JP5703682B2 - Risk calculation device and risk calculation method - Google Patents

Description

  The present invention relates to a risk level calculation device that calculates the risk level of a blind spot for a vehicle driver.

  2. Description of the Related Art Conventionally, an apparatus and a method for avoiding a collision by predicting an object jumping from a place that is a blind spot for a driver of the host vehicle are known. For example, in Patent Document 1, a road width of a blind spot area where a jump is assumed is calculated, and a target that may jump out from the road width is predicted. In addition, a vehicle travel support device and a vehicle travel support method are described in which the estimated movable range of the pop-out target is estimated and the risk of collision is set.

JP 2006-260217 A

  Here, according to the vehicle travel support apparatus as described above, it is possible to calculate the degree of risk for jumping out from a specific blind spot. Here, in the vehicular driving support apparatus as described above, when there are a plurality of blind spots, jumping out of all blind spots is assumed. Therefore, in an environment where there are many blind spots, such as an actual road environment, driving out of many blind spots is assumed, and the driver assistance system that alerts the driver and performs the intervention control operates too much. There was a problem.

  Accordingly, an object of the present invention is to provide a risk level calculation device that can perform an appropriate risk level calculation according to a blind spot structure by selecting a blind spot that sets a risk level based on the blind spot structure. To do.

In order to solve the above-described problem, a risk calculation apparatus according to the present invention includes a blind spot area detection unit that detects a blind spot area for a driver of the host vehicle, and a blind spot configuration that configures a blind spot detected by the blind spot area detection unit. Blind spot composition recognition means for acquiring information of a blind spot composition, which is information of the object and includes any of the height, width, and occupied area ratio of the blind spot composition, and a blind spot composition based on the information of the blind spot composition If the height of the blind spot component is lower than the preset height , the risk setting means for setting the risk of the moving object jumping out from the blind spot area formed by the And a blind spot selecting means for excluding the blind spot component from the risk setting target by the danger setting means when the occupied area ratio is lower than a predetermined occupied area ratio .

  In the risk level calculation device according to the present invention, it is possible to calculate the appropriate risk level according to the blind spot component by setting the risk level of the blind spot area formed by the blind spot component based on the information of the blind spot component. it can. Thus, by setting an appropriate degree of risk for the blind spot area, it is possible to suppress unnecessary operation of the driving support system that alerts the driver and performs intervention control.

  The risk degree calculation apparatus according to the present invention can be excluded from the blind spot for predicting the pop-out when it is determined from the information on the blind spot composition that the blind spot area formed by the blind spot composition is low. As a result, the number of blind spots for which the risk level is calculated can be reduced, so that the calculation load can be reduced.

  Further, it is preferable that the risk level calculation means sets a risk level of the blind spot area based on a height of the blind spot component.

  Since the size of the blind spot area formed by the blind spot structure varies depending on the height of the blind spot structure, whether or not a moving object existing behind the blind spot structure can be recognized changes. For this reason, it is possible to set an appropriate risk level by setting the risk level of the blind spot area based on the height of the blind spot component.

  Furthermore, it is preferable that the risk level calculation means sets the risk level of the blind spot area formed by the blind spot component to be higher as the height of the blind spot component is higher.

  The higher the blind spot component height, the larger the blind spot area at the place where the object is installed, and the number of moving objects that cannot be recognized increases. For this reason, the higher the height of the blind spot component, the higher the risk of the blind spot region formed by the blind spot component, thereby appropriately setting the risk level for the blind spot region.

  Moreover, it is preferable that the risk level calculation means sets the risk level of the blind spot area based on the width of the blind spot structure.

  Since the size of the blind spot area formed by the blind spot structure varies depending on the width of the blind spot structure, whether or not a moving object existing behind the blind spot structure can be recognized changes. For this reason, an appropriate risk level can be set by setting the risk level of the blind spot area based on the width of the blind spot component.

  Furthermore, it is preferable that the risk level calculation means sets the risk level of the blind spot area formed by the blind spot component higher as the width of the blind spot component increases.

  The wider the blind spot component, the larger the blind spot area at the place where the object is installed, and the number of moving objects that cannot be recognized increases. Therefore, the risk level for the blind spot area can be appropriately set by setting the risk level of the blind spot area formed by the blind spot structure higher as the width of the blind spot structure is wider.

  The risk calculation method according to the present invention also recognizes a blind spot area detecting step for detecting an area that becomes a blind spot for the driver of the host vehicle and information on a blind spot component that constitutes the blind spot detected by the blind spot area detecting means. A blind spot component recognizing step, and a risk level setting step of setting a risk level of a moving object popping out from a blind spot area formed by the blind spot component based on information on the blind spot component.

  In the risk level calculation method according to the present invention, it is possible to perform an appropriate risk level calculation according to the blind spot component by setting the risk level of the blind spot area formed by the blind spot component based on the information of the blind spot component. it can. This suppresses unnecessary operation of the driving support system that alerts the driver and performs intervention control.

  ADVANTAGE OF THE INVENTION According to this invention, the danger level calculation apparatus which can perform a highly accurate risk level calculation by selecting a blind spot based on a blind spot structure can be provided.

It is a block block diagram of the risk degree calculation apparatus which concerns on embodiment of this invention. It is a flowchart about operation | movement of the risk calculation apparatus which concerns on embodiment of this invention. It is a figure which shows the example of the image of a blind spot structure for demonstrating an occupation area rate. It is a flowchart about the blind spot selection process which concerns on embodiment of this invention. It is a graph which shows the relationship between the height of a blind spot component, and a danger level.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, the same or equivalent elements will be denoted by the same reference numerals, and redundant description will be omitted.

  FIG. 1 is a block diagram showing a risk degree calculation apparatus according to an embodiment of the present invention. As shown in FIG. 1, the risk degree calculation device 1 of this embodiment includes an ECU 2, a camera 11, a radar 12, a wheel speed sensor 13, a yaw rate sensor 14, a rudder angle sensor 15, a display device 31, a speaker 32, and an actuator 33. I have. This risk level calculation device 1 is mounted on an automobile or the like.

  The ECU 2 is connected with a sensor for measuring the surrounding environment including the camera 11 and the radar 12.

  The camera 11 is a device that images the periphery of the host vehicle. The periphery of the host vehicle is at least the front, and the side and the rear are imaged as necessary. The camera 11 images the periphery of the host vehicle, and transmits data of the captured image to the ECU 2 as an image signal.

  The radar 12 is a device for detecting objects around the host vehicle. The periphery of the host vehicle is at least the front, and also detects the side and rear objects as necessary. Examples of the radar 12 include a laser radar and a millimeter wave radar. The radar 12 transmits an electromagnetic wave while scanning in a horizontal plane, receives a reflected wave that returns after being reflected by an object, and detects information related to the transmission and reception. The radar 12 transmits the detected transmission / reception information to the ECU 2 as a radar signal.

  Furthermore, a wheel speed sensor 13, a yaw rate sensor 14, and a steering angle sensor 15 are connected to the ECU 2. The wheel speed sensor 13 is a sensor that detects the rotational speed of the wheel of the host vehicle. The wheel speed sensor 13 transmits the detected rotation speed of the wheel to the ECU 2 as a wheel speed signal. The yaw rate sensor 14 is a sensor that detects the yaw rate of the host vehicle. The yaw rate sensor 14 transmits the detected yaw rate to the ECU 2 as a yaw rate signal. The steering angle sensor 15 is a sensor that detects the steering angle of the host vehicle. For example, the steering angle sensor 4 detects the steering angle of the host vehicle by detecting the rotation angle (steering angle) of the steering shaft. The steering angle sensor 15 transmits the detected steering angle to the ECU 2 as a steering angle signal.

  The display device 31 is a display installed in the vehicle, displays various information according to the driving support signal output from the ECU 2, and notifies the driver. The speaker 32 outputs a predetermined sound according to the driving support signal from the ECU 2. Thus, the display 31 and the speaker 32 perform screen display and audio output as HMI (Human Machine Interface).

  The actuator 33 is a brake actuator or accelerator actuator that causes the ECU 2 to intervene in the driving operation of the driver based on the driving support signal to drive the brake or accelerator of the host vehicle.

  The ECU 2 includes a CPU [Central Processing Unit], various memories, and the like, and performs overall control of the risk degree calculation device 1. The ECU 2 loads each application program stored in the memory and executes it by the CPU, thereby performing a blind spot area detection unit 21, a blind spot component recognition unit 22, a blind spot selection unit 23, a blind spot risk setting unit 24, and a collision probability calculation. The unit 25 and the driving support control unit 26 are configured. In the present embodiment, the blind spot area detection unit 21 is a blind spot area detection unit described in the claims, the blind spot component recognition unit 22 is a blind spot component recognition unit, and the blind spot selection unit 23 is a blind spot selection unit. The blind spot risk setting unit 24 corresponds to a risk setting means.

  The blind spot area detection unit 21 detects an object (hereinafter referred to as a blind spot constituent in this specification) that causes a blind spot on the basis of image information from the camera 11 and radar information from the radar 12 at regular intervals. The blind spot information formed by the component is acquired. For example, the blind spot information includes a blind spot area, a blind spot position, and the number of blind spots, and all or a part of these are acquired. The blind spot information detected by the blind spot area detection unit 21 is transmitted to the blind spot component recognition unit 22. The detection of the blind spot may be performed using both the camera 11 and the radar 12, or the detection result of one of the sensors having high detection performance may be used. The blind spot can be detected by applying a conventionally known method.

  The blind spot composition recognizing unit 22 acquires information on the detected blind spot composition. The information of the blind spot composition includes, for example, the position, size (lateral length, depth length, height), shape, type (wall, pillar, tree, vehicle, etc.), occupation of the blind spot composition. There are the area ratio, the distance to the blind spot composition, the direction in which the blind spot composition exists, and the relative speed, and all or some of these are acquired. Information on the blind spot composition acquired by the blind spot composition recognition unit 22 is transmitted to the blind spot sorting unit 23.

  The blind spot selection unit 23 determines whether there is a possibility that a moving object exists in the blind spot area formed by the constituent and unexpectedly pops out based on the information of the blind spot constituent. If it is determined that there is no moving object in the blind spot area formed by the blind spot structure, or if it is determined that the possibility of the moving object popping out is extremely low, the blind spot area is excluded from the predicted blind spot. The control signal is transmitted to the blind spot risk level setting unit 24.

  The blind spot risk level setting unit 24 sets the risk level of a moving object that may exist in a blind spot area formed by the blind spot component based on the information of the blind spot component. In addition, when a control signal that is excluded from the blind spot predicted by popping out is transmitted from the blind spot selector 23, the risk level of the blind spot area is not set.

  The collision probability calculation unit 25 calculates the collision probability between the host vehicle and the moving object predicted to jump out based on the risk set in the blind spot risk setting unit 24. The collision probability calculation unit 25 predicts the future position of the host vehicle after a predetermined time based on the measurement results of the wheel speed sensor 13, the yaw rate sensor 14, and the rudder angle sensor 15, and the moving speed of the moving object popping out from the blind spot area. The probability that the own vehicle and the moving object collide is calculated by predicting the future position of the moving object after a predetermined time from the moving direction. The calculation of the collision probability can be performed by applying a conventionally known method.

  The driving support control unit 26 performs control for avoiding a collision based on the collision probability calculated by the collision probability calculation unit 25. For example, when it is determined that the collision probability between the own vehicle and a moving object popping out from the blind spot area is high, the driver is alerted by the display device 31 or the speaker 32, or the brake control signal or the engine control signal is sent to the actuator. By performing intervention control by transmitting to 33, a collision is avoided.

  Next, the operation of the risk degree calculation apparatus according to this embodiment will be described.

  FIG. 2 is a flowchart of the operation of the risk level calculation apparatus according to the present embodiment. The series of processes shown in FIG. 2 is repeatedly executed at a predetermined cycle set in advance in the obstacle recognition ECU 2, for example.

  When the driving support process is started, detection results of various sensors mounted on the host vehicle such as the camera 11, the radar 12, the wheel speed sensor 13, the yaw rate sensor 14, and the rudder angle sensor 15 are read (S1). The read measurement value is stored in the RAM in the ECU 2.

  Subsequently, a blind spot existing around the vehicle is detected based on the image information and the radar information (S2). An index (i = 1, 2,..., N) is set and stored for the detected blind spot (S3). Thereafter, the index i is set to 1 (S4), and information on the blind spot structure corresponding to the index is acquired (S5). As information on the blind spot composition, for example, the size (width, height, depth) of the blind spot composition, the type of the blind spot composition, and the occupation area ratio of the blind spot composition are acquired.

  The size of the blind spot component is calculated based on, for example, the distance between the host vehicle and the blind spot component acquired by the radar 12 and the area occupied by the blind spot component in the image. Also, the type of blind spot component is recognized by pattern matching using a template, for example.

  The occupation area ratio of the blind spot component will be described with reference to FIG. FIG. 3 is an image obtained by cutting out the minimum rectangular area surrounding the blind spot component from the captured image of the camera 11. The shaded area in FIG. 3 shows the background area other than the blind spot structure. The area occupied by the blind spot composition is the ratio of the range occupied by the blind spot composition to the area of the cut image. The occupied area ratio is calculated, for example, by binarizing a blind spot component and a region other than the blind spot component from the clipped image, and dividing the number of pixels.

  Subsequently, a blind spot selection process is performed (S6). The blind spot selection process will be described with reference to FIG. In the blind spot selection process, it is first determined whether or not the detected blind spot component is a wall (S21). If it is determined that the structure is a wall, it is determined whether the height of the wall is equal to or greater than a set value Hw (for example, 1 m) (S22). If the height of the wall is equal to or greater than the set value Hw, the blind spot area with index i is set as a blind spot area for setting the risk level (S30). On the other hand, in S22, if the height of the wall is lower than the set value Hw, it is unlikely that the moving object will jump out of the blind spot area, so the blind spot area of index i is excluded from the blind spot area where the risk level is set. (S31).

  When it is determined in S21 that the blind spot structure is not a wall, it is determined whether or not the detected blind spot structure is a pillar (S23). If it is determined that the structure is a pillar, it is determined whether or not the width of the pillar is greater than or equal to a set value Wc (for example, 0.5 m) (S24). If the column width is equal to or greater than the set value Wc, the blind spot area with index i is set as a blind spot area for which the risk level is set (S30). On the other hand, in S24, when the column width is narrower than the set value Wc, it is considered that there is a low possibility that a moving object is present in the blind spot area. Therefore, the blind spot area of index i is a blind spot area in which the risk level is set. (S31).

  In S23, when it is determined that the blind spot structure is not a pillar, it is determined whether or not the detected blind spot structure is a tree (S25). If it is determined that the structure is a tree, it is determined whether the occupied area ratio of the tree is equal to or greater than a set value At (for example, 60%) (S26). When the occupation area ratio of the tree is equal to or greater than the set value At, the blind spot area with index i is set as a blind spot area for which the risk level is set (S30). On the other hand, in S26, when the occupied area ratio of the tree is lower than the set value At, it is predicted that the own vehicle can be seen from the moving object even if the moving object exists in the blind spot area. Considering that the pop-out probability is low, the blind spot area of index i is excluded from the blind spot area where the risk level is set (S31).

  When it is determined in S25 that the blind spot structure is not a tree, it is determined whether or not the detected blind spot structure is a vehicle (S27). If it is determined that the structure is a vehicle, it is determined whether or not the height of the vehicle is greater than or equal to a set value Hc (for example, 1 m) (S28). When the height of the vehicle is equal to or greater than the set value Hc, it is determined whether or not the occupied area ratio of the vehicle is equal to or greater than the set value Ac (for example, 60%) (S29). If the height of the tree is equal to or greater than the set value Hc and the occupation area ratio is equal to or greater than the set value Ac, the blind spot area with index i is set as a blind spot area for which the risk level is set (S30). On the other hand, in S28, when the vehicle height is lower than the set value Hc or when the vehicle occupation area ratio is smaller than the set value Ac, it is considered that there is a low possibility that a moving object will jump out of the blind spot area, and the index i. The blind spot area is excluded from the blind spot area where the risk level is set (S31).

  If it is determined in S27 that the blind spot structure is not a vehicle, the blind spot area with index i is set as a blind spot area for which the risk level is set (S30).

  When S30 or S31 ends, it returns to FIG. 2, and it is determined whether or not the risk level is set for the blind spot area of index i (S7). If the blind spot area of index i is excluded from the blind spot area where the risk level is set in S7, the process proceeds to S10. On the other hand, when the blind spot area of index i is set as a blind spot area for which the risk level is set, a risk level setting process is performed (S8). The risk level is set based on information such as the position and size of the blind spot area and the type of blind spot structure. This risk level is determined by referring to a risk level determination table stored in the ECU 2 in advance. For example, based on near-miss cases that occurred in the past, the degree-of-risk determination table quantifies and stores the degree of risk that a moving object will pop out depending on the position and size of the blind spot area, the type of blind spot structure, and the like. The determined risk is changed based on the information of the blind spot structure recognized in S5. Specifically, the degree of risk is changed based on the height and width of the blind spot structure.

  FIG. 5 is a graph showing the relationship between the height of the wall, which is an example of the blind spot structure, and the degree of danger. In this graph, the horizontal axis is the height of the wall, and the vertical axis is the risk level. The blind spot risk level setting unit 24 sets the risk level to a high value when the height is high, and sets the risk level to a low value as the height is low, according to the height of the blind spot component. For example, in the range where the height of the wall is 0 to Hw (G1), the blind spot area formed by the wall by the blind spot selector 23 is set as a blind spot where the risk level is not set, and the calculation of the risk level is omitted. In the range where the wall height is Hw to Hm (G2), the higher the wall height, the higher the degree of danger. When the height of the wall exceeds the set value Hm (G3), the pop-out probability of the moving object existing in the blind spot area formed by the wall does not change and is set to a constant value. Here, the set values Hw and Hm are values set in the ECU 2 in advance. This value is set based on past near-miss cases. Note that the height of the blind spot component is the maximum height of the blind spot component.

  The relationship between the width of the pillar that is the blind spot component and the pop-out probability is the same as that in FIG. That is, the blind spot risk level setting unit 24 sets the risk level to a high value when the width is wide, and sets the risk level to a low value as the width is narrow, according to the width of the column. Further, in the range of the column width from 0 to Wc, the blind spot area formed by the pillar by the blind spot selector 23 is set as a blind spot where the risk level is not set, and the calculation of the risk level is omitted. In the range where the column width is Wc to Wm, the higher the column width, the higher the degree of danger. When the column width exceeds the set value Wm, the pop-out probability of the moving object existing in the blind spot area formed by the column does not change and is set to a constant value. Here, the set values Ww and Wm are values set in the ECU 2 in advance. This value is set based on past near-miss cases. The width of the blind spot component is the maximum width of the blind spot component.

  When the prediction of the degree of risk from the blind spot area is completed, the collision probability between the moving object popping out from the blind spot area and the host vehicle is calculated (S9). Based on the measurement results of the wheel speed sensor 13, the yaw rate sensor 14, and the rudder angle sensor 15 acquired in S <b> 1, the future position of the host vehicle after a predetermined time is predicted, and the moving object that jumps out from the blind spot area is after the predetermined time By predicting the future position, the probability that the host vehicle collides with a moving object is calculated.

  Next, the index i is incremented (+1) (S10), and the detected number n of blind spots is compared with the index i (S11). If the index i is equal to or less than the number n of detected blind spots, the process proceeds to S5, and the same process is executed for all detected blind spots. On the other hand, when the index i is larger than the number n of detected blind spots, the risk level calculation process is terminated. The calculated risk level is transmitted to the driving support control unit 26. For example, when the risk level is higher than a predetermined value, a control signal such as alerting the driver or intervention control is generated and driving support is performed. Prevent collisions with moving objects that pop out of the blind spot.

  As described above, in the risk level calculation apparatus according to the present embodiment, the risk level of the blind spot area formed by the blind spot component is set based on the information of the blind spot component. When it is judged from the information of the objects that make up the blind spot that the object is unlikely to jump out of the blind spot area, the driver assistance system that alerts the driver and performs intervention control is activated by reducing the risk level. Therefore, it is possible to calculate the risk level with high accuracy.

  In the conventional risk level calculation device, when there are a plurality of blind spot areas, the risk levels are calculated for all the blind spot areas, and thus there is a problem that real-time processing is difficult. On the other hand, the risk level calculation apparatus according to the present invention can exclude a blind spot area that is determined to have a low risk level based on the information of the blind spot structure from the blind spot that predicts popping out. As a result, the number of blind spots for which the risk level is calculated can be reduced, so that the calculation load can be reduced.

  In addition, although preferred embodiment of this invention was described, this invention is not limited to the said embodiment. For example, in the present embodiment, in the blind spot selection process, a determination is made as to whether or not to calculate the risk level by comparing the height, width, and occupation area ratio of a wall, a pillar, a tree, and a vehicle with a set value. Yes. However, the blind spot structures to be compared are not limited to walls, pillars, trees, and vehicles, and the judgment conditions are not limited to the height, width, and occupied area ratio of the blind spot components. For example, when the number of light vehicles is a certain number or less, it may be excluded from the risk calculation target. As described above, the blind spot composition other than the wall, the pillar, the tree, and the vehicle and the conditions thereof may be set and stored in advance, and the blind spot selection process may be performed.

DESCRIPTION OF SYMBOLS 1 ... Risk level calculation apparatus, 2 ... ECU, 11 ... Camera, 12 ... Radar, 13 ... Wheel speed sensor, 14 ... Yaw rate sensor, 15 ... Steering angle sensor, 21 ... Blind spot area detection part, 22 ... Blind spot structure recognition part , 23 ... Blind spot selection unit, 24 ... Blind spot risk setting unit, 25 ... Collision probability calculation unit, 26 ... Driving support control unit, 31 ... Display device, 32 ... Speaker, 33 ... Actuator.

Claims (6)

A blind spot area detecting means for detecting a blind spot area for the driver of the host vehicle;
A blind spot that is information of a blind spot constituting the blind spot detected by the blind spot area detecting means, and that acquires information on the blind spot constituent including any of a height, a width, and an occupied area ratio of the blind spot constituent. A component recognition means;
A degree-of-risk setting means for setting a degree of risk of a moving object popping out from a blind spot area formed by the blind spot component based on the information of the blind spot component;
When the height of the blind spot component is lower than a preset height, when the width of the blind spot component is narrower than a preset width, or the occupation area ratio is more than a predetermined occupation area ratio When it is low, the risk calculating device includes a blind spot selecting unit that excludes the blind spot component from a risk setting target by the risk setting unit.   The risk level calculation device according to claim 1, wherein the risk level setting unit sets a risk level of a blind spot area formed by the blind spot component based on a height of the blind spot component.   The risk level calculation device according to claim 2, wherein the risk level setting unit sets a higher risk level of a blind spot area formed by the blind spot component as the height of the blind spot component increases.   The risk level calculation device according to any one of claims 1 to 3, wherein the risk level setting unit sets a risk level of a blind spot area formed by the blind spot component based on a width of the blind spot component.   The risk level calculation device according to claim 4, wherein the risk level setting unit sets a higher risk level of a blind spot area formed by the blind spot component as the width of the blind spot component increases. A risk calculation method executed by the risk calculation device according to any one of claims 1 to 5 ,
A blind spot area detecting step in which the blind spot area detecting means detects a blind spot area for a driver of the host vehicle;
The blind spot component recognizing means is information of a blind spot component constituting the blind spot detected in the blind spot area detecting step, and includes the height, width, and occupied area ratio of the blind spot component. A blind spot composition recognition step for acquiring information on the composition,
A risk level setting step in which the risk level setting means sets a risk level of a moving object popping out from a blind spot area formed by the blind spot component based on the information of the blind spot component;
When the blind spot selection means has a height of the blind spot component lower than a preset height, a width of the blind spot constituent is narrower than a preset width, or an area occupied by the blind spot constituent When the rate is lower than a predetermined occupation area rate, the blind spot selection step of excluding the blind spot region formed by the blind spot component from the risk setting target in the risk setting step, Calculation method.

Images