JP2020197506A - Object detector for vehicles - Google Patents

Abstract

To eliminate the influence of other objects that generate a multipath for a reflected radar wave from a detection object and reliability discriminate between a ghost and a real object when detecting an object on the outside of a vehicle by a radar device.SOLUTION: A reflection point determination unit 31 determines whether or not there is a reflection point on a first object in the surrounding of the host vehicle that reflects a radar wave reflected by the first object. When a second object is detected following the first object while there is a reflection point, a first ghost determination unit 32 determines, on the basis of the relative speed of the reflection point to the host vehicle, the relative speed of the first object to the host vehicle and the relative speed of the second object to the host vehicle, whether or not a first ghost condition that the second object is provisionally assumed as a ghost holds true. Furthermore, when the first ghost condition of the second object holds true, a second ghost determination unit 33 determines, on the basis of a second ghost condition pertaining to a physical relationship where at least the second object is detected, whether or not the second object is a ghost.SELECTED DRAWING: Figure 1

Description

The present invention relates to a vehicle object detection device that detects an object existing around a vehicle by using a radar device.

In vehicles such as automobiles, a camera or radar device is installed to detect objects existing outside the vehicle, avoiding collisions with obstacles in front, controlling following vehicles in front, warning control and steering against vehicle wobbling and lane departure. A system that executes various support controls for the driver such as control has been put into practical use, and an automatic driving system that does not require the driving operation of the driver is being put into practical use.

In particular, radar devices are less affected by weather such as rain, snow, fog, and backlight compared to image sensors that use visible or infrared light such as cameras, and are vehicles that travel in various changing environments. It is regarded as a promising sensor for detecting objects in the outside world.

However, the radar device has a problem that due to the deviation of the radio wave in the time axis direction, a ghost that is received as if an object exists even though the object does not actually exist is generated. It is necessary to surely distinguish it from the actual object.

Therefore, for example, in Patent Document 1, the radar beam is scanned substantially parallel to the road surface on which the vehicle travels, and an object within the radar beam scanning range is detected every one or more times of the radar beam scanning. Proposed a technique for determining whether or not the first detection object, which is a detection object on the own lane in the road surface, is a ghost, based on the second detection object, which is a detection object on the adjacent lane adjacent to the own lane. Has been done.

Japanese Unexamined Patent Publication No. 2000-147115

Patent Document 1 determines whether or not the first detected object is a ghost, with the distance and relative velocity between the second detected object and the first detected object, the angle formed by each direction, and the like as necessary conditions. The condition is that the second detection object is an actual entity that actually exists.

However, when the first detection object is a real object, the radar wave reflected by the first detection object, which is a real object, is reflected by another object, and the reflected wave is reflected by the first detection object again to perform radar. A multi-pass that is received by the device may occur, and it may be difficult to identify whether the second detected object is a real object or a ghost.

That is, if there is another object that generates multipath for the radar reflected wave from the detection target in the driving environment of the own vehicle, the ghost may be erroneously detected as an actual entity due to the influence. On the contrary, there is a possibility that the actual object is mistakenly judged as a ghost and removed, which may cause a serious hindrance to the traveling control of the vehicle.

The present invention has been made in view of the above circumstances, and when an object outside the vehicle is detected by a radar device, the influence of another object that generates multipath on the radar reflected wave from the detection target is eliminated. An object of the present invention is to provide a vehicle object detection device capable of reliably distinguishing a ghost from an actual object.

The vehicle object detection device according to one aspect of the present invention is a vehicle object detection device that transmits a radar wave from a radar device, receives the reflected wave reflected by the object, and detects the object existing around the own vehicle. A reflection point determination unit that determines whether or not there is a reflection point that reflects the radar wave reflected by the first object around the own vehicle with respect to the first object detected by the radar device. When the radar device detects the second object following the first object in the presence of the reflection point, the relative speed of the reflection point with respect to the own vehicle and the relative speed of the first object with respect to the own vehicle. A first ghost determination unit that determines whether or not the first ghost condition for tentatively ghosting the second object is satisfied based on the relative speed and the relative speed of the second object with respect to the own vehicle. When the first ghost condition of the second object is satisfied, it is determined whether or not the second object is a ghost, at least based on the second ghost condition related to the positional relationship in which the second object is detected. It is provided with a second ghost judgment unit.

According to the present invention, when an object outside the vehicle is detected by the radar device, the influence of another object that generates multipath on the radar reflected wave from the detection target is eliminated, and the ghost and the actual object are ensured. Can be identified.

Configuration diagram of the object detection device for vehicles Explanatory drawing showing the influence of the front object on the radar wave behind the own vehicle Explanatory drawing showing the positional relationship between the own vehicle, the following vehicle, the front object, and the approaching vehicle Flowchart of ghost judgment routine

Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a configuration diagram of a vehicle object detection device. The vehicle object detection device 1 shown in FIG. 1 detects an object around the vehicle by an in-vehicle radar device 10, and connects the radar device 10, the traveling environment detection device 20, and the radar status monitoring device 30 with the communication bus 100. It is configured to be connected so that bidirectional communication is possible via.

The radar device 10 uses, for example, a millimeter wave as a radar transmission wave, and detects the relative velocity, distance, and orientation of an object existing around the own vehicle with respect to the own vehicle by a well-known FMCW (Frequency Modulated Continuous Wave) method or a pulse method. To do. The radar device 10 includes, for example, a plurality of radar units 11, and each radar unit 11 is arranged at a plurality of locations in the vehicle. FIG. 1 shows an example in which the radar device 10 includes four radar units 11.

The radar unit 11 combines or integrates a transmitting antenna that radiates a radar transmitting wave into space and a receiving antenna that receives a reflected wave that is reflected by an object and returned, and is a unit together with a circuit board for signal processing. It is a modified version. The radar unit 11 is arranged at the left and right corners of the front and rear bumpers of the vehicle, for example.

In the radar device 10, the distance of the object can be calculated based on, for example, a signal processing algorithm based on the time difference between the radar transmission wave and the reception wave. Further, the velocity (relative velocity) of an object can be calculated based on a signal processing algorithm based on the Doppler effect or a change in distance. Further, the orientation of the object can be calculated based on a signal processing algorithm using the amplitude and phase of the received wave. The calculation process of the distance, speed, and direction of the object using the radar device 10 can be performed by using a general algorithm, and detailed description thereof will be omitted here.

The driving environment detection device 20 processes a locator unit 21 that detects the vehicle position of the own vehicle based on signals from a plurality of navigation satellites such as a GNSS (Global Navigation Satellite System) satellite, and an image of the surroundings of the own vehicle. It is configured to include an image processing unit 22 that recognizes the external environment.

The locator unit 21 receives a signal including information on the orbit and time transmitted from a plurality of navigation satellites, and based on the received signal, sets the self-position of the own vehicle, and includes longitude, latitude, altitude, and time information. Position as an absolute position. In addition, when the positioning accuracy deteriorates due to the capture state of the signal (radio wave) from the satellite or the influence of multipath due to the reflection of the radio wave, the locator unit 21 determines the traveling direction and the moving distance of the own vehicle. Position your vehicle's position as a relative position change.

Further, the locator unit 21 is provided with a map database DB, and identifies the position of the map database DB on the map data from the position data of the own vehicle that has been positioned. The map database DB is a database that holds a high-precision map created for driving control including automatic driving, and is stored in a large-capacity storage medium such as an HDD (hard disk drive) or SSD (solid state drive). ..

In detail, a high-precision map is a multidimensional map that holds static information such as road shapes and connections between roads and dynamic information such as traffic information collected by infrastructure communication in multiple layers. It is configured as (dynamic map). The road data includes the type of white line, the number of driving lanes, the width of the driving lane, the point sequence data indicating the center position in the width direction of the driving lane, the curvature of the traveling lane, the traveling azimuth of the traveling lane, the speed limit, etc. include.

The image processing unit 22 processes the image captured by the in-vehicle camera unit 22a to recognize the external environment around the own vehicle. As the camera unit 22a, a monocular camera or a stereo camera composed of two cameras that capture the same object from different viewpoints is used.

When a stereo camera is used, the external environment is recognized three-dimensionally by stereo-processing a pair of left and right images captured by the stereo camera. In the camera unit 22a as a stereo camera, for example, two shutter-synchronized color cameras having image pickup elements such as CCD and CMOS near the rearview mirror inside the front window in the upper part of the vehicle interior have a predetermined baseline length in the vehicle width direction. It is arranged on the left and right.

The pair of left and right images captured by the stereo camera are matched, and the amount of pixel shift (parallax) at the corresponding position between the left and right images is obtained. Then, this pixel shift amount is converted into luminance data or the like to generate a distance image. Based on the principle of triangular survey, the points on the distance image are coordinate-converted to points in the real space centered on the own vehicle, and the left and right white lines, obstacles, and own vehicle that divide the driving lane of the road on which the own vehicle travels. Vehicles and the like traveling in front of the vehicle are three-dimensionally recognized.

The white lines on the left and right of the road can be recognized by extracting a point cloud that is a candidate for the white line from the image and calculating a straight line or a curve connecting the candidate points. For example, in the white line detection area set on the image, edges whose brightness changes by a predetermined value or more are detected on a plurality of search lines set in the horizontal direction (vehicle width direction), and one set of white lines is detected for each search line. By detecting the start point and the white line end point, an intermediate region between the white line start point and the white line end point is extracted as a white line candidate point.

Then, the white line is recognized by processing the time series data of the spatial coordinate position of the white line candidate point based on the amount of vehicle movement per unit time and calculating a model that approximates the left and right white lines. As the approximate model of the white line, an approximate model in which linear components obtained by Hough transform are connected, or a model approximated by a curve such as a quadratic equation can be used.

The radar status monitoring device 30 monitors the status of object detection by the radar device 10 based on the information from the traveling environment detection device 20, and the object detected by the radar device 10 is an actual entity or does not actually exist. Determine if it is a ghost. Therefore, the radar status monitoring device 30 includes a reflection point determination unit 31, a first ghost determination unit 32, and a second ghost determination unit 33.

In the reflection point determination unit 31, another object serving as a reflection point that reflects the radar wave reflected by the first object is formed in the traveling environment of the own vehicle with respect to the object (first object) detected by the radar device 10. Determine if it exists. For example, when a radar wave is transmitted from the radar unit 11 at the rear of the own vehicle and a following vehicle traveling behind the own vehicle is detected, if a reflection point exists in front of the own vehicle, the reflected wave reflected by the following vehicle is multi-layered. Ghosts due to paths occur, increasing the possibility of falsely detecting ghosts as real objects.

FIG. 2 is an explanatory diagram showing the influence of the front object on the radar wave behind the own vehicle. In the scene shown in FIG. 2, an object SG that serves as a reflection point for reflecting radar waves exists in front of the own vehicle Cs, and a following vehicle as a first object OBJ1 to be detected exists behind the own vehicle Cs. doing. When the radar wave is transmitted from the radar unit 11 at the rear of the own vehicle Cs in a scene where the object SG (reflection point) exists at a position opposite to the first object OBJ1 with the own vehicle Cs in between. , This radar wave is reflected by the first object (following vehicle) OBJ1, the reflected wave is received by the radar unit 11, and the first object OBJ1 is detected.

At the same time, the reflected wave reflected by the first object OBJ1 travels toward the object SG in front of the own vehicle Cs and is reflected by the object SG. The radar wave reflected by the object SG hits the first object OBJ1 again, is reflected and received by the radar unit 11, and a ghost is generated in the radar unit 11. Therefore, when the radar unit 11 detects the second object OBJ2 approaching from the rear side of the first object OBJ1, the second object OBJ2 is an object even if the second object OBJ2 actually exists. There is a high possibility that the ghost generated by the reflected wave of SG is erroneously determined and removed.

The reflection point determination unit 31 determines whether or not an object as a reflection point that may generate a ghost due to such multipath exists in the traveling environment of the own vehicle. Specifically, the reflection point determination unit 31 displays the destination and the road exit around the own vehicle based on the own vehicle position information and map information from the locator unit 21 and the image recognition information from the image processing unit 22. It is determined whether or not there is a reflection point that reflects the radar wave of the preceding vehicle, which is taller than the own vehicle, such as the signboard, the upper wall surface of the tunnel, and the trailer.

When the first ghost determination unit 32 determines that the reflection point exists by the reflection point determination unit 31 and further detects the second object following the first object, the relative speed of the reflection point with respect to the own vehicle and the relative speed with respect to the own vehicle. Based on the relative speed of the first object and the relative speed of the second object with respect to the own vehicle, it is determined whether or not the first ghost condition for tentatively ghosting the second object is satisfied.

Explaining with the example of FIG. 2, the relative speed between the own vehicle Cs and the object (reflection point) SG is Vos, the relative speed between the own vehicle Cs and the first object (following vehicle) OBJ1 is Vrel1, and the own vehicle Cs and the second When the relative velocity with the object OBJ2 is Vrel2, it is determined whether or not the first ghost condition shown in the following equation (1) is satisfied within the allowable range set from the velocity resolution of the radar unit 11.
2 × (Vos + Vrel1) = Vos + Vrel2 → Vos + 2 × Vrel1 = Vrel2… (1)

In equation (1), when the relative velocity (Vos + Vrel2) of the second object OBJ2 with respect to the object SG is equal to twice the relative velocity (Vos + Vrel1) of the first object OBJ1 with respect to the object SG, the second object OBJ2 is the first. This means that there is a high possibility that the reflected wave of the object OBJ is reflected by the object SG, then reflected by the first object OBJ1 again, received by the radar unit 11, and detected. Therefore, the first ghost determination unit 32 tentatively determines that the second object OBJ2 is a ghost when the condition of the equation (1) is satisfied. Based on this provisional ghost judgment, the second object OBJ2 is to be removed as noise.

When the first ghost condition for the second object is satisfied, the second ghost determination unit 33 determines whether the second object is really a ghost based on at least the second ghost condition related to the positional relationship in which the second object is detected. The final judgment is made to prevent the second object from being accidentally removed. The second ghost condition includes, for example, a lane condition in which the second object exists in an area where the vehicle can travel, such as on an adjacent lane adjacent to the lane of the own vehicle, a reflection point, the own vehicle, the first object, and the second object. Mainly the distance condition related to the distance between the vehicle and the vehicle, including the condition related to the size of the second object such as a truck or a two-wheeled vehicle, the condition of the driving road such as an expressway or a general road, etc., these conditions are appropriately weighted. And combine.

When determining the lane condition, the second ghost determination unit 33 acquires the vehicle position information and map information from the locator unit 21, and the image recognition information from the image processing unit 22, and the lane in which the vehicle is currently traveling. Find out if there is a lane next to. As shown in FIG. 2, when an adjacent lane LNr exists next to the traveling lane LNs of the own vehicle and the second object OBJ2 is detected on the adjacent lane LNr (lane condition is satisfied), the second ghost determination unit 33 determines that the second object OBJ2 is not a ghost but a real object (approaching vehicle) as the second ghost condition is not satisfied. On the other hand, when the adjacent lane does not exist or the second object OBJ2 is outside the adjacent lane (lane condition is not satisfied), the second ghost determination unit 33 determines that the second ghost condition is satisfied and the second object OBJ2 is a ghost. to decide.

On the other hand, when determining the distance condition, the second ghost determination unit 33 examines the relationship between the distances DB, DS, and DE as shown in FIG. FIG. 3 is an explanatory diagram showing the positional relationship between the own vehicle, the following vehicle, the front object, and the approaching vehicle. In FIG. 3, the distance from the own vehicle Cs to the first object (following vehicle) OBJ1 is DD, the distance from the object (reflection point) SG to the first object OBJ1 is DS, and the distance from the own vehicle Cs to the second object (approaching vehicle). With the distance to OBJ2 as DE, the second ghost determination unit 33 determines that the relationship between these distances DD, DS, and DE is within the permissible range set by the distance resolution of the radar unit 11 under the condition of the following equation (2). Find out if you are satisfied with.
DS + DB = DE ... (2)

Equation (2) is the sum of the distance DB from the own vehicle Cs to the first object (following vehicle) OBJ1 and the distance DS from the object (reflection point) SG to the first object OBJ1, that is, the first object from the own vehicle Cs. It means that the path length of the radio wave reflected by the OBJ1 to the object SG is equal to the path length of the radio wave from the own vehicle Cs to the second object OBJ2. Therefore, when the first ghost determination unit 32 tentatively determines that the second object OBJ2 is a ghost and the condition of equation (2) is satisfied, the second ghost determination unit 33 determines that the second ghost condition is satisfied. The second object OBJ2 is determined to be a ghost.

The above lane conditions and distance conditions may be used individually or in combination. However, when combining lane conditions and distance conditions, lane conditions are prioritized. For example, if the lane condition is satisfied while the second object is tentatively determined to be a ghost, it is determined that the second object is a real object regardless of the distance condition, and the lane condition is not satisfied. However, if the distance condition is satisfied, it is determined that the second object is a ghost.

When the second ghost determination unit 33 finally determines that the second object is a ghost, it removes the data of the second object, and finally determines that the second object is not a ghost but a real object. The data of the first object and the second object are transmitted to a travel control device or the like (not shown) via the communication bus 100. The travel control device executes travel control while ensuring safety for the object detected by each radar unit 11 of the radar device 10.

Next, the ghost determination process in the radar status monitoring device 30 will be described with reference to the flowchart of FIG. FIG. 4 is a flowchart of the ghost determination routine. Note that FIG. 4 shows a processing example when the rear region of the own vehicle is monitored by the radar unit 11, but the same applies to the case where the front region is monitored.

In step S1, the radar status monitoring device 30 checks whether or not a vehicle (rear vehicle) is detected behind the own vehicle by the radar unit 11 at the rear of the vehicle. If the rear vehicle is not detected, the radar status monitoring device 30 exits this routine and returns to the process of step S1 in the next control cycle, and if the rear vehicle is detected, the process from step S1 to step S2. move on.

In step S2, the radar status monitoring device 30 acquires the position of the rear vehicle (distance from the own vehicle to the rear vehicle) and the relative speed of the rear vehicle with respect to the own vehicle from the radar unit 11. Next, the process proceeds to step S3, and the radar status monitoring device 30 checks whether or not a moving object (rear approaching object) approaching the own vehicle is detected by the radar unit 11 following the rear vehicle.

If the rear approaching object is not detected in step S3, the radar status monitoring device 30 exits this routine from step S3 and resumes the processing of step S1 in the next control cycle. When a rear approaching object is detected, the radar status monitoring device 30 proceeds from step S3 to step S4, and the position of the rear approaching object (distance from the own vehicle to the rear approaching object) from the radar unit 11 and the own vehicle. Gets the relative velocity of the object approaching backwards to.

Next, the radar status monitoring device 30 proceeds from step S4 to the process of step S5, and reads the own vehicle position information, the map information, and the image recognition information from the locator unit 21 and the image processing unit 22 of the traveling environment detection device 20. Then, the radar status monitoring device 30 is a vehicle rather than the own vehicle such as a sign indicating the destination or the road exit, the upper wall surface of the tunnel, the trailer, etc. in front of the own vehicle from the own vehicle position information, the map information, and the image recognition information. It is determined whether or not there is a reflection point that reflects radar waves of a high preceding vehicle or the like.

When there is no reflection point in front of the own vehicle, the radar status monitoring device 30 proceeds from step S5 to step S9, and determines that the rear approaching object is an actual entity (rear approaching vehicle). Then, the data of the vehicle approaching behind is transmitted together with the data of the vehicle behind, and the routine is exited.

On the other hand, when the reflection point exists in front of the own vehicle, the radar condition monitoring device 30 proceeds from step S5 to step S6, and whether or not the first ghost condition for tentatively ghosting the rear approaching object is satisfied. Find out. If the first ghost condition is not satisfied, the radar status monitoring device 30 proceeds from step S6 to step S9, determines that the rear approaching object is an entity, and transmits the data of the rear approaching object together with the data of the rear vehicle.

On the other hand, when the first ghost condition is satisfied in step S6, the radar condition monitoring device 30 further investigates whether or not the second ghost condition is satisfied in step S7, and whether the object is a rear approaching object or is really a ghost. Determine if it is an entity that exists in. As a result, if the second ghost condition is satisfied, the radar status monitoring device 30 determines in step S8 that the rear approaching object is a ghost and removes the data. If the second ghost condition is not satisfied, the step In S9, it is determined that the rear approaching object is a real object.

As described above, in the present embodiment, it is determined whether or not there is a reflection point that reflects the radar wave reflected by the first object around the own vehicle with respect to the first object detected by the radar device. When the radar device detects the second object following the first object in the presence of the reflection point, the relative velocity of the reflection point with respect to the own vehicle, the relative velocity of the first object with respect to the own vehicle, and the second object with respect to the own vehicle. Based on the relative velocity of the object, it is determined whether or not the first ghost condition for tentatively ghosting the second object is satisfied. Then, when the first ghost condition of the second object is satisfied, it is determined whether or not the second object is a ghost, at least based on the second ghost condition related to the positional relationship in which the second object is detected. As a result, when an object outside the vehicle is detected by the radar device, the influence of other objects that generate multipath on the radar reflected wave from the detection target is eliminated, and the ghost and the actual object are reliably distinguished. be able to.

1 Vehicle object detection device 10 Radar device 11 Radar unit 20 Driving environment detection device 21 Locator unit 22 Image processing unit 30 Radar status monitoring device 31 Reflection point judgment unit 32 1st ghost judgment unit 33 2nd ghost judgment unit

Claims (4)

A vehicle object detection device that transmits radar waves from a radar device, receives reflected waves reflected by an object, and detects the object existing around the own vehicle.
A reflection point determination unit that determines whether or not there is a reflection point that reflects the radar wave reflected by the first object around the own vehicle with respect to the first object detected by the radar device.
When the radar device detects the second object following the first object in the presence of the reflection point, the relative velocity of the reflection point with respect to the own vehicle and the relative velocity of the first object with respect to the own vehicle. And a first ghost determination unit that determines whether or not the first ghost condition for tentatively ghosting the second object is satisfied based on the relative speed of the second object with respect to the own vehicle.
When the first ghost condition of the second object is satisfied, it is determined whether or not the second object is a ghost, at least based on the second ghost condition related to the positional relationship in which the second object is detected. A vehicle object detection device including a second ghost determination unit. In the first ghost determination unit, the value obtained by adding the relative speed of the reflection point with respect to the own vehicle and the speed twice the relative speed of the first object with respect to the own vehicle is the second value with respect to the own vehicle. The vehicle object detection device according to claim 1, wherein it is determined that the first ghost condition is satisfied when the relative speed of the object is substantially equal to the relative speed of the object. The second ghost determination unit includes the condition that the second object exists in the region where the vehicle can travel in the second ghost condition, and when the second object exists in the region, the second object The vehicle object detection device according to claim 1 or 2, wherein it is determined that the object is not a ghost. The second ghost determination unit includes the condition of the distance between the own vehicle, the reflection point, the first object, and the second object in the second ghost condition, and from the own vehicle to the first object. When the sum of the distance from the reflection point and the distance from the reflection point to the first object substantially matches the distance from the own vehicle to the second object, it is determined that the second object is a ghost. The vehicle object detection device according to claim 1 or 2.

Images