The camera 2 collects data relating to the road on which the motor vehicle 1 is located and transmits it to the computer 3. In the implementation illustrated in [Fig. 1], the camera 2 participating in the implementation of the method is the front camera of the motor vehicle 1.

[0037] The motor vehicle 1 also includes a computer program 4 comprising instructions which, when the program is executed by the computer 3 of the motor vehicle 1, lead the latter to implement the method of estimating the position of the motor vehicle 1 on the road.

[0038] As illustrated in [Fig.2], route 5 on which vehicle 1 is located comprises four traffic lanes, one of which is considered lane 6 on which the motor vehicle is located, delimited by a first line 7 and a second line 8.

[0039] In the following description, the first line 7 is considered to correspond to the left line of lane 6 in which motor vehicle 1 is located and the second line 8 is considered to correspond to the right line of lane 6 in which motor vehicle 1 is located. The lane 6 in which motor vehicle 1 is located is called the lane considered 6.

[0040] Route 5 lines 7, 8 may include ground markings such as a single line, a dashed line or a double line, and / or boundary devices such as barriers, low walls or cones.

[0041] The process of estimating the position of the motor vehicle on a road includes a first step which is shown schematically in [Fig.3] and which consists of estimating the number of lanes NbV.

[0042] During a first substep, a channel number index NbV is initialized to 1.

[0043] During a second sub-step, the N types of lines 7 and 8 of the track in question 6 are detected. Lines 7 and 8 may be passable or impassable by motor vehicle 1. A passable line means that an adjacent lane to the lane in question exists. Indeed, a passable line separates the lane in question from an adjacent lane delimited by said passable line and another line. A passable line is generally discontinuous and an impassable line is generally continuous.

[0044] If the process detects at least one traversable line during the second substep, it studies the distance D between said traversable line and the other line of the track adjacent to the track in question 6, delimited by said traversable line. If the distance D is less than or equal to a predetermined threshold S, the process increments the track number index NbV by 1. Advantageously, the predetermined threshold S is less than or equal to 5 meters, and preferably equal to 3.5 meters.

[0045] The NbV lane estimation step is performed as a loop. If the type N of the line under study is traversable, the type N of the other line of a first lane adjacent to the lane under consideration is in turn detected. If said other line of the first adjacent lane is also traversable, the type N of the other line of a second lane adjacent to the first adjacent lane is detected. For each of the traversable lines detected, the process increments the NbV lane number index by 1.

[0046] The process thus studies each line until an impassable line is detected.

[0047] When the process detects an impassable line to the left and right of vehicle 1, the loop execution is stopped and the total number of lanes (NbV) is generated. This is because if an impassable line is detected, it means that there is no lane adjacent to the lane in question on the side of said impassable line.

[0048] If the process detects an impassable line of the considered track 6, it considers that there is no longer an adjacent track in the direction of the impassable line and stops the loop in said direction.

[0049] According to one embodiment, the NbV lane estimation step is completed when an impassable line has been detected on the left of the motor vehicle and when an impassable line has been detected on the right of the motor vehicle, or when the NbV lane number index is equal to 5. Indeed, the method considers that a route 5 includes a number of NbV lanes less than or equal to 5.

[0050] When the method is implemented on a road segment such as [Fig. 2], the lane number index is first incremented by 1. Then, the method identifies the left lane 7 and the right lane 8 as traversable. Indeed, camera 2 detects broken lines and concludes that these are traversable by motor vehicle 1.

[0051] The distance D between the left line 7 of the considered track 6 and another line 9 of an adjacent track 10 being less than 5 meters, the track number index NbV is incremented by 1.

[0052] The distance D between the straight line 8 of the track in question 6 and another line 11 of an adjacent track 12 being also less than 5 meters, the track number index NbV is incremented by 1.

[0053] The process then examines the nature N of the lines in the adjacent lanes. Line 9 of the adjacent lane 10, to the left of the lane in question 6, is a traversable line. The lane number index NbV is therefore incremented by 1 again. Line 11 of the adjacent lane 12, to the right of the lane in question 6, is an impassable line. Indeed, camera 2 detects a solid line. The process therefore stops the loop performed on the right side of road 5.

[0054] Furthermore, a line 13 of an adjacent lane 14 to the adjacent lane 10 to the left of the lane considered 6 being detected as an impassable line, the loop made on the left side of route 5 is also stopped.

[0055] The process therefore generates a final number of NbV channels equal to 4, that is to say the estimated number of channels.

[0056] The process of estimating the position of motor vehicle 1 on the road then includes a second step of estimating the number N°V of the lane considered 6. The [Fig.2] illustrates motor vehicle 1 positioned on lane number 3 of road 5, considering a direction going from left to right.

[0057] A road plan is established assuming that the lane number at the leftmost end of Route 5 is 1 and the lane number located at the far right end MRL of route 5 is equal to the number of NbV lanes generated by the lane number estimation step.

[0058] In addition, the number N°V of the lane considered 6 is determined by analyzing one or more scenarios, the aim being to find the scenario corresponding to the current situation of vehicle 1.

[0059] According to a first scenario, if the left line 7 of the lane in question 6 is detected as impassable, then the number N°V of the lane in question 6 is equal to 1. Indeed, if the line to the left of vehicle 1 is impassable, this means that there is no lane adjacent to the left of the lane in question 6. Vehicle 1 is therefore located in the leftmost lane MLL of route 5.

[0060] If the current situation of vehicle 1 does not correspond to the first scenario, a second scenario is analyzed. If the other lane 9 of the adjacent lane 10 on the left side of the lane in question 6 is impassable, the lane number (N°V) of the lane in question 6 is equal to 2. This scenario means that there is only one lane 10 adjacent to the lane in question 6 on the left side of said lane in question 6. Vehicle 1 is therefore located in the second lane of route 5.

[0061] If this second scenario is not verified, the process examines a third scenario in which the lane number N°V of the lane in question 6 is equal to the lane number NbV determined previously when the straight line 8 of the lane in question 6 is impassable. Indeed, if there is no lane adjacent to the right of the lane in question 6, this means that vehicle 1 is located in the rightmost MRL lane of route 5.

[0062] Furthermore, the process analyzes a fourth scenario if the third scenario does not correspond to the current situation of vehicle 1. According to the fourth scenario, the lane number (N°V) of the lane in question, 6, is equal to the subtraction of the lane number (NbV) and 1 if the other lane (11) of the adjacent lane (12) to the right of the lane in question, 6, is impassable. Indeed, if the adjacent lane (12) to the right of the lane in question, 6, is the rightmost MRL lane of Route 5, the lane number (N°V) of the lane in question, 6, is the second-to-last lane of Route 5 when considering a direction from left to right.

[0063] According to the implementation described, the process assumes that a route 5 cannot have more than five lanes. Therefore, if none of the previous scenarios are verified, this means that vehicle 1 is located in the third lane of route 5.

[0064] After finding the scenario corresponding to the situation of motor vehicle 1, the process generates a number N°V of the lane considered 6.

[0065] According to the situation shown in [Fig. 2], the first, second, and third scenarios are not verified. Indeed, the left-hand line 7 of the considered track 6, the other line 9 of the adjacent track 10 to the left of the track in question 6 and the right line 8 of the track in question 6 are passable.

[0066] The fourth scenario, however, is verified since the other line 11 of the adjacent lane 12 to the right of the lane in question 6 is impassable. The process therefore generates the lane number N°V of the lane in question 6, which is equal to the subtraction of the number of lanes NbV and 1, that is, 3. Vehicle 1 is located on the third lane of route 5.

[0067] The process for estimating the position of the motor vehicle on the road also includes an optional third step of determining the leftmost lane MLL of the road and the rightmost lane MRL of the road.

[0068] Indeed, the nearest detected crossing line to the left line 7 of the considered lane 6 delimits the leftmost MLL lane of Route 5. Therefore, when motor vehicle 1 is located in the leftmost MLL lane of Route 5, the V-number of the considered lane 6 is equal to 1. According to the example illustrated in [Fig. 2], the process detects the crossing line 13 of the adjacent lane 14 as the nearest crossing line to the left line 7 of the considered lane 6 and thus deduces that the leftmost MLL lane with V-number equal to 1 is the adjacent lane 14. As illustrated in [Fig. 4], the left line Since 7 of the road considered 6 is impassable, the number N°V of the road considered 6 is therefore equal to 1.

[0069] Furthermore, the impassable line detected closest to the right-hand line Lane 8 of lane 6 delimits the rightmost MRP lane of route 5. Thus, when motor vehicle 1 is located in the rightmost MRP lane of route 5, the lane number (N°V) of lane 6 is equal to the total number of lanes (NbV). According to the example shown in [Fig. 2], the process detects the impassable line 11 of adjacent lane 12 as the impassable line 11 closest to the straight line 8 of lane 6 and therefore deduces that the rightmost MRP lane with lane number (N°V) equal to the total number of lanes (NbV) is adjacent lane 12. As shown in [Fig. 5], since the straight line 8 of lane 6 is impassable, the lane number (N°V) of lane 6 is therefore equal to the number of lanes (NbV).

[0070] The advantage of determining whether vehicle 1 is located in the leftmost lane MEE or in the rightmost lane MRP allows, in particular, to understand the speed of surrounding vehicles, to optimize overtaking of surrounding vehicles or even to anticipate a possible departure from the road.

[0071] The information from camera 2 may include noise that leads to a misrepresentation of the nature of the lines for a specific duration, typically a few seconds. Therefore, in some cases In these situations, even if camera 2 correctly perceives the nature N of the lines, the number of lanes NbV and the number N°V of the lane considered 6 do not correspond with reality.

[0072] In the example illustrated in [Fig. 6], an merge lane 15 occurs on a two-lane road. Camera 2 identifies this merge lane 15 as a third lane. The process therefore generates a distorted lane count (NbV) and lane number (N°V) for the duration that the merge lane 15 exists.

[0073] In the example illustrated in [Fig. 7], an exit lane 16 appears on a two-lane road. Camera 2 identifies the exit lane 16 as a third lane, and the process in this case also generates a number of lanes NbV and a number N°V of the lane in question 6 that are distorted for the duration that the exit lane exists.

[0074] Camera 2 must not consider an merge lane 15 or an exit lane 16 as an additional lane. These are temporary roads that are not useful to vehicle 1, which is already on the road and intends to remain there.

[0075] In a vehicle 1 driving mode, the intersection between said vehicle 1 and on-ramp 15 or off-ramp 16 lasts only a few seconds. However, a few seconds are sufficient to transmit undesirable information to the adaptive speed control of vehicle 1 and consequently create dangerous situations.

[0076] The method for estimating the position of the motor vehicle on the road therefore includes a fourth step of robusting the previous estimates, in which reinforcement signals are detected by camera 2 and allow the validation, invalidation, or delay of the validation or invalidation of the actions for the lane count (NbV), the lane number (N°V) of the lane in question (6), the leftmost lane (MLL), and the rightmost lane (MRL). In other words, if the method validates the action, the lane count (NbV) and the lane number (N°V) of the lane in question (6) will be estimated again, and if the method invalidates the action, no estimate is recalculated. If the method delays the decision to validate or invalidate the action, no action is performed for a certain period before the method determines whether to validate or invalidate the action.

[0077] In one implementation mode, the detected reinforcement signals include information relating to the motor vehicle 1 such as the vehicle status or the activation of the turn signals and / or information relating to road signs such as the reading of direction signs and road markings.

[0078] Thus, in the estimation robustness step, when the process detects a lane change of vehicle 1, it confirms a change of route 5 and therefore validates the action of estimating the number of lanes NbV and the number N°V of the lane considered 6 so that route 5 on which vehicle 1 is located can be re-examined. For example, a lane change by vehicle 1 is detected by the activation of a turn signal and / or by crossing a line.

[0079] In the case of detection of an merge lane 15 as illustrated in [Fig. 6], the process invalidates or delays the validation or invalidation of the estimation action. For example, an merge lane 15 is detected by reading a yield sign and / or a no-left or no-right turn sign.

[0080] As illustrated in [Fig. 8], if camera 2 detects a no-right-turn sign 17, the process delays the decision regarding the validation or invalidation of the estimation action because it is uncertain whether a merging lane 15 is nearby. Conversely, if a yield sign is detected a few tens of meters further on, a merging lane 15 is present, in which case the process invalidates the estimation action. Delaying the decision regarding the validation or invalidation of the estimation action upon detection of a no-right-turn sign 17 allows for a reassessment a few tens of meters further on to determine whether a merging lane 15 is present or if the situation is different.

[0081] In the case of detection of an exit lane 16, as in [Fig. 7], the process invalidates the estimation action so that the estimates of the lane number NbV and the lane number N°V of the lane in question 6 are not re-estimated, which would risk the adaptive speed control of vehicle 1 being inaccurately activated. For example, an exit lane 16 is detected by reading a speed limit sign and / or a direction sign.

[0082] In the case of detection of direction signs as illustrated in [Fig.9], the process validates the action of estimating the number of lanes NbV and the number N°V of the lane in question 6. Indeed, direction signs 18 generally indicate a change in the topology of the road 5 and it is therefore important that the number of lanes NbV and the number N°V of the lane in question 6 be estimated again.

[0083] Similarly, in the case of detecting signs indicating a change in the topology of route 19 as illustrated in [Fig.10], the process validates the action of estimating the number of lanes NbV and the number N°V of the lane considered 6.

[0084] Of course, the invention remains within the scope of the invention even if the method has a different order or steps. Indeed, it is possible to design a step for estimating the number V of the lane in question, 6, considering a direction from right to left. Furthermore, camera 2 is capable of reading all signs and road markings in order to implement the step of robustening the method's estimates.

Claims

Demands

1. A method for estimating the position of a motor vehicle (1) on a road (5) comprising at least one traffic lane (6, 10, 12, 14), each lane (6, 10, 12, 14) being delimited by a first line (7, 8, 9, 13) and a second line (7, 8, 9, 11) located on either side of the lane (6, 10, 12, 14), the method being carried out by means of at least one camera (2) of the motor vehicle (1) and characterized in that it comprises: - a step for estimating the number of lanes (NbV) comprising the following sub-steps: • Initialization of a channel number index (NbV) to 1, • for the lane in question (6) corresponding to the lane in which the motor vehicle (1) is located, detection of the nature (N) of the first and second lines (7, 8) of said lane (6), the nature (N) of each line that can be crossed or not crossed by the motor vehicle (1), a crossable line separating the lane in question (6) from an adjacent lane delimited by said crossable line and another line, • when a traversable line is detected, the nature (N) of the other line of the track adjacent to the track in question is detected (6), • Incrementing the number of lanes index (NbV) by 1 for each traversable line detected, the value of the number of lanes index (NbV) corresponding to the estimated number of lanes, - a step of estimating the number (N°V) of the road considered (6) based on the estimated number of roads and at least one impassable line detected.

2. A method according to claim 1, wherein the step of estimating the number of lanes (NbV) comprises a substep of estimating the distance (D) between the first traversable line of the lane in question (6) and the other line of the adjacent delimited lane by said first line and / or the distance (D) between the second traversable line of the lane in question (6) and the other line of the adjacent lane delimited by said second line, the substep of incrementing the index of the number of lanes (NbV) being carried out only if said distance (D) is less than or equal to a predetermined threshold (S).

3. A method according to any one of claims 1 and 2, wherein the number of the track (N°V) considered is determined by analyzing the following scenarios: - the number (N°V) of the road in question (6) is equal to 1 if the first line of the road in question (6) is impassable, - the number (N°V) of the road in question (6) is equal to 2 if the other line of the road adjacent to the road in question (6) delimited by the first line of the road in question (6) is impassable, - the number (N°V) of the road in question (6) is equal to the number of roads (NbV) if the second line of the road in question is impassable, - the number (N°V) of the road in question (6) is equal to the subtraction of the number of roads (NbV) and 1 if the other line of the road adjacent to the road in question (6) delimited by the second line of the road in question (6) is impassable, - the number (N°V) of the road in question (6) is equal to 3 if none of the previous scenarios are verified.

4. A method according to any one of claims 1 to 3, comprising a step of determining the leftmost lane (MLL) of the road (5).

5. A method according to any one of claims 1 to 4, comprising a step of determining the rightmost lane (MRL) of the road (5).

6. A method according to any one of claims 1 to 5, carried out using images from the front camera of the motor vehicle (1).

7. A method according to any one of claims 1 to 6, comprising a step of robustifying the estimates (NbV, N°V) in which detected strengthening signals lead validation of the estimation action, invalidation of the estimation action, or delay in the validation or invalidation of the estimation action.

8. Method according to claim 7, wherein the reinforcement signals include information relating to the motor vehicle (1) and / or information relating to road signs.

9. Motor vehicle (1) comprising a computer (3) capable of performing the method according to any one of claims 1 to 8.

10. Computer program (4) comprising instructions which, when the program is executed by a computer (3) of the motor vehicle (1), cause the latter to implement the method according to any one of claims 1 to 8.

Images