JP7363574B2 - object detection device - Google Patents

Description

The present invention relates to an object detection device that detects objects existing around a vehicle (hereinafter referred to as the vehicle).

Conventionally, objects such as sonar have been used to detect objects around the vehicle, and the detection results are used for vehicle control such as automatic parking. For example, there is an object detection device proposed in Patent Document 1.

This device includes a vote count calculation section, a cluster setting section, a grid discrimination section, and a moving object discrimination section. The vote number calculation unit calculates the number of votes, which is the number including the detection points of the distance measurement sensor, for each grid of the grid map based on the detection results of the detection points by the distance measurement sensor such as a radar sensor. The cluster setting unit sets clusters, which are objects to be detected, on the grid map by clustering the detection points based on the detection results of the detection points by the distance measuring sensor. The grid determining unit determines whether the grid occupied by the cluster is a stationary grid or a moving grid based on the total number of votes for each grid calculated multiple times. The moving object determination unit determines whether the cluster is a moving object or a stationary object based on the arrangement of a stationary grid and a moving grid included in the grid occupied by the cluster. Using these, it is possible to detect objects around the vehicle and to determine whether a cluster of objects to be detected is a moving object or a stationary object.

JP2017-187422A

However, in Patent Document 1, a grid map is used to make it possible to determine whether a moving object is a stationary object, but if the coordinates of an object are created using the grid map, a quantization error due to abstraction with the grid may occur. Occur.

In view of the above points, an object of the present invention is to provide an object detection device that can suppress the occurrence of quantization errors and perform accurate object detection even when abstracted into a grid using a grid map. do.

In order to achieve the above object, the invention according to claim 1 detects objects (100) existing around the own vehicle using a grid map in which the surroundings of the own vehicle (V) are divided into a plurality of grids. The sensing information of the distance sensor (10), which is an object detection device and measures the position of an object by outputting a search wave and acquiring the reflected wave of the search wave reflected by an object, is input, and a grid map is generated. an object recognition unit (31) that creates a grid map showing the detection coordinates of the detected position of the object included in the sensing information of the ranging sensor, and a main grid that is a grid corresponding to the detection coordinates; a grid creation unit (32) that creates a grid map showing the positions of objects around the host vehicle by setting a predetermined number of votes for surrounding grids located around the main grid; an object identification unit that identifies a grid in which the number of votes exceeds a predetermined threshold value in the grid map created by the creation unit as a grid in which an object exists, and sets detection coordinates corresponding to the identified grid as actual detection coordinates of the object; (33), the grid creation unit sets a first number of votes smaller than the threshold value in the main grid, and sets a second number of votes smaller than the first number in the surrounding grid, and In the grid map created by the object recognition unit, a voting number larger than a threshold value is set for a main grid that includes a plurality of detected coordinates and for main grids that are adjacent and continuous.

In this way, the detected coordinates of the object included in the sensing information of the ranging sensor are converted to grids on a grid map with low resolution, and the number of votes is set for the main grid and surrounding grids. Furthermore, the grid in which the object exists is identified by comparing the number of votes with a threshold value. Then, for a grid in which the presence of an object is detected based on the grid map, the coordinates of the grid are not output, but the coordinates detected by the distance sensor are output as they are.

Therefore, data that more accurately indicates the position of the object can be output than outputting grid coordinates. Therefore, even if the object is abstracted into a grid using a grid map, by outputting detection coordinates corresponding to the grid, it is possible to suppress the occurrence of quantization errors and perform accurate object detection.

Note that the reference numerals in parentheses attached to each component etc. indicate an example of the correspondence between the component etc. and specific components etc. described in the embodiments to be described later.

FIG. 1 is a block diagram of a vehicle control system according to a first embodiment. FIG. 7 is a diagram showing an example of the number of votes set for each grid when the center position of the diagram is the detected coordinates of an object. It is a flowchart of object detection processing. FIG. 7 is a diagram illustrating an example of a case where detected coordinates shown as a comparative example are converted to a grid on a grid map, and then the coordinates of the grid where an object is detected are output. FIG. 6 is a diagram illustrating an example of a case where, after converting the detected coordinates described as the first embodiment into a grid on a grid map, the sonar detected coordinates corresponding to the grid in which an object is detected are output.

Embodiments of the present invention will be described below based on the drawings. Note that in each of the following embodiments, parts that are the same or equivalent to each other will be described with the same reference numerals.

(First embodiment)
A vehicle control system equipped with an object detection device according to this embodiment will be described. Here, vehicle control will be explained using as an example parking assist control that performs automatic parking. It doesn't matter if it's something that does this.

As shown in FIG. 1, the vehicle control system 1 includes a sonar 10, an electronic control unit (hereinafter referred to as ECU) 30 that constitutes an object detection device, various actuators 40, and a display 50.

The sonar 10 corresponds to a distance measuring sensor. The sonar 10 outputs ultrasonic waves as exploration waves at every predetermined sampling period, and obtains the reflected waves to obtain information such as the relative speed and distance to the target, and the azimuth where the target exists. The position measurement results are sequentially output to the ECU 30 as sensing information. When the sonar 10 detects an object, the sonar 10 outputs sensing information including detection coordinates, which are the coordinates of the detected position. The detected coordinates of the object are determined using mobile triangulation, and since the distance to the object changes as the vehicle moves, it is determined based on changes in measurement results at each sampling period. There is.

Although only one sonar 10 is shown here, it is actually provided at multiple locations on the vehicle. The sonar 10 includes, for example, a front sonar in which a plurality of sonar units are arranged in a row in the left-right direction of the vehicle on the front and rear bumpers, and a side sonar arranged in a side position of the vehicle. The sampling period of the sonar 10 is set depending on the location of the sonar 10, for example, a first period for a front sonar, and a second period less than the first period for a side sonar.

Here, the sonar 10 is taken as an example of the distance measurement sensor, but examples of the distance measurement sensor include millimeter wave radar and LIDAR (Light Detection and Ranging). Millimeter wave radar performs measurements using millimeter waves as a search wave, and LIDAR performs measurements using laser light as a search wave, and both output search waves within a predetermined range, such as in front of a vehicle. , perform measurements within that output range.

The ECU 30 performs various controls to detect objects and realize parking support control for automatic parking, and is composed of a microcomputer equipped with a CPU, ROM, RAM, I/O, etc. There is. Specifically, the ECU 30 detects objects existing around the vehicle by creating a grid map based on sensing information from the sonar 10, and selects the detected coordinates of the objects. Furthermore, in the case of this embodiment, the ECU 30 performs parking assistance control based on the object detection result, and controls various actuators 40 based on the parking assistance control to perform automatic parking. Specifically, the ECU 30 is configured to include an object recognition section 31, a grid creation section 32, an output selection section 33, and a control section 34 as functional sections that execute various controls.

The object recognition unit 31 creates a grid map showing the positions of objects around the own vehicle based on sensing information input from the sonar 10 using a grid map in which the surroundings of the own vehicle are divided into a plurality of grids. For example, the object recognition unit 31 creates a grid map of a range of several meters by several meters, which is larger than the own vehicle and has the own vehicle as the center. The grid map is a map in which the area around the host vehicle is divided into a plurality of grids by dividing the area around the vehicle in a grid pattern in a plan view. Here, the object recognition unit 31 creates a grid map in which, for example, squares of several to tens of centimeters are used as one grid, that is, one unit area, and divided into a grid pattern in the longitudinal direction of the own vehicle and in the left and right directions of the own vehicle. There is. Then, the object recognition unit 31 recognizes in which grid on the grid map the detected coordinates of the object are placed.

Note that the resolution of the sonar 10 is high, and it is possible to configure the grid map with finer grids, but in order to reduce memory capacity and processing load, the data is thinned out and coarsened by increasing the grid size. Good too. Therefore, if the detected coordinates of an object are abstracted into a grid using a grid map, a quantization error will occur as described above.

The grid creation unit 32 converts the detected coordinates into a grid expressed by the number of votes, based on the recognition result of the object recognition unit 31 as to which grid on the grid map the detected coordinates of the object are placed. The number of votes is a value that is an index indicating the probability of existence of an object, and here, the larger the number of votes, the higher the probability of existence of the object. Specifically, the grid corresponding to the detected coordinates included in the sensing information from the sonar 10 is set as a main grid, the grids located around the main grid are set as surrounding grids, and the number of votes corresponding to each is set. Thereby, the grid creation unit 32 creates a grid map that reflects the detected coordinates included in the sensing information of the sonar 10.

For example, as shown in FIG. 2, the main grid is a grid corresponding to the detection coordinates of the object indicated by the sensing information of the sonar 10, that is, the coordinates indicating the position where the sonar 10 detects the presence of the object. This figure shows a case where the center of the figure is the main grid. Furthermore, grids located around the main grid are defined as surrounding grids. Then, a first number of votes is set for the main grid, and a second number of votes smaller than the first number is set for the peripheral grids. In this embodiment, the second voting number is set for each of the eight surrounding grids adjacent to the main grid. For example, for the first number of votes, a positive number of votes, in this case X, is set. The second number of votes is also set to be a positive number of votes, here Y. Note that both the first and second votes are positive votes here, but the first vote may be a positive vote, and the second vote may be 0 or negative. need not all be positive.

Furthermore, the number of votes is set for each detected coordinate. That is, if there are two detected coordinates in the same grid, that grid becomes the main grid, but the number of votes is set by adding the first numbers of votes for the two grids. When the first number of votes is X, 2×X, which is the number of votes for two, is set as the number of votes for that main grid. The same applies to the surrounding grid, and the number of votes is set by adding the two second numbers of votes. Furthermore, if the detected coordinates are included in adjacent grids, the grid containing the detected coordinates is both the main grid and the peripheral grid, so the first vote count is the number of adjacent main grids. The number of votes obtained by adding the second number of votes for several minutes is set.

The output selection unit 33 corresponds to an object identification unit, and identifies a grid where an object exists from the grid map created by the grid creation unit 32, and sends the detected coordinates of the object detected to exist in that grid to the control unit 34. Output. For example, the output selection unit 33 stores a threshold value of the number of votes, and if the number of votes for each grid in the grid map is less than the threshold value, it is determined that the grid does not have an object, and if it is equal to or greater than the threshold value, an object exists. It is determined that the grid is Here, the threshold value is set to Th, and a grid in which the total number of votes is equal to or greater than Th is specified as a grid in which an object exists. Although the threshold Th can be set arbitrarily, it is made to satisfy, for example, X<Th≦X+Y. Then, of the grid map created by the object recognition unit 31, the detected coordinates of the object corresponding to the grid in which the object is identified as being present are outputted to the control unit 34 as the detected coordinates of the actual object.

At this time, it is also possible to output the coordinates of the grid where the object exists to the control unit 34. However, since the resolution of the sonar 10 is high and the resolution of the grid map data is low and coarse, it would be more detailed data if the detected coordinates indicated by the sensing information of the sonar 10 were output as they were. For example, when outputting the coordinates of a grid where an object exists, the center coordinates of the grid will be output. However, the detected coordinates of the object and the center coordinates of the grid do not necessarily match. Therefore, it is possible to output data that more accurately indicates the position of the object by outputting the detected coordinates of the object as is, rather than outputting the coordinates of the grid. Therefore, even if the object is abstracted into a grid using a grid map, by outputting detection coordinates corresponding to the grid, it is possible to suppress the occurrence of quantization errors and perform accurate object detection.

In addition, if there are multiple detected coordinates included in the sensing information of the sonar 10 in the same grid, each of the multiple detected coordinates can be output, which is more accurate than outputting data from one rough grid. Data can be output.

Furthermore, since it is sufficient to simply output the detected coordinates of the object included in the sensing information of the sonar 10, there is no need to calculate or store grid coordinates again, reducing the memory capacity and processing load of the ECU 30. be able to.

Then, for a grid in which there is one detected coordinate of an object, the number of votes below the threshold value is added, and if the detected coordinates of an object also exist in the same grid or an adjacent grid, that grid The number of votes for exceeds a threshold. Therefore, even if only one detected coordinate of an object is included as noise, the number of votes for the main grid corresponding to that detected coordinate does not exceed the threshold value. Therefore, by quantizing the detected coordinates into a grid using a grid map, it is also possible to remove noise.

The control unit 34 performs parking support control for automatic parking. The parking assistance control includes generation of a parking route from the current position to the planned parking position, route following control for moving the vehicle along the generated parking route, and display control related to the parking assistance control. Parking route generation is performed based on the detected coordinates of the object transmitted from the output selection section 33. For example, the control unit 34 generates a parking route so as to avoid the detected coordinates of the object. Route following control is control that performs vehicle motion control such as acceleration/deceleration control and steering control in order to move the own vehicle along the parking route generated by parking route generation. In addition, if an object is newly detected in the parking route during parking assistance control, the control unit 34 regenerates the parking route or moves the own vehicle using route following control until the object moves. By stopping the robot, contact with objects can be avoided.

Display control displays information related to parking assistance control to inform the driver that parking assistance control is in progress, and to display information such as the parking route and the state of the own vehicle, such as where on the parking route the driver is moving. It controls the display. The control unit 34 controls the display 50 as display control.

Note that in this embodiment, the ECU 30 is configured as one and each functional unit is provided within the ECU 30, but it may be configured as a combination of multiple ECUs, and only a portion, for example, only the control unit 34 is provided. may be configured by one or more ECUs. Examples of the various ECUs when the control unit 34 is composed of a plurality of ECUs include a steering ECU that performs steering control, a power unit control ECU that performs acceleration/deceleration control, and a brake ECU.

Specifically, although not shown, the control unit 34 acquires detection signals output from each sensor mounted on the own vehicle, such as an accelerator position sensor, a brake pedal force sensor, a steering angle sensor, and a shift position sensor. . Then, the control section 34 detects the state of each section based on the acquired detection signals, and outputs control signals to the various actuators 40 in order to move the own vehicle following the parking route.

The various actuators 40 are various driving control devices related to running and stopping the own vehicle, and include an electronically controlled throttle 41, a brake actuator 42, an EPS (Electric Power Steering) motor 43, a transmission 44, and the like. These various actuators 40 are controlled based on control signals from the control unit 34, and the running direction, steering angle, and braking/driving torque of the own vehicle V are controlled. Thereby, parking assistance control including route following control for moving the own vehicle along the parking route and parking it at the planned parking position is realized.

The display device 50 includes a display 51 in a navigation system (hereinafter referred to as a navigation display), a display 52 provided in a meter (hereinafter referred to as a meter display), and the like. Here, during parking assistance control, the navigation display 51 and meter display 52 display the parking route and the position in the parking route where the vehicle is moving.

The vehicle control system 1 according to this embodiment is configured as described above. Next, the operation of the vehicle control system 1 according to this embodiment will be explained. The parking assistance control performed by the ECU 30 of the vehicle control system 1 includes parking route generation, route following control, and display control, but since these are as described above, their explanation will be omitted. Here, object detection including creation of a grid map used for parking route generation and output of object detection coordinates will be described with reference to FIG. 3. FIG. 3 is a flowchart of the object detection process executed by the ECU 30. This object detection process is executed at every predetermined control cycle, for example every cycle shorter than the sampling cycle of the sonar 10, when a start switch of the own vehicle such as an ignition switch is turned on. The object recognition section 31, grid creation section 32, and output selection section 33 of the ECU 30 function as a processing section and perform this object detection processing.

First, in step S100, it is determined whether sonar recognition has been performed, that is, whether the sensing information output from the sonar 10 includes information indicating that an object has been recognized. If the sensing information of the sonar 10 includes information indicating that an object has been recognized, the process advances to step S110, and if the sensing information does not include the information, the process is repeated.

If an object is detected by the sonar 10, each process from step S110 onwards is executed. In step S110, a grid corresponding to the detection coordinates included in the sensing information of the sonar 10 is searched. This grid becomes the main grid in the current control cycle. This process is performed for each detected coordinate, and a main grid is set for each detected coordinate. In step S115, the number of votes for the main grid searched in step S110 is added. That is, in the grid map created up to the previous control cycle, the number of votes for the main grid in the current control cycle is added. Here, X, which is the first number of votes, is added as the number of votes, and if two detection coordinates are included in the same grid, two first votes are added for that grid. The number of votes will be 2×X.

In step S120, the number of votes is added to eight surrounding grids located around the main grid. Here, Y, which is the second number of votes, is added as the number of votes. Surrounding grids are also set for each main grid that is set for each detection coordinate, so for the surrounding grid of a main grid that contains two detection coordinates in the same grid, the second vote count for two The votes are added up and the number of votes becomes 2×Y. Further, when adjacent grids include detection coordinates, the main grids are set to be consecutively adjacent, but each grid is also a surrounding grid for the other main grid. Therefore, for adjacent main grids, the second number of votes added as the surrounding grid is added to the first number of votes added as the main grid, and the number of votes becomes X+Y. In this way, the grid map is updated.

Next, the process advances to step S140, and output selection processing is performed. That is, the grid in which the object exists is specified from the grid map created in step S130, and the detected coordinates of the object detected to exist in the grid are output to the control unit 34. Specifically, grids in which the total number of votes is equal to or greater than a threshold value, in this case equal to or greater than Th, are determined to be grids in which an object exists from among the grid maps. Then, among the detected coordinates included in the sensing information of the sonar 10, the detected coordinates corresponding to those determined to be the grid in which the object is present are output to the control unit 34.

In this way, object detection including creation of a grid map used for parking route generation and output of the detected coordinates of the object is completed, and the detected coordinates of the object are transmitted to the control unit 34. Note that among the processes described here, the processes in steps S100 and S110 are executed by the object recognition unit 31, the processes in steps S120 and S130 are executed by the grid creation unit 32, and the processes in step S140 are executed by the output selection unit 33.

When the detected coordinates of the object are transmitted to the control unit 34, the control unit 34 executes parking route generation, route following control, and display control as parking assistance control. As a result, after a parking route is created while avoiding objects, the various actuators 40 are controlled to perform automatic parking that follows the parking route, and the display 50 shows the parking route and the position on the parking route. A display will be displayed to indicate whether the vehicle is moving.

As explained above, in the vehicle control system 1 of this embodiment, the detected coordinates of an object included in the sensing information of the sonar 10 are converted to a grid on a grid map with a low resolution, and the number of votes is assigned to the main grid and surrounding grids. It is set. Furthermore, the grid in which the object exists is identified by comparing the number of votes with a threshold value. Then, for grids in which the presence of an object is detected based on the grid map, the coordinates of the grid are not output, but the coordinates detected by the sonar 10 are output as they are. Therefore, data that more accurately indicates the position of the object can be output than outputting grid coordinates. Therefore, even if the object is abstracted into a grid using a grid map, by outputting detection coordinates corresponding to the grid, it is possible to suppress the occurrence of quantization errors and perform accurate object detection.

In addition, if there are multiple detected coordinates included in the sensing information of the sonar 10 in the same grid, each of the multiple detected coordinates can be output, which is more accurate than outputting data from one rough grid. Data can be output.

Furthermore, since it is sufficient to simply output the detected coordinates of the object included in the sensing information of the sonar 10, there is no need to calculate or store grid coordinates again, reducing the memory capacity and processing load of the ECU 30. be able to.

Furthermore, the number of votes is set in the main grid and surrounding grids for each detected coordinate of an object. Then, for a grid in which there is one detected coordinate of an object, the number of votes below the threshold value is added, and if the detected coordinates of an object also exist in the same grid or an adjacent grid, that grid The number of votes for exceeds a threshold. Therefore, even if only one detected coordinate of an object is included as noise, the number of votes for the main grid corresponding to that detected coordinate does not exceed the threshold value. Therefore, by quantizing the detected coordinates into a grid using a grid map, it is possible to remove noise.

Specifically, the mechanism of the object detection processing described above will be explained with reference to FIGS. 4 and 5. FIG. 4 is a comparative example, and shows a case where the detected coordinates of an object are converted into a grid and then the coordinates of the grid where the object exists are output. FIG. 5 is an example of this embodiment, and shows a case where the detected coordinates of an object are converted into a grid and then the detected coordinates of the object by the sonar 10 corresponding to the grid where the object is present are output. .

First, a comparative example will be described based on FIG. 4. As shown in (1) of FIG. 4, a grid that shows the positions of objects around the vehicle using a grid map that divides the area around the vehicle into multiple grids based on sensing information input from the sonar 10. Create a map. That is, a grid map of a predetermined range centered on the own vehicle is created, and it is recognized in which grid on the grid map the detected coordinates of the object are placed.

Next, as shown in (2) in Figure 4, based on the recognition result of which grid on the grid map the object's detected coordinates are placed, the detected coordinates are converted into a grid expressed by the number of votes. . At this time, if a plurality of detected coordinates of an object are included in the same grid, the number of votes corresponding to the number of detected coordinates is set. Therefore, since the main grid in the upper right corner of the figure includes two detected coordinates of objects, 2×X, which is the number of votes for two, is set as the number of votes for that main grid. On the other hand, since the main grid at the upper left in the figure includes only one detected coordinate of an object, X, which is the number of votes for one, is set as the number of votes for that main grid. In addition, the two adjacent grids containing the detected coordinates of the object in the second row from the bottom are both the main grid and the peripheral grid, so X+Y, which is the sum of the number of first votes and the number of second votes, is the number of votes. Set as a number.

Thereafter, grids whose number of votes is greater than or equal to a threshold value, Th in this case, are identified as grids in which objects exist. At this time, as shown in (3), the number of votes for the upper left grid in the figure is X, which is not greater than the threshold Th, so the coordinates of the grid can be prevented from being output. Therefore, even if noise is included in the sensing information from the sonar 10 as the detected coordinates of an object, it can be removed.

Further, for a grid in which the number of votes exceeds a threshold and an object is identified as being present, the coordinates of that grid, specifically the center coordinates of that grid, are output to the control unit 34. Therefore, in the upper right grid in the figure, only one coordinate is output even though two detected coordinates of the object were included in step (1). In addition, in both the grid at the top right of the figure and the two adjacent grids in the second row from the bottom, the detected coordinates of the actual object do not match the center coordinates of the grid, but the coordinates that are output are the center coordinates of the grid. Become. Therefore, it is not possible to output data that accurately indicates the position of the object.

Next, this embodiment will be described based on FIG. 5. (1) and (2) in FIG. 5 are the same as those in FIG. 4 (1) and (2). Then, as shown in Figure 5 (2), when the detected coordinates are converted into a grid expressed by the number of votes, the grid where the number of votes is greater than or equal to a threshold is identified as the grid where the object is present. . At this time, as shown in (3), the number of votes for the upper left grid in the figure is X, which is not greater than the threshold Th, so the coordinates of the grid can be prevented from being output. Therefore, even if noise is included in the sensing information from the sonar 10 as the detected coordinates of an object, it can be removed.

Further, for a grid in which the number of votes exceeds the threshold and it is specified that an object exists, the coordinates detected by the sonar 10 of the object included in the grid are directly output to the control unit 34, instead of the coordinates of the grid. That is, (1) and (2) in FIG. 5 are compared, and the detected coordinates of the object indicated by (1) included in the grid in which the object is identified in (2) are extracted and output. Therefore, in the upper right grid in the figure, if two detected coordinates of an object are included in step (1), those are output as they are as two coordinates. In addition, in the grid at the top right of the figure and the two adjacent grids in the second row from the bottom, the detected coordinates of the actual object do not match the center coordinates of the grid, but the coordinates that do not match are accurate. Output. Therefore, data that accurately indicates the position of the object can be output.

(Other embodiments)
Although the present disclosure has been described based on the embodiments described above, it is not limited to the embodiments, and includes various modifications and modifications within equivalent ranges. In addition, various combinations and configurations, as well as other combinations and configurations that include only one, more, or fewer elements, are within the scope and scope of the present disclosure.

(1) For example, the number of votes and the numerical value of the predetermined threshold value used when detecting an object described in the first embodiment are merely examples, and can be set arbitrarily. For example, although the first number of votes is X and the second number of votes is Y, the first number of votes and the second number of votes may be other values. Further, the first and second voting numbers do not necessarily have to be positive values, and may be set to 0 or negative values. In addition, the threshold value for determining a grid where an object exists may be set to a value that corresponds to the number of votes. For example, if the number of first votes and the number of second votes are set to other values, the threshold value can also be set to a value that corresponds to the number of votes. You can set it accordingly.

In addition, when the same grid contains multiple coordinates detected by the sonar 10, the number of first votes for that number is added to set the number of votes for that grid, but it is not necessary to set the number of first votes for that number. There is no need to add up the votes. In other words, for grids that include multiple detected coordinates, since there is a high probability that an object exists, a predetermined value that is equal to or greater than a threshold is the number of votes, including a method of adding up the number of first votes for the number of detected coordinates. It should be set as .

(2) Furthermore, in the first embodiment, a grid map was created based on the sensing information of the sonar 10 and object detection was performed, but a grid map was created based on the sensing information of other sensors other than the sonar 10. It's okay. Moreover, it is also possible to detect objects by creating a grid map based on the sensing information of each of the various sensors using various types of sensors including the sonar 10 and other sensors.

For example, the object recognition unit 31 creates a grid map showing the detected coordinates of an object based on sensing information from various types of sensors, and then integrates the grid maps to create one grid map showing the detected coordinates of the object. Create. Thereafter, the grid creation unit 32 converts the detected coordinates into a grid expressed by the number of votes. Then, the grid map created by the object recognition unit 31 and the grid map created by the grid creation unit 32 after conversion are compared, and the identification coordinates of the object included in the grid in which the object is identified as being present are output.

The grid created by the object recognition unit 31 used for comparison at this time may be a grid after integration, or a grid map created based on sensing information from various types of sensors. When creating a grid map after integration, it becomes possible to output more accurate detected coordinates of objects based on the characteristics of each of the various types of sensors. In other words, distance measurement sensors such as the Sonar 10 are not good at measuring the width of an object but are good at measuring the distance to the object, and vehicle-mounted cameras are not good at measuring the distance to the object but are good at measuring the distance to the object. is good at measuring. Therefore, sensor fusion can be performed by integrating grid maps created based on sensing information from various types of sensors, and a grid map can be created that takes advantage of the unique characteristics. In this way, by creating an integrated grid map that takes into account the characteristics of many types of sensors, it is possible to increase the accuracy of object detection and eliminate false detections that occur when detecting objects using only one ranging sensor. It becomes possible to suppress the

(3) In the first embodiment, in addition to the main grid corresponding to the detected coordinates of the object indicated by the sensing information of the sonar 10, the number of votes is added for all the surrounding grids located around the main grid. , only some of them may be used. Further, the number of votes may be made different for surrounding grids.

For example, in the case of Sonar 10, the detection accuracy in the width direction is not very high, but the detection accuracy in the distance to the object is high, so when an object is recognized, there is a high possibility that there is nothing closer to the object. . For this reason, the number of votes may not be added to a grid corresponding to a side closer to the sonar 10 than the grid of the detection coordinates among the surrounding grids, or the number of votes may be added to a smaller number than other grids.

(4) In the first embodiment described above, the object detection process shown in FIG. 3 is executed when the starting switch of the host vehicle is turned on, but it is always executed when the starting switch is turned on. It doesn't have to be done. For example, the object detection process may be executed when a switch (not shown) instructing execution of parking assistance control is turned on. In other words, when the execution of control using the result of the object detection process is instructed, the object detection process may also be executed.

(5) Note that the processing unit and its method described in the present disclosure are provided by configuring a processor and memory programmed to execute one or more functions embodied by a computer program. It may also be realized by a dedicated computer. Alternatively, the processing units and techniques described in this disclosure may be implemented by a dedicated computer provided by a processor configured with one or more dedicated hardware logic circuits. Alternatively, the processing unit and the method described in the present disclosure may be implemented by a combination of a processor and memory programmed to perform one or more functions and a processor configured by one or more hardware logic circuits. It may be implemented by one or more dedicated computers configured. The computer program may also be stored as instructions executed by a computer on a computer-readable non-transitory tangible storage medium.

1 Vehicle control system 10 Sonar 30 ECU
31 Object recognition section 32 Grid creation section 33 Output selection section 34 Control section 40 Actuator 50 Display device

Claims (3)

An object detection device that detects objects (100) existing around the own vehicle using a grid map in which the surroundings of the own vehicle (V) are divided into a plurality of grids,
Sensing information of a ranging sensor (10) that measures the position of the object by outputting a search wave and acquiring a reflected wave from the search wave reflected by the object is input, and using the grid map, , an object recognition unit (31) that creates a grid map indicating the detection coordinates of the detected position of the object included in the sensing information of the ranging sensor;
The position of the object existing around the own vehicle is indicated by setting a predetermined number of votes for a main grid that is a grid corresponding to the detected coordinates and a surrounding grid located around the main grid. a grid creation unit (32) that creates a grid map;
A grid in which the number of votes in the grid map created by the grid creation unit exceeds a predetermined threshold is identified as a grid in which the object exists, and the detection coordinates corresponding to the identified grid are used to detect the actual object. It has an object specifying unit (33) that uses coordinates ,
The grid creation unit sets a first number of votes smaller than the threshold for the main grid, and sets a second number of votes smaller than the first number of votes for the surrounding grid, and the object recognition unit In the grid map created by the object detection device, the object detection device sets a voting number larger than the threshold value to the main grid that includes a plurality of the detection coordinates and to the main grids that are adjacent to each other and are continuous. . When a plurality of the detected coordinates are included in the same grid in the grid map created by the object recognition unit, the grid creation unit calculates the number of the detected coordinates included in the grid as the number of votes for the grid. The object detection device according to claim 1 , wherein a value larger than the threshold is set by adding up the first number of votes. The grid creation section is configured to include, in the grid map created by the object recognition section, adjacent and continuous main grids with as many as the number of adjacent main grids based on the first number of votes. The object detection device according to claim 1 or 2 , wherein a numerical value larger than the threshold is set by adding the second number of votes.

Images