JP2012008718A - Obstacle avoiding apparatus - Google Patents

Abstract

An obstacle avoidance device that operates a collision avoidance system at an appropriate timing in accordance with the recognized obstacle and / or the reliability of the recognition while accurately recognizing an obstacle around the host vehicle. provide.
An obstacle avoidance device according to the present invention includes a plurality of detection means for detecting an obstacle, and information detected by the plurality of detection means based on information about the obstacle detected by the plurality of detection means. Correlation determination means for determining whether or not the detection result has correlation, and obstacle avoidance control means for controlling the collision avoidance system according to the determination result determined by the correlation determination means. The avoidance control unit operates the collision avoidance system at a timing earlier than the timing when the avoidance control unit determines that there is a correlation when the correlation determination unit determines that there is no correlation.
[Selection] Figure 1

Description

  The present invention relates to an obstacle avoidance device that avoids or reduces collision damage with obstacles around the host vehicle, and more specifically, detects obstacles around the host vehicle and detects the detected obstacles. The present invention relates to an obstacle avoidance device that avoids collision or reduces collision damage according to the above.

  2. Description of the Related Art Conventionally, vehicles have been equipped with a pre-crash safety system (PCS), and a technique for detecting an obstacle around the host vehicle and avoiding or mitigating a collision with the obstacle has been widely used.

  Examples of techniques for detecting obstacles around the host vehicle include image recognition and radar recognition. In the image recognition technique, an image around the host vehicle is acquired by a camera mounted on the host vehicle, and a preset reference pattern (for example, a vehicle and a pedestrian) is searched from the acquired image. The radar recognition technology detects an obstacle around the own vehicle by a laser radar and a millimeter wave radar mounted on the own vehicle, and the obstacle is a vehicle according to the intensity of a reflected wave from the obstacle. Or whether it is a pedestrian.

  Furthermore, a technique combining the above-described image recognition technique and radar recognition technique has also been proposed (for example, Patent Document 1). In the technique described in Patent Document 1, first, the type of an obstacle is determined based on an image captured by a camera mounted on the host vehicle. The reflection intensity from the obstacle to be detected by the radar is set according to the determined obstacle type.

  Thus, conventionally, recognition accuracy of obstacles around the host vehicle is improved by using image recognition technology and / or radar recognition technology.

JP 2009-174900 A

  However, if both the image recognition technique and the radar recognition technique are used as in the technique described in Patent Document 1 described above, the recognition accuracy itself of the obstacle around the host vehicle is improved, but the recognized obstacle and Depending on the reliability of the recognition, the collision avoidance system cannot be operated at an appropriate timing.

  Therefore, an object of the present invention is to operate the collision avoidance system at an appropriate timing according to the recognized obstacle and / or the reliability of the recognition while accurately recognizing the obstacle around the own vehicle. An obstacle avoidance device is provided.

In order to achieve the above object, an obstacle avoidance apparatus according to the present invention is an obstacle avoidance apparatus that detects an obstacle around a host vehicle and operates a collision avoidance system that avoids a collision with the obstacle. Based on the information on the obstacles detected by the plurality of detection means and the plurality of detection means, it is determined whether or not the detection results detected by the plurality of detection means are correlated. Correlation determining means, and obstacle avoidance control means for controlling the collision avoidance system according to the determination result determined by the correlation determining means. The obstacle avoidance control means is correlated with the correlation determining means. When it is determined that there is a correlation, the collision avoidance system is operated at a timing earlier than the operation timing when it is determined that there is no correlation.
With this configuration, the plurality of detection units detect obstacles, and the correlation determination unit determines whether or not each detection result detected by the plurality of detection units has a correlation. The obstacle avoidance control means can operate the collision avoidance system at an appropriate timing according to the recognized obstacle and / or the reliability of the recognition. it can.

Preferably, the plurality of detection means includes a radar detection means for detecting an obstacle via a radar, and an image detection means for acquiring an image around the host vehicle by a camera and detecting the obstacle from the image. To do.
With this configuration, the radar detection unit detects an obstacle via the radar, and the image detection unit acquires an image around the host vehicle by the camera and detects the obstacle from the image. It can be recognized accurately, and the reliability of the obstacle recognition is improved.

More preferably, the correlation determination unit includes a radar detection result determination unit that determines the type of an obstacle based on the radar reflection intensity in the radar detection unit, and a predetermined predetermined value for the image acquired by the image detection unit. Image detection result determination means for determining the type of obstacle by matching the image pattern, obstacle type determined by the radar detection result determination means, and obstacle type determined by the image detection result determination means And obstacle estimation means for estimating an obstacle based on the above.
With this configuration, the radar detection result determination means determines the type of obstacle based on the radar reflection intensity in the radar detection means, and the image detection result determination means is preset for the image acquired by the image detection means. The obstacle type is determined by matching the predetermined image pattern, and the obstacle estimation means estimates the obstacle based on the determined obstacle type. And the reliability of the obstacle recognition is further improved.

A more preferable obstacle avoidance control means is estimated by the obstacle estimation means when the obstacle type determined by the radar detection result determination means is the same as the obstacle type determined by the image detection result determination means. The collision avoidance system that operates in response to the obstacle that has been made is operated at a timing earlier than the operation timing when the type of the obstacle is not the same.
With this configuration, the obstacle avoidance control unit is configured to determine whether the obstacle type determined by the radar detection result determination unit is the same as the obstacle type determined by the image detection result determination unit. In order to control the operation timing of the collision avoidance system that operates in response to the obstacle estimated by the object estimation means, collision avoidance at an appropriate timing according to the recognized obstacle and / or the reliability of the recognition. The system can be activated.

More preferably, the obstacle avoidance control unit is configured such that the obstacle type determined by the radar detection result determination unit is the same as the obstacle type determined by the image detection result determination unit, and the obstacle is set in advance. A collision avoidance system that operates in response to the obstacle estimated by the obstacle estimation means when the obstacle is a predetermined obstacle set in advance. It is characterized by operating at an earlier timing than when it is not specified.
With this configuration, the obstacle avoidance control unit operates in response to the obstacle estimated by the obstacle estimation unit depending on whether the obstacle is specified as a predetermined obstacle set in advance. Therefore, the collision avoidance system can be operated at a more appropriate timing according to the recognized obstacle and / or the reliability of the recognition.

The obstacle avoidance device of the present invention further comprises recording means in which a look-up table indicating a collision avoidance system that operates in response to an obstacle is recorded in advance, and the obstacle avoidance control means is included in the recording means. The collision avoidance system is controlled by referring to the recorded lookup table and according to the determination result determined by the correlation determination means.
With such a configuration, a lookup table indicating a collision avoidance system that operates in response to an obstacle is recorded in advance in the recording means, and the obstacle avoidance control means refers to the lookup table and the collision avoidance system Therefore, it is possible to operate an optimum collision avoidance system according to any situation among various types of collision avoidance systems.

In order to achieve the above object, the obstacle avoidance method of the present invention includes an obstacle avoidance device that detects an obstacle around the host vehicle and operates a collision avoidance system that avoids a collision with the obstacle. Obstacle avoidance method to be executed, based on a plurality of detection steps for detecting an obstacle and information on obstacles detected at the plurality of detection steps, to each detection result detected at the plurality of detection steps. An obstacle avoidance control step comprising: a correlation determination step for determining whether or not there is a correlation; and an obstacle avoidance control step for controlling the collision avoidance system according to the determination result determined in the correlation determination step. Then, when it is determined that there is a correlation in the correlation determination step, the collision avoidance system is operated at an earlier timing than the timing when it is determined that there is no correlation. And characterized in that.
With this configuration, an obstacle around the host vehicle is detected in order to detect an obstacle in a plurality of detection steps and to determine whether or not each detection result detected in the plurality of detection steps is correlated in the correlation determination step. An object can be recognized with high accuracy, and in the obstacle avoidance control step, the collision avoidance system can be operated at an appropriate timing according to the recognized obstacle and / or the reliability of the recognition.

  Moreover, in order to achieve the said objective, each process which each structure of the obstacle avoidance apparatus of this invention mentioned above can be regarded as the obstacle avoidance method which gives a series of process procedures. This method is provided in the form of a program for causing a computer to execute a series of processing procedures. This program may be installed in a computer in a form recorded on a computer-readable recording medium.

  As described above, according to the obstacle avoidance device of the present invention, while recognizing obstacles around the host vehicle with high accuracy, the obstacle avoidance device can be appropriately selected according to the recognized obstacle and / or the reliability of the recognition. The collision avoidance system can be activated at the timing.

1 is a functional block diagram showing an obstacle avoidance device 100 according to an embodiment of the present invention. The flowchart which shows the flow of a process of the obstacle avoidance method 200 which the obstacle avoidance apparatus 100 which concerns on one Embodiment of this invention performs. Diagram showing the relationship between the distance from the vehicle to the target and the intensity of the radar reflected wave The information related to the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220 and the information related to the obstacle acquired in the image acquisition step 230 and the image determination step 240 are fused (fusion). Showing an obstacle type estimation method 400 for estimating the type of an obstacle The figure which shows the timing which operates the collision avoidance system 160 in a pedestrian (1st estimation result 431) and a low threshold stationary target (2nd estimation result 432). The figure which shows the timing (it respond | corresponds to the speed of the own vehicle) which operates the collision avoidance system 160 in a stationary vehicle (3rd estimation result 441) and a high threshold stationary target (4th estimation result 442). The figure which shows the timing which operates the collision avoidance system 160 in a preceding vehicle (5th estimation result 451) and an oncoming vehicle (6th estimation result 452). The figure which shows the timing (it respond | corresponds to the lateral position required for collision avoidance) which operates the collision avoidance system 160 in a stationary vehicle (3rd estimation result 441) and a high threshold stationary target (4th estimation result 442). The figure which shows the collision avoidance system operated according to the obstruction estimated by the obstruction estimation means 133 and / or the reliability of the said estimation

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a functional block diagram showing an obstacle avoidance device 100 according to an embodiment of the present invention. In FIG. 1, the obstacle avoidance device 100 includes radar detection means 110, image acquisition means 120, correlation determination means 130, recording means 140, obstacle avoidance control means 150, and collision avoidance system 160. . The correlation determination unit 130 includes a radar detection result determination unit 131, an image determination unit 132, and an obstacle estimation unit 133.

  The radar detection unit 110 detects an obstacle around the host vehicle via the radar. Specifically, the radar detection unit 110 is a laser radar and a millimeter wave radar mounted on the host vehicle, and detects a reflected wave from an obstacle around the host vehicle.

  The image acquisition unit 120 is a camera mounted on the host vehicle, and captures an image around the host vehicle using the camera.

  Correlation determining means 130 correlates to information related to the obstacle based on information related to the obstacle detected by radar detecting means 110 and information related to the obstacle recognized from the image acquired by image acquiring means 120, respectively. It is determined whether there is sex.

  More specifically, the radar detection result determination unit 131 calculates the relative speed between the host vehicle and the obstacle based on the reflected wave from the obstacle around the host vehicle detected by the radar detection unit 110. Then, based on the calculated relative speed, the radar detection result determination unit 131 determines whether the obstacle is a stationary target, a preceding vehicle traveling in the same direction as the own vehicle, or traveling in a direction opposite to the own vehicle. Determine if it is an oncoming vehicle.

  Furthermore, the radar detection result determination unit 131 determines whether the obstacle is a pedestrian or a vehicle based on the intensity of the reflected wave from the obstacle around the host vehicle detected by the radar detection unit 110.

  The image determination unit 132 searches the reference pattern set in advance for the image around the host vehicle acquired by the image acquisition unit 120, and when there is an obstacle around the host vehicle, the obstacle A person or a vehicle. Specifically, a pedestrian pattern and a vehicle pattern are set in advance as a reference pattern, and the pedestrian pattern and the vehicle pattern are matched in order in the image area.

  The obstacle estimation unit 133 estimates an obstacle around the host vehicle based on the result determined by the radar detection result determination unit 131 and the result determined by the image determination unit 132. Specifically, it is determined whether or not the obstacle type determined by the radar detection result determination unit 131 matches the obstacle type determined by the image determination unit 132, and the obstacle type is determined. Is estimated.

  For example, when the obstacles determined by the radar detection result determination unit 131 and the image determination unit 132 are the same for pedestrians (when there is a correlation as a pedestrian), the obstacle estimation unit 133 determines that the obstacles are Estimated to be a pedestrian. Similarly, when the obstacles determined by the radar detection result determination unit 131 and the image determination unit 132 coincide with each other in the vehicle (when there is a correlation as a vehicle), the obstacle estimation unit 133 determines that the obstacle is a vehicle. It is estimated that. When the obstacle is a vehicle, it is more preferable if it can be estimated whether the vehicle is a stationary vehicle, a preceding vehicle, or an oncoming vehicle.

  On the other hand, when the type of obstacle determined by the radar detection result determination unit 131 and the type of obstacle determined by the image determination unit 132 do not match or when the determination result is unclear (no correlation) ), The obstacle estimation means 133 may not estimate that the obstacle is a pedestrian or a vehicle. Even if the obstacle is not determined by the image determination unit 132 as a pedestrian or a vehicle, the obstacle estimation unit 133 may, for example, calculate the difference between the vehicle and the obstacle calculated by the radar detection result determination unit 131. Based on the relative speed, the obstacle is estimated to be a stationary target. In other words, the obstacle estimation unit 133 may be configured to estimate within an estimable range based on the result determined by the radar detection result determination unit 131 and the result determined by the image determination unit 132. I do not care.

  Information relating to the collision avoidance system 160 that operates in response to the obstacle estimated by the obstacle estimating means 133 is recorded in the recording means 140 in advance as a lookup table. For example, when the obstacle is a pedestrian or a vehicle, the timing of operating the collision avoidance system 160 and the specific collision avoidance system 160 to be operated are associated. Here, the collision avoidance system 160 is, for example, a PCS, and includes a brake, a seat belt, a driver's seat airbag, a pedestrian protection hood airbag, and the like of the host vehicle.

  The obstacle avoidance control unit 150 controls the collision avoidance system 160 according to the determination result determined by the correlation determination unit 130. More specifically, the obstacle avoidance control unit 150 refers to the lookup table recorded in the recording unit 140 and operates the collision avoidance system 160 corresponding to the obstacle estimated by the obstacle estimation unit 133.

  Next, the flow of processing will be described in detail for the obstacle avoidance method executed by the obstacle avoidance apparatus 100 according to an embodiment of the present invention. FIG. 2 is a flowchart showing a process flow of the obstacle avoidance method 200 executed by the obstacle avoidance apparatus 100 according to the embodiment of the present invention. 2, the obstacle avoidance method 200 includes a radar detection step 210, a radar detection result determination step 220, an image acquisition step 230, an image determination step 240, an obstacle estimation step 250, and a collision avoidance step 260. Including.

  In the radar detection step 210, the radar detection means 110 detects an obstacle around the host vehicle via the radar. Specifically, the radar detection unit 110 detects a reflected wave from an obstacle around the host vehicle.

  In the radar detection result determination step 220, the radar detection result determination means 131 calculates the relative speed Vr between the host vehicle and the obstacle based on the reflected wave detected in the radar detection step 210. The radar detection result determination unit 131 determines whether the obstacle is a stationary target or a preceding vehicle traveling in the same direction as the own vehicle based on the speed V0 of the own vehicle and the calculated relative speed Vr. It is determined whether the vehicle is an oncoming vehicle traveling in the opposite direction to the host vehicle.

  First, when the absolute value (| V0−Vr |) of the difference between the speed V0 of the host vehicle and the relative speed Vr is less than a predetermined value (ΔV), it is determined that the obstacle is a stationary target (| V0− Vr | <ΔV). On the other hand, when the absolute value (| V0−Vr |) of the difference between the speed V0 of the host vehicle and the relative speed Vr is equal to or greater than a predetermined value (ΔV), it is determined that the obstacle is a moving target (| V0− Vr | ≧ ΔV). Here, the predetermined value (ΔV) is set in consideration of the speed of the pedestrian so that it can be determined as a stationary target even if the pedestrian is moving.

  When the relative speed Vr is positive, the obstacle is approaching the host vehicle, and when the host vehicle is traveling and the relative speed Vr is high (for example, the host vehicle Speed V0 = 50 km / hr, relative speed Vr = 100 km / hr), it can be determined that the obstacle is an oncoming vehicle.

  Next, the radar detection result determination unit 131 determines whether the obstacle is a pedestrian or a vehicle based on the intensity of the reflected wave detected in the radar detection step 210. FIG. 3 is a diagram illustrating the relationship between the distance from the host vehicle to the target and the intensity of the radar reflected wave. In FIG. 3, a high threshold 310 and a low threshold 320 are set for the intensity of the radar reflected wave corresponding to the distance from the host vehicle to the target. When the intensity of the reflected wave detected in the radar detection step 210 is equal to or higher than the high threshold 310, the radar detection result determination unit 131 determines that the obstacle is a vehicle (vehicle determination area 311). When the intensity of the reflected wave detected in the radar detection step 210 is equal to or higher than the low threshold 320 and lower than the high threshold 310, the radar detection result determination unit 131 determines that the obstacle is a pedestrian (pedestrian Determination area 321). When the intensity of the reflected wave detected in the radar detection step 210 is less than the low threshold 320, the radar detection result determination unit 131 determines that the reflected wave is noise (noise determination area 330).

  Thus, in the radar detection result determination step 220, the radar detection result determination means 131 determines whether the obstacle is a stationary target, a preceding vehicle, or an oncoming vehicle based on the reflected wave detected in the radar detection step 210. Is judged.

  In the image acquisition step 230, the image acquisition unit 120 acquires an image around the host vehicle.

  In the image determination step 240, the image determination unit 132 matches a preset pedestrian pattern and vehicle pattern with respect to the image around the host vehicle acquired in the image acquisition step 230.

  The pedestrian pattern is, for example, the direction of the pedestrian (forward, backward, leftward, rightward, etc.) and the size of the pedestrian (pedestrian far away from the host vehicle, pedestrian close to the host vehicle, etc.) There may be various pedestrian patterns based on it. Furthermore, it may be a pedestrian pattern on a bicycle.

  The vehicle patterns are, for example, various vehicle patterns based on the direction of the vehicle (forward, backward, leftward, rightward, etc.), the size of the vehicle (a vehicle far from the host vehicle, a vehicle close to the host vehicle, etc.), etc. There is no problem.

  Thus, in the image determination step 240, the image determination means 132 matches the pedestrian pattern and the vehicle pattern with respect to the image around the host vehicle acquired in the image acquisition step 230, so that the obstacle is a pedestrian. The vehicle is determined.

  In the obstacle estimation step 250, the obstacle estimation means 133 estimates an obstacle around the host vehicle based on the result determined in the radar detection result determination step 220 and the result determined in the image determination step 240. . In other words, the obstacle estimation unit 133 includes information related to the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220, and information related to the obstacle acquired in the image acquisition step 230 and the image determination step 240. Is fused to estimate the type of the obstacle.

  FIG. 4 shows the fusion of the information related to the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220 and the information related to the obstacle acquired in the image acquisition step 230 and the image determination step 240. 5 is a flowchart showing an obstacle type estimation method 400 for estimating the type of the obstacle.

  Based on the information on the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220 and the information on the obstacle acquired in the image acquisition step 230 and the image determination step 240, the type of the obstacle is estimated. The obstacle type estimation method 400 will be described in detail below with reference to FIG.

  In FIG. 4, the obstacle type estimation method 400 includes a first determination step 410 to a fifth determination step 450.

  In the first determination step 410, it is determined whether or not the obstacle is a stationary target. This is determined in the radar detection result determination step 220 based on the speed V0 and the relative speed Vr of the host vehicle.

  When it is determined in the first determination step 410 that the obstacle is a stationary target (Yes in the first determination step 410), in the second determination step 420, the obstacle is determined based on the intensity of the radar reflected wave. Judge whether it is a pedestrian or a vehicle. Specifically, in the radar detection result determination step 220, the intensity of the reflected wave detected in the radar detection step 210 is compared with the high threshold 310 and the low threshold 320 as shown in FIG. Or the walking vehicle determination area 321 or the noise determination area 330 is determined.

  In the second determination step 420, when it is determined that the intensity of the reflected wave is the pedestrian determination area 321 shown in FIG. 3 (pedestrian determination area in the second determination step 420), in the third determination step 430, It is determined whether it is a pedestrian image pattern. Specifically, in the image determination step 240, it is determined whether the obstacle is a pedestrian by matching a preset pedestrian pattern with respect to the image around the host vehicle acquired in the image acquisition step 230. ing.

  When it is determined that the obstacle is a pedestrian (Yes in the third determination step 430), the obstacle is estimated to be a pedestrian (first estimation result 431). On the other hand, when it is determined that the obstacle is not a pedestrian (No in the third determination step 430), the obstacle is estimated to be a low threshold stationary target (second estimation result 432).

  Here, the first estimation result 431 is determined based on the information about the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220, and the obstacle is determined to be a pedestrian, and the image acquisition step 230 and Based on the information regarding the obstacle acquired in the image determination step 240, the obstacle is determined to be a pedestrian. Therefore, in the obstacle estimation step 250, the first estimation result 431 estimated by the obstacle estimation unit 133 has high reliability.

  When it is determined in the second determination step 420 that the intensity of the reflected wave is the vehicle determination area 311 shown in FIG. 3 (vehicle determination area in the second determination step 420), in the fourth determination step 440, the vehicle image It is determined whether or not it is a pattern. Specifically, in the image determination step 240, it is determined whether the obstacle is a vehicle by matching a vehicle pattern set in advance with respect to the image around the host vehicle acquired in the image acquisition step 230. .

  When it is determined that the obstacle is a vehicle (Yes in the fourth determination step 440), it is estimated that the obstacle is a vehicle (third estimation result 441). On the other hand, when it is determined that the obstacle is not a vehicle (No in the fourth determination step 440), the obstacle is estimated to be a high threshold stationary target (fourth estimation result 442).

  Here, the third estimation result 441 is determined that the obstacle is a vehicle based on the information about the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220, and the image acquisition step 230 and the image Based on the information regarding the obstacle acquired in the determination step 240, it is determined that the obstacle is a vehicle. Therefore, the third estimation result 441 estimated by the obstacle estimation unit 133 in the obstacle estimation step 250 is highly reliable.

  When it is determined in the first determination step 410 that the obstacle is not a stationary target (No in the first determination step 410), it is determined in the fifth determination step 450 whether or not the obstacle is a preceding vehicle. . This is determined in the radar detection result determination step 220 based on the speed V0 and the relative speed Vr of the host vehicle.

  When it is determined that the obstacle is a preceding vehicle (Yes in fifth determination step 450), it is estimated that the obstacle is a preceding vehicle (fifth estimation result 451). On the other hand, when it is determined that the obstacle is not a preceding vehicle (No in the fifth determination step 450), it is estimated that the obstacle is an oncoming vehicle (sixth estimation result 452).

  Here, the fifth estimation result 451 and the sixth estimation result 452 are based on only the information regarding the obstacle acquired in the radar detection step 210 and the radar detection result determination step 220, and the obstacle is the preceding vehicle and the oncoming vehicle. It is determined. This is because the determination result is highly reliable even if it is not determined based on the information regarding the obstacle acquired in the image acquisition step 230 and the image determination step 240.

  In the collision avoidance step 260, the obstacle avoidance control means 150 controls the collision avoidance system 160 according to the result estimated in the obstacle estimation step 250. Specifically, the obstacle avoidance control unit 150 controls the timing for operating the collision avoidance system 160 based on the obstacle type indicated by the obstacle type estimation method 400 in FIG.

  FIG. 5 is a diagram illustrating timings at which the collision avoidance system 160 is activated for a pedestrian (first estimation result 431) and a low threshold stationary target (second estimation result 432). The first estimation result 431 indicating that the obstacle in the obstacle estimation step 250 is a pedestrian is highly reliable. Therefore, the obstacle avoidance control means 150 uses the collision avoidance system 160 for avoiding a collision with a pedestrian based on the first estimation result 431 indicating that the obstacle is a pedestrian. Compared with the timing in the case of a stationary target (second estimation result 432), it is operated at an earlier timing.

  FIG. 6 is a diagram illustrating the timing (corresponding to the speed of the host vehicle) for operating the collision avoidance system 160 in a stationary vehicle (third estimation result 441) and a high threshold stationary target (fourth estimation result 442). is there. The third estimation result 441 indicating that the obstacle in the obstacle estimation step 250 is a stationary vehicle has high reliability. Therefore, the obstacle avoidance control unit 150 uses the collision avoidance system 160 for avoiding a collision with a stationary vehicle based on the third estimation result 441 indicating that the obstacle is a stationary vehicle. Compared with the timing in the case of a stationary target (fourth estimation result 442), the operation is performed at an earlier timing.

  FIG. 7 is a diagram illustrating timing for operating the collision avoidance system 160 in the preceding vehicle (the fifth estimation result 451) and the oncoming vehicle (the sixth estimation result 452). As described above, with respect to the fifth estimation result 451 indicating that the obstacle is the preceding vehicle and the sixth estimation result 452 indicating that the obstacle is the oncoming vehicle in the obstacle estimation step 250, the radar detection step 210 and The reliability is sufficiently high even if the determination is made only with the information related to the obstacle acquired in the radar detection result determination step 220. Therefore, the obstacle avoidance control means 150 operates the collision avoidance system 160 for avoiding a collision with the preceding vehicle at an appropriate timing based on the fifth estimation result 451 indicating that the obstacle is the preceding vehicle. The collision avoidance system 160 for avoiding the collision with the oncoming vehicle is operated at an appropriate timing based on the sixth estimation result 452 indicating that the obstacle is the oncoming vehicle.

  In FIGS. 5 to 7 described above, the distance / relative speed from the host vehicle to the target according to the speed V0 of the host vehicle is the timing for operating the collision avoidance system 160 for avoiding a collision with an obstacle. However, the present invention is not limited to this. For example, the timing of operating the collision avoidance system 160 for avoiding a collision with an obstacle is indicated by the distance / relative speed from the own vehicle to the target according to the lateral position to avoid the collision with the obstacle. It doesn't matter.

  FIG. 8 corresponds to the timing at which the collision avoidance system 160 is operated in the stationary vehicle (third estimation result 441) and the high threshold stationary target (fourth estimation result 442) (the lateral position necessary for collision avoidance). ). Here, the lateral position necessary for collision avoidance is calculated based on, for example, the center position of the host vehicle, the lateral width of the host vehicle, and the lateral width of the obstacle. When the obstacle is a vehicle, it is easy to accurately estimate how much the own vehicle should move in order to avoid a collision with the obstacle. In FIG. 8, as in FIG. 6, the obstacle avoidance control unit 150 is configured to prevent the collision with the stationary vehicle based on the third estimation result 441 indicating that the obstacle is a stationary vehicle. 160 is operated at an earlier timing than the timing when the obstacle is a high threshold stationary target (fourth estimation result 442). Further, the larger the lateral position necessary for collision avoidance, the earlier the collision avoidance system 160 is operated.

  Further, as shown in FIGS. 5 to 7, the timing at which the collision avoidance system 160 is operated is shown in FIG. 8 as the timing indicated by the distance / relative speed from the host vehicle to the target according to the speed V0 of the host vehicle. As shown, in order to avoid a collision with an obstacle, it may be set by combining timings indicated by the distance / relative speed from the host vehicle to the target according to the lateral position.

  In addition, about the obstacle avoidance control means 150 controlling the collision avoidance system 160, you may control the collision avoidance system classification to operate besides the timing which operates the collision avoidance system 160.

  FIG. 9 is a diagram showing a collision avoidance system that operates according to the obstacle estimated by the obstacle estimating means 133 and / or the reliability of the estimation. When the obstacle is estimated to be a pedestrian and the reliability of the estimation result is high (for example, the first estimation result 431 shown in FIG. 4), the obstacle is estimated to be a low threshold stationary target ( For example, the collision avoidance system 160 is operated at an earlier timing (operation timing T1 (earlier than T2)) than the second estimation result 432). Furthermore, in this case, vehicle brakes and pedestrian protection hood airbags are set as types of collision avoidance systems to be operated.

  Similarly, when the obstacle is estimated to be a stationary vehicle and the reliability of the estimation result is high (for example, the third estimation result 441 shown in FIG. 4), the obstacle is estimated to be a high threshold stationary target. The collision avoidance system 160 is operated at an earlier timing (operation timing T3 (earlier than T4)) than in the case (for example, the fourth estimation result 442). Further, in this case, vehicle brakes, driver's seat airbags, and seat belts are set as types of collision avoidance systems to be operated.

  In this way, if a collision avoidance system that operates according to the obstacle estimated by the obstacle estimating means 133 and / or the reliability of the estimation is set, for example, a pedestrian who is weak in traffic, and / or Alternatively, an appropriate collision avoidance system (collision damage reduction system) can be operated at an appropriate timing in consideration of driver safety.

  Moreover, you may set in detail the timing which operates for every collision avoidance system classification to operate. It is preferable to operate each appropriate collision avoidance system type at an appropriate timing according to the obstacle.

  As described above, according to the obstacle avoidance device 100 and the obstacle avoidance method 200 according to an embodiment of the present invention, while recognizing an obstacle around the host vehicle with high accuracy, the recognized obstacle and / or Alternatively, the collision avoidance system can be operated at an appropriate timing according to the reliability of the recognition.

  Furthermore, according to the obstacle avoidance device 100 and the obstacle avoidance method 200 according to an embodiment of the present invention, when an obstacle is erroneously recognized or when an obstacle is unclear, a collision is wasted. It is possible to prevent the avoidance system 160 from operating.

  Furthermore, if the collision avoidance system type to be operated is controlled in addition to the timing to activate the collision avoidance system 160, an appropriate collision avoidance system can be operated at an appropriate timing. The safety of the driver and / or the driver can be ensured more appropriately.

  In the present embodiment, the obstacle type estimation method 400 is illustrated in FIG. 4 as an obstacle estimation method, but the obstacle estimation method is not limited to this.

  For example, in the first determination step 410, whether or not the obstacle is a stationary target is determined based on the speed V0 and the relative speed Vr of the host vehicle. It may be determined whether the target is a low speed moving target or a high speed moving target.

  In the second determination step 420, as shown in FIG. 3, the vehicle determination area 311, the pedestrian determination area 321, and the noise determination area 330 are determined using the high threshold 310 and the low threshold 320. However, the present invention is not limited to this. For example, when the obstacle is a pedestrian, the intensity of the radar reflected wave may become unstable. For this reason, when the intensity of the radar reflected wave is not less than the low threshold 320 and less than the high threshold 310, it is determined that the obstacle is a pedestrian, and when the intensity of the radar reflected wave is not less than the high threshold 310, It may be determined that the object is a vehicle or a pedestrian. If the intensity of the radar reflected wave is equal to or higher than the high threshold 310, both the third determination step 430 and the fourth determination step 440 in FIG. 4 may be performed. In all cases where the intensity of the radar reflected wave is higher than the high threshold value 310, the radar reflected wave intensity is within a predetermined range from the high threshold value 310 without determining that the obstacle is a vehicle or a pedestrian. Only in some cases, it may be determined that the obstacle is a vehicle or a pedestrian.

  Further, in the third determination step 430 and the fourth determination step 440, it is determined whether or not the obstacle is a pedestrian and a vehicle by matching a preset pedestrian pattern and a vehicle pattern. The images to be matched are not limited to these. For example, obstacles may be determined in more detail by presetting and matching image patterns of bicycles, motorcycles, large vehicles, streetcars, and the like. In this case, if the intensity of the reflected wave corresponding to each of the plurality of obstacles is set and the determination in the second determination step 420 is performed, the obstacle estimation with higher accuracy and higher reliability can be performed.

  Further, when there is no correlation between the information related to the obstacle detected by the radar detection unit 110 and the information related to the obstacle recognized from the image acquired by the image acquisition unit 120, The collision avoidance system 160 corresponding to the estimation result was operated (for example, the second estimation result 432 and the fourth estimation result 442). However, for example, a priority is set in advance based on the information related to the obstacle detected by the radar detection unit 110 and the information related to the obstacle recognized from the image acquired by the image acquisition unit 120. Obstacles may be estimated based on the information to be processed. By prioritizing information in the detection means with high reliability, it is possible to estimate obstacles with high reliability.

  Further, when the obstacle is not estimated based on any priority information, for example, the obstacle estimated based on the information on the obstacle detected by the radar detection unit 110 and the image acquisition unit 120 are acquired. Preparations may be made to activate the collision avoidance system 160 corresponding to both obstacles estimated based on information relating to obstacles recognized from the images. Among the plurality of obstacles that can be detected, for example, the obstacle is narrowed down to two obstacles at an early stage. Therefore, if the obstacle can be estimated by further detection means, the appropriate collision avoidance system 160 can be operated quickly. .

  Furthermore, in this embodiment, the radar detection unit 110 and the image acquisition unit 120 are used as the unit for detecting an obstacle, but the present invention is not limited to this. Other detection means may be used as long as obstacles around the host vehicle can be recognized with high accuracy and high reliability.

  The present invention is applied to an obstacle avoidance device that avoids collision with an obstacle around the host vehicle or reduces collision damage, and is particularly useful for a PCS or the like mounted on a vehicle.

DESCRIPTION OF SYMBOLS 100 Obstacle avoidance apparatus 110 Radar detection means 120 Image acquisition means 130 Correlation determination means 140 Recording means 150 Obstacle avoidance control means 160 Collision avoidance system 131 Radar detection result determination means 132 Image determination means 133 Obstacle estimation means 200 Obstacle avoidance Method 210 Radar detection step 220 Radar detection result determination step 230 Image acquisition step 240 Image determination step 250 Obstacle estimation step 260 Collision avoidance step 310 High threshold 320 Low threshold 311 Vehicle determination area 321 Pedestrian determination area 330 Noise determination area 400 Obstacle Type estimation method 410, 420, 430, 440, 450 Each determination step 431, 432, 441, 442, 451, 452 in the obstacle type estimation method Each determination result in the obstacle type estimation method

Claims (7)

An obstacle avoidance device that detects an obstacle around the host vehicle and operates a collision avoidance system that avoids a collision with the obstacle,
A plurality of detecting means for detecting the obstacle;
Correlation determining means for determining whether or not each detection result detected by the plurality of detection means has a correlation based on information on the obstacle detected by the plurality of detection means;
An obstacle avoidance control means for controlling the collision avoidance system according to the determination result determined by the correlation determination means,
The obstacle avoidance control unit operates the collision avoidance system at a timing earlier than the timing when the obstacle determination control unit determines that there is a correlation when the obstacle determination control unit determines that there is no correlation. An obstacle avoidance device characterized by the above. The plurality of detection means include
Radar detecting means for detecting the obstacle via a radar;
The obstacle avoidance device according to claim 1, further comprising: an image detection unit that acquires an image around the host vehicle using a camera and detects the obstacle from the image. The correlation determination means includes
Radar detection result determination means for determining the type of the obstacle based on the radar reflection intensity in the radar detection means;
An image detection result determination unit that determines the type of the obstacle by matching a predetermined image pattern set in advance with respect to the image acquired by the image detection unit;
Obstacle estimation means for estimating the obstacle based on the obstacle type determined by the radar detection result determination means and the obstacle type determined by the image detection result determination means. The obstacle avoidance device according to claim 2, wherein:   The obstacle avoidance control means, when the obstacle type determined by the radar detection result determination means and the obstacle type determined by the image detection result determination means are the same, 4. The collision avoidance system that operates corresponding to the obstacle estimated by the estimation means is operated at a timing earlier than the timing when the type of the obstacle is not the same. 5. Obstacle avoidance device.   The obstacle avoidance control means is the same as the obstacle type determined by the radar detection result determination means and the obstacle type determined by the image detection result determination means. Is identified as a predetermined obstacle set in advance, a collision avoidance system that operates in response to the obstacle estimated by the obstacle estimating means The obstacle avoidance device according to claim 4, wherein the obstacle avoidance device is operated at an earlier timing than an operation timing when it is not specified to be an object. A recording unit in which a look-up table indicating a collision avoidance system that operates in response to the obstacle is recorded in advance;
The obstacle avoidance control means refers to a look-up table recorded in the recording means, and controls the collision avoidance system according to the determination result determined by the correlation determination means. The obstacle avoidance device according to claim 1. An obstacle avoidance method executed by an obstacle avoidance device that detects an obstacle around the host vehicle and operates a collision avoidance system that avoids a collision with the obstacle,
A plurality of detection steps for detecting the obstacle;
A correlation determination step for determining whether or not each detection result detected in the plurality of detection steps has a correlation, based on information on the obstacle detected in the plurality of detection steps;
An obstacle avoidance control step for controlling the collision avoidance system according to the determination result determined in the correlation determination step,
In the obstacle avoidance control step, when the correlation determination step determines that there is a correlation, the collision avoidance system is operated at a timing earlier than the timing when it is determined that there is no correlation. The obstacle avoidance method characterized by this.

Images