Method for detecting an obstacle in front of a vehicle

Abstract

Method for detecting an obstacle in front of a vehicle (10) using a radar sensor (16), wherein the radar sensor (16) monitors a monitoring area in front of the vehicle (10), wherein a classification of an object (24, 25) detected by the radar sensor in the monitoring area is carried out, wherein the monitoring area is divided into monitoring sub-areas (20, 21) and a first classification rule is used for classification in a first monitoring sub-area (20) and a second classification rule is used for classification in a second monitoring sub-area (21), characterized in that the first monitoring sub-area (20) extends between the radar sensor (16) and a ground intersection line (22) and the second monitoring sub-area (21) extends between the ground intersection line (22) and a maximum monitoring distance (23), wherein the ground intersection line (22) is a line,which is defined by the intersection of a radar beam of the radar sensor (16) with the ground, wherein the first classification rule is independent of the intrinsic velocity of the object (25) detected in the monitoring area, and the second classification rule comprises classifying the object (24) detected in the monitoring area as a function of the intrinsic velocity of the object (24) detected in the monitoring area.

Description

State of the art

[0001] Methods for detecting an obstacle in front of a vehicle using a radar sensor are known from automotive applications. For obstacle classification, feature spaces are used in combination with classification methods. Features can include performance values, variance curves, etc. These methods are also based on the vehicle's longitudinal approach to the object. State-of-the-art classification algorithms are highly dependent on the installation position of a radar sensor, as the feature space is non-deterministically dependent on installation parameters such as the installation position or obscuration by a vehicle emblem. Therefore, a time-consuming modification and validation of the classification algorithm is necessary for each vehicle model.

[0002] The methods known from the prior art therefore have the disadvantage that using a radar sensor to detect an obstacle in front of a vehicle is not possible without complex application. The methods known from the prior art also have the disadvantage that they assume a longitudinal approach of the vehicle to the detected object, so that object detection of objects to the side of the vehicle's direction of movement is impossible or difficult. For these reasons, the methods known from the prior art are not suitable for detecting obstacles in front of an off-highway vehicle using a radar sensor.

[0003] The document DE 102 55 797 A1 discloses a method for detecting the surroundings ahead of a road vehicle by means of an environment detection system.

[0004] The document DE 10 2010 007 415 A1 discloses a radar signal processing method for estimating an object height of an object above a reflecting surface.

[0005] The document DE 10 2009 041 555 A1 discloses a method for warning a vehicle driver of an object by monitoring an area surrounding a vehicle.

[0006] The document US 2011 / 0 025 548 A1 discloses a method for tracking targets detected by sensors in a host vehicle.

[0007] The document US 2017 / 0 309 997 A1 discloses a radar detection system for a vehicle with a transmitter and a receiver configured to receive the transmitted radio signal reflected by objects in the environment.

[0008] The document DE 10 2006 020 387 A1 discloses a method for detecting and identifying objects with a small height extension on a roadway located in front of a vehicle in the direction of travel. Disclosure of the invention

[0009] The method according to the invention for detecting an obstacle in front of a vehicle by means of a radar sensor, wherein the radar sensor monitors a monitoring area in front of the vehicle, wherein a classification of an object detected by the radar sensor in the monitoring area is carried out, has the advantage that the monitoring area is divided into monitoring sub-areas and for a first monitoring sub-area a first classification rule is used for classification and for a second monitoring sub-area a second classification rule is used for classification, wherein the first

[0010] The first monitoring sub-area extends between the radar sensor and a ground intersection line, and the second monitoring sub-area extends between the ground intersection line and a maximum monitoring distance, the ground intersection line being a line defined by an intersection of a radar lobe of the radar sensor with the ground, the first classification rule being independent of an intrinsic speed of the object detected in the monitoring area, and the second classification rule comprising classifying the object detected in the monitoring area as a function of the intrinsic speed of the object detected in the monitoring area.

[0011] Advantageous embodiments are the subject of the subclaims.

[0012] In a top view of the vehicle with the radar sensor, the ground intersection line is approximately a circular segment. Between the vehicle and the circular segment of the ground intersection line, the beams emitted by the radar sensor do not hit the ground. In the area between the ground intersection line and the maximum detection distance, the beams emitted by the radar sensor do hit the ground.

[0013] It is advantageous that the ground intersection line is determined using the radar sensor. This allows for a self-calibrating process. In a particularly advantageous embodiment, the radar sensor scans the area in front of the vehicle, and the ground intersection line is determined by locating the radar reflections emanating from the ground. The area in which ground reflections can be located then forms the area between the ground intersection line and the maximum monitoring distance, so that the boundary of this area towards the radar sensor represents the ground intersection line.

[0014] In a particularly advantageous embodiment, the first classification rule is a rule that states that both moving and stationary objects are classified as obstacles.

[0015] In a particularly advantageous embodiment, the second classification rule specifies that only moving objects are classified as obstacles. This allows for a robust distinction between an obstacle and a ground reflection.

[0016] It is advantageous that the second classification rule comprises classifying the object detected in the surveillance area as an obstacle if the radar sensor detects that the object detected in the surveillance area has an inherent speed other than zero or had this at an earlier point in time. An earlier point in time is, in particular, a previous measurement cycle. Objects are therefore classified as obstacles if they are moving or have moved at an earlier point in time, i.e., in a previous measurement cycle. In a particularly advantageous embodiment, objects that have moved in an earlier measurement cycle are only classified as an obstacle if the objects are registered in every measurement cycle between a current measurement cycle and the previous measurement cycle in which the object in question was first classified as an obstacle.In other words, in this advantageous embodiment, objects are only classified as obstacles if they can be identified as a moving object without interruption in each measurement cycle since initial detection.

[0017] It is advantageous if the vehicle is an off-highway vehicle. When using the method according to the invention in conjunction with an off-highway vehicle, it can be exploited in a particularly advantageous manner that the method according to the invention provides a method which does not require an application and thus makes it possible to mount a radar sensor at almost any location on the off-highway vehicle and put it into operation directly. When using the method according to the invention in conjunction with an off-highway vehicle, it can also be advantageously used that the method according to the invention is a method which is very robust in terms of the occurrence of ground reflections of varying strengths, so that the method according to the invention can be used particularly advantageously for construction site vehicles or agricultural machinery.Agricultural machinery typically travels on surfaces that are very uneven and can therefore cause very strong ground reflections.

[0018] Advantageous is a device which is configured to carry out each step of the method according to the invention.

[0019] A computer program that causes a control unit to execute each step of the method according to the invention when the computer program is executed on the control unit is advantageous. A storage medium on which the computer program is stored is also advantageous.

[0020] An exemplary embodiment of the invention is presented in more detail below. It shows: Short description of the drawings Fig. 1 a schematic representation of a vehicle comprising a device configured to carry out an embodiment of the method according to the invention; Fig. 2 a schematic plan view of the vehicle; Fig. 3 a schematic representation of the sequence of an embodiment of the method according to the invention.

[0021] Fig. 1 shows a schematic representation of a vehicle (10) comprising a device configured to carry out an exemplary embodiment of the method according to the invention. The vehicle (10) can be, in particular, an off-highway vehicle, in particular an agricultural machine or a construction vehicle. The vehicle (10) comprises a control unit (12), which in turn comprises a storage medium (14). The vehicle (10) also comprises a radar sensor (16). The radar sensor (16) is arranged on the vehicle (10) in such a way that an area in front of the vehicle (10) can be monitored for the presence of obstacles. The radar sensor (16) is also arranged on the vehicle (10) in such a way that a monitoring area (20, 21) can be divided into two monitoring sub-areas.A first monitoring sub-area (20) comprises the area between the radar sensor (16) and a ground intersection line (22), the ground intersection line being defined by the radar beam of the radar sensor (16) touching the ground in the area of ​​the ground intersection line. The monitoring area also comprises a second monitoring sub-area (21) extending from the ground intersection line (22) to a maximum monitoring distance (23). The radar sensor (16) is configured such that objects (24, 25) detected by the radar sensor (16) can be classified, so that as a result of the classification, a detected object (24, 25) is possibly recognized as an obstacle and a corresponding reaction, for example the issuance of a warning, is initiated. A direction of travel of the vehicle (10) is indicated by the arrow (18).

[0022] Fig. 2 shows a schematic plan view of the Fig. 1 vehicle shown (10). Fig. 2 serves to better illustrate the definition of the floor section line (22), which in plan view has approximately the shape of a circular segment. The Fig. Reference symbols already described in Figure 1 are not described again here. The radar beam of the radar sensor (16) strikes the ground in the second monitoring sub-area (21), so that a plurality of ground reflections (26) are detected in the second monitoring sub-area (21). For reasons of clarity, not all ground reflections are provided with reference symbols. A line approximately shaped like a segment of a circle on a flat ground, which delimits the ground reflections (26) toward the vehicle (10), defines the ground intersection line (22).

[0023] Fig. 3 shows a schematic sequence of an embodiment of the method according to the invention. Fig. The embodiment illustrated in Figure 3 starts in step 100. In step 100, the ground in front of the vehicle (10) is scanned using the radar sensor (16). This is followed by step 110.

[0024] In step 110, the ground intersection line (22) is determined based on the radar reflections received in step 100. For this purpose, a line is defined that delimits the ground reflections (26) detected in step 100 from the vehicle (10). Step 120 is then performed. In an alternative embodiment, the position of the ground intersection line (22) is read from a memory. In this alternative embodiment, the ground intersection line (22) is previously determined from an installation height, a viewing angle, and an inclination of the radar sensor (16) and stored in the memory.

[0025] In step 120, a check is performed to determine whether any objects (24, 25) are present in the surveillance area (20, 21). In other words, in step 120, a check is performed to determine whether any radar reflections are present in the surveillance area (20, 21) that were not clearly identified as ground reflections (26). If no objects (24, 25) are present in the surveillance area, step 100 follows again after step 120. If objects (24, 25) are present in the surveillance area, step 130 follows after step 120.

[0026] In step 130, the position of each object (24, 25) detected in step 120 is determined. Furthermore, each object (24, 25) detected in step 120 is assigned a distinguishable index. Step 140 is then performed.

[0027] In step 140, a check is performed to determine whether the object (24, 25) with the lowest index is located in the first monitoring sub-area (20). If the object (24) with the checked index is not located in the first monitoring sub-area (20), step 160 is then performed. If the object (25) is located in the first monitoring sub-area (20), step 150 is performed following step 140.

[0028] In step 150, the object (25) with the current index is classified using the first classification rule. The first classification rule classifies any object (25) whose radar echo exceeds a predefined threshold as an obstacle. The intrinsic speed of the object (25) to be classified is not taken into account within the scope of the first classification rule. Following step 150, step 170 is performed.

[0029] In step 170, it is checked whether the object (25) classified in step 150 is an obstacle. If it is an obstacle, step 190 is performed after step 170. If it is not an obstacle, step 140 is performed again after step 170. When step 140 is performed again, the index is incremented.

[0030] In step 190, the detected obstacle triggers a reaction. This reaction may, for example, be a visual or acoustic warning that is brought to the attention of a user of the vehicle (10). The reaction in step 190 may also be an automatic intervention in the operation of the vehicle (10), for example, emergency braking.

[0031] In step 160, the object to be checked is classified based on the second classification rule. The second classification rule provides for an object (24) to be classified as an obstacle only if the object (24) has an inherent speed other than zero or had an inherent speed other than zero in a previous run of the exemplary embodiment of the method according to the invention. In an advantageous embodiment, a stationary object (24) is only classified as an obstacle within the framework of the second classification rules if the stationary object (24) in the current cycle of the method according to the invention has been detected in every previously completed cycle since the first cycle in which the object had an inherent speed other than zero. Step 180 is then carried out.

[0032] In step 180, a check is made to determine whether the object (24) classified in step 160 represents an obstacle or not. If the object (24) classified in step 160 represents an obstacle, step 190 is performed following step 180. If the object (24) does not represent an obstacle, step 140 is performed again following step 180, with the index being incremented upon repeating step 140.

[0033] The presented embodiment of the method according to the invention is repeated cyclically (in Fig. 3 not shown), so that the floor cutting line (22) is always updated.

[0034] The presented embodiment of the method according to the invention thus enables reliable detection of obstacles with the aid of a radar sensor (16) which does not need to be calibrated after being mounted on a vehicle (10), but rather calibrates itself by detecting the ground intersection line (22) and subsequently dividing the monitoring area into a first partial monitoring area (20) and a second partial monitoring area (21).

Claims

[1] Method for detecting an obstacle in front of a vehicle (10) by means of a radar sensor (16), wherein the radar sensor (16) monitors a monitoring area in front of the vehicle (10), wherein a classification of an object (24, 25) detected in the monitoring area by means of the radar sensor is carried out, wherein the monitoring area is divided into monitoring sub-areas (20, 21) and a first classification rule is used for classification for a first monitoring sub-area (20) and a second classification rule is used for classification for a second monitoring sub-area (21), characterized byin that the first monitoring sub-area (20) extends between the radar sensor (16) and a ground intersection line (22) and the second monitoring sub-area (21) extends between the ground intersection line (22) and a maximum monitoring distance (23), wherein the ground intersection line (22) is a line defined by an intersection of a radar lobe of the radar sensor (16) with the ground, wherein the first classification rule is independent of an intrinsic speed of the object (25) detected in the monitoring area, and the second classification rule comprises classifying the object (24) detected in the monitoring area as a function of the intrinsic speed of the object (24) detected in the monitoring area. [2] Method according to claim 1, characterized by that the ground intersection line (22) is determined by means of the radar sensor (16). [3] Method according to claim 1 or 2, characterized bythat the second classification rule comprises classifying the object (24) detected in the monitoring area as an obstacle if it is detected by means of the radar sensor (16) that the object (24) detected in the monitoring area has or had at an earlier time an own speed different from zero. [4] Method according to one of the preceding claims, characterized by that the vehicle (10) is an off-highway vehicle. [5] Apparatus arranged to carry out each step of the method according to any one of claims 1 to 4. [6] A computer program that causes a control unit to carry out each step of the method according to any one of claims 1 to 4 when the computer program is executed on the control unit. [7] Storage medium on which the computer program according to claim 6 is stored.

Images