JP2010132056A - Sensing device, sensing method, and vehicle control device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To take into account the degrees of reliability of sensing results when sensing obstacles by using the sensing results through an on-vehicle radar or an on-vehicle camera. <P>SOLUTION: A sensing device includes preliminarily a reliability table for storing the degrees of reliability of sensing results through individual on-vehicle sensors by associating them with environmental conditions when the sensing results are captured from the individual on-vehicle sensors (such as night, rain, or the like) or sensing objects that become sensing objects of the individual on-vehicle sensors (such as long distance objects, pedestrians or the like), respectively. The sensing device calculates the degrees of reliability of the sensing results by referring to the reliability table based on the environmental conditions during capture of the sensing results and the sensing objects indicated by the sensing results. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a detection device, a detection method, and a vehicle control device.

  Conventionally, for the purpose of avoiding collisions with obstacles, etc., sensing the surroundings of the route of the host vehicle with a radar or in-vehicle camera, and comprehensively judging each sensing result to detect and recognize obstacles Exists.

  For example, Patent Document 1 proposes a technique for accurately recognizing only a vehicle ahead by adding a sensing result obtained by an image sensor to a sensing result obtained by a laser radar.

JP 2003-84064 A

  The reliability of the sensing result varies depending on the environment in which various sensors such as a radar and a camera mounted on the vehicle are placed and the objects detected by the various sensors. However, the above-described conventional technique has a problem that the reliability of the sensing result is not taken into account when an obstacle is detected using the sensing result.

  Therefore, the present invention has been made to solve the above-described problems of the prior art, and considers the reliability of the sensing result obtained from each in-vehicle sensor when an obstacle is detected using the sensing result. It is an object of the present invention to provide a detection device, a detection method, and a vehicle control device that can be used.

  In order to solve the above-described problems and achieve the object, the present invention is an environmental condition when a sensing result is acquired from each in-vehicle sensor that senses an obstacle around the vehicle, or a sensing target of each in-vehicle sensor. A reliability storage means for storing the reliability of the sensing result by each in-vehicle sensor in association with each detection target, and when the sensing result by each in-vehicle sensor is acquired, Based on the detected object indicated by the sensing result, the reliability storage means is referred to, a reliability calculating means for calculating the reliability of the sensing result, the reliability calculated by the reliability calculating means, and the sensing And a detection determination unit that determines whether or not the obstacle is detected using a result.

  In addition, the present invention is based on each in-vehicle sensor that senses an obstacle around the vehicle based on the environmental condition at the time of acquiring the sensing result and the detected object indicated by the sensing result when the sensing result by each in-vehicle sensor is acquired. Refer to the reliability storage means for storing the reliability of the sensing result by each in-vehicle sensor in association with the environmental condition when the sensing result is acquired or the detection object to be sensed by each in-vehicle sensor. And determining whether or not the obstacle has been detected using the reliability calculation step for calculating the reliability of the sensing result, the reliability calculated by the reliability calculation step, and the sensing result. A detection determination step.

  ADVANTAGE OF THE INVENTION According to this invention, when detecting an obstruction using the sensing result by the radar mounted in a vehicle or a vehicle-mounted camera, the reliability of a sensing result can be considered.

  Hereinafter, embodiments of a detection device, a detection method, and a vehicle control device according to the present invention will be described in detail with reference to the accompanying drawings. In addition, below, after describing Example 1 as one embodiment of the detection device according to the present invention, other embodiments included in the detection device according to the present invention will be described.

  In the following first embodiment, the outline and features of the detection device according to the first embodiment, the configuration and processing of the detection device will be described in order, and finally the effects of the first embodiment will be described.

[Outline and Features of Detection Device (Example 1)]
First, the outline and features of the detection apparatus according to the first embodiment will be described with reference to FIG. FIG. 1 and FIG. 2 are diagrams for explaining the outline and features of the detection apparatus according to the first embodiment.

  The detection apparatus according to the first embodiment is mounted on a vehicle, and outlines the detection of an obstacle existing around the traveling vehicle. And the detection apparatus which concerns on Example 1 has the main characteristics in the point which considers the reliability of a sensing result, when detecting an obstruction using the sensing result by the radar mounted in a vehicle, or a vehicle-mounted camera.

  This main feature will be specifically described. As shown in FIG. 1, the detection apparatus according to the first embodiment calculates the reliability of the sensing result when the sensing result obtained by the radar, the in-vehicle camera, and the environment sensor is acquired (see FIG. 1). 1 (1)).

  Specifically, as shown in FIG. 1, the detection device is configured such that an environmental condition (for example, night or rain) when a sensing result is acquired from each in-vehicle sensor that senses an obstacle around the vehicle while the vehicle is running. ) Or a reliability table that stores the reliability of the sensing result of each in-vehicle sensor in advance in association with a detection target (for example, a long-distance object or a pedestrian) to be sensed by each in-vehicle sensor. Prepare.

  Then, the detection device calculates the reliability of the sensing result by referring to the reliability table based on the environmental condition at the time of sensing result acquisition and the detected object indicated by the sensing result. For example, as shown in FIG. 2, the reliability of the sensing result by the radar corresponding to the environmental condition “night” is calculated as “high”, and the reliability of the sensing result by the in-vehicle camera corresponding to the environmental condition “night”. The degree is calculated as “low”. Furthermore, the reliability of the sensing result by the radar corresponding to the detected object “far object” indicated by the sensing result is calculated as “high”, and the sensing by the in-vehicle camera corresponding to the detected object “far object” indicated by the sensing result. The reliability of the result is calculated as “low”.

  After calculating the reliability, the detection device changes a detection threshold set in advance for each type of vehicle-mounted sensor based on the reliability of the sensing result in order to determine whether an obstacle has been detected (FIG. 1). (See (2)).

  Specifically, the detection device changes the detection threshold set in advance for each type of on-vehicle sensor to be lower than the initial value based on the reliability calculated previously. For example, in the case shown in FIG. 2, since the reliability of the in-vehicle camera corresponding to the environmental condition “night” and the detected object “far-distance object” indicated by the sensing result is low, the detection threshold for the in-vehicle camera sensing result is initially set. Change lower than the value.

  After the detection threshold is changed, the detection device compares the detection threshold set in advance for each type of vehicle-mounted sensor, the detection threshold that has been changed to a lower value, and the sensing value of the sensing result corresponding to each detection threshold. Is used to detect the obstacle (see (3) in FIG. 1). Specifically, when it is determined that each sensing value acquired by the radar and the vehicle-mounted camera exceeds the detection threshold for the radar and the previously detected vehicle-mounted camera, It is determined that an object has been detected.

  When it is determined that an obstacle has been detected, the detection device notifies the driver that the obstacle has been detected (for example, that a long-distance object has been detected) (see (4) in FIG. 1). ).

  As described above, the detection apparatus according to the first embodiment, when detecting an obstacle using a sensing result obtained by a radar or an in-vehicle camera mounted on a vehicle, as described above, is a reliability of the sensing result. Can be added.

[Configuration of Detection Device (Example 1)]
Next, the configuration of the detection apparatus according to the first embodiment will be described with reference to FIGS. FIG. 3 is a diagram illustrating the configuration of the detection device according to the first embodiment.

  As shown in FIG. 3, the detection apparatus according to the first embodiment is mounted on a vehicle together with the environmental sensor group 100, the radar 200, the radar recognition unit 210, the in-vehicle camera 220, the image recognition unit 230, the display 240 and the speaker 250. .

  The environmental sensor group 100 measures or detects an environmental condition when a sensing result is acquired from each in-vehicle sensor that senses an obstacle around the vehicle. As shown in FIG. 3, an illuminance sensor 110 for measuring illuminance, a temperature sensor 120 for measuring temperature, a raindrop sensor 130 for detecting rain, a gyro sensor 140 for measuring road undulations, and the like An infrastructure information device 150 that acquires and accumulates infrastructure information is included. The infrastructure information device 150 has a communication function executed between vehicles and between roads and vehicles, and acquires information on infrastructures such as roadside structures such as guardrails and guard ropes and roadside structures such as tunnels by communication. Accumulate. The sensing result measured or detected by the environmental sensor group 100 is output to the environment specifying unit 320 described later.

  The radar 200 irradiates a radar (for example, 76.5 GHz millimeter wave), and distance and speed with an obstacle (pedestrian or other vehicle) in the radar irradiation range (sensing range) in front of or on the side of the vehicle. Measure the direction, etc., and obtain it as a sensing result. A sensing result by the radar 200 is output to the radar recognition unit 210.

  The radar recognition unit 210 recognizes a sensed obstacle to some extent from the sensing result input from the radar 200, and outputs the recognition result to a reliability calculation unit 330 described later. For example, from a distance, speed, and direction from an obstacle input as a sensing result from the radar 200, a long distance object, a vehicle, a pedestrian, or the like is recognized to some extent, and a reliability calculation unit 330 described later on the recognition result and the sensing result. Output to.

  The in-vehicle camera 220 captures a sensing range (imaging direction) set in front of or on the side of the vehicle and acquires it as a sensing result. A sensing result by the in-vehicle camera 220 is output to the image recognition unit 230.

  The image recognizing unit 230 performs preprocessing (for example, sharpness, contrast adjustment, saturation adjustment) for emphasizing the outline of the obstacle in the image input from the in-vehicle camera 220, and then the obstacle displayed in the image The object is recognized, and the recognition result and the sensing result are output to the reliability calculation unit 330 described later.

  The display 240 outputs notification contents from a notification control unit 360 described later as an image. The speaker 250 outputs notification contents from the notification control unit 360 described later by voice.

  The detection apparatus 300 includes a reliability table 1 (310), a reliability table 2 (311), a reliability table 3 (312), an environment identification unit 320, a reliability calculation unit 330, a detection threshold change unit 340, and a detection determination unit 350. And a notification control unit 360.

  The reliability table 1 (310) stores in advance the reliability of the sensing result of the radar or the in-vehicle camera in association with the classification item related to the environmental condition. For example, as shown in FIG. 4, when the classification item is “night”, the reliability of the sensing result is “radar”, and the reliability of the sensing result is “in-vehicle camera”. Data is stored in advance. FIG. 4 is a diagram illustrating a configuration example of the reliability table 1 according to the first embodiment.

  The reliability table 2 (311) is associated with classification items related to infrastructure information such as roadside structures such as guard rails and guard ropes and road structures such as tunnels, and the reliability of the sensing result of the radar or the in-vehicle camera. Is stored in advance. For example, as shown in FIG. 5, when the classification item is “guardrail”, the reliability of the sensing result is “on-vehicle camera”, and the reliability of the sensing result is “radar”. Data is stored in advance. FIG. 5 is a diagram illustrating a configuration example of the reliability table 2 according to the first embodiment.

  The reliability table 3 (312) stores in advance the reliability of the sensing result of the radar or the in-vehicle camera in association with a detection target such as a long-distance object or a pedestrian. For example, as shown in FIG. 6, when the classification item is “long-distance object”, the sensing result reliability is “radar” and the sensing result reliability is “vehicle camera”. The data that there is is stored in advance. FIG. 6 is a diagram illustrating a configuration example of the reliability table 3 according to the first embodiment.

  The environment specifying unit 320 specifies environmental conditions around the host vehicle based on the sensing result input from the environment sensor group 100. Specifically, the environment identification unit 320 analyzes the sensing results from the illuminance sensor 110, the temperature sensor 120, the raindrop sensor 130, the gyro sensor 140, and the infrastructure information device 150, so that night (dark in a tunnel or the like), rain, Identify environmental conditions such as predetermined road undulations and tunnels. The environmental condition specified by the environment specifying unit 320 is output to the reliability calculation unit 330.

  The reliability calculation unit 330 uses the environmental conditions input from the environmental sensor group 100 and the recognition results input from the radar recognition unit 210 and the image recognition unit 230 to determine the reliability of the sensing results from the radar 200 and the in-vehicle camera 220. calculate.

  Specifically, the reliability calculation unit 330 is based on the environmental condition at the time of sensing result input from the environmental sensor group 100 and the recognition result input from the radar recognition unit 210 and the image recognition unit 230 based on the reliability table 1 ( 311) to reliability table 3 (312), the reliability of the sensing result is calculated.

  For example, when the environmental condition at the time of sensing result acquisition is “night”, referring to the reliability table 1 (310), the sensing result of the radar 200 corresponding to “night” as shown in FIG. The reliability is calculated as “high”, and the reliability of the sensing result by the in-vehicle camera 220 corresponding to “night” is calculated as “low”. Furthermore, when the recognition result from the radar 200 and the in-vehicle camera 220 is “far object”, the reliability table 3 (312) is referred to, and as shown in FIG. The reliability of the sensing result obtained by the radar 200 is calculated as “high”, and the reliability of the sensing result obtained by the in-vehicle camera 220 corresponding to “distant object” is calculated as “low”.

  Then, the reliability calculation unit 330 outputs the reliability calculation result (see FIG. 7) of the sensing results obtained by the radar 200 and the in-vehicle camera 220 to the detection threshold value changing unit 340 and inputs from the radar recognition unit 210 and the image recognition unit 230. The sensed result is output to the detection determination unit 350. FIG. 7 is a diagram illustrating an example of a reliability calculation result according to the first embodiment.

  If the environmental condition at the time of sensing result acquisition is “night” and the recognition result from the radar 200 and the in-vehicle camera 220 is “pedestrian”, refer to the reliability table 1 (310), As shown in FIG. 8, the reliability of the sensing result by the radar 200 and the in-vehicle camera 220 corresponding to “night” is calculated as “high” and “low”, respectively. Further, referring to the reliability table 3 (312), as shown in the figure, the reliability of the sensing result by the radar 200 and the in-vehicle camera 220 corresponding to “pedestrian” is set to “low” and “high”, respectively. calculate.

  Then, the reliability calculation unit 330 outputs the reliability calculation result (see FIG. 8) of the sensing results obtained by the radar 200 and the in-vehicle camera 220 to the detection threshold value changing unit 340 and inputs from the radar recognition unit 210 and the image recognition unit 230. The sensed result is output to the detection determination unit 350. FIG. 8 is a diagram illustrating an example of a reliability calculation result according to the first embodiment.

  If the environmental conditions at the time of sensing result acquisition are “night” and “rain” and the recognition results from the radar 200 and the in-vehicle camera 220 are “pedestrians”, the reliability table 1 (310) is stored. Referring to FIG. 9, the reliability of the sensing results by the radar 200 and the in-vehicle camera 220 corresponding to “night” or “rain” is “high” and “low”, “high” and “low”, respectively. And calculate. Further, referring to the reliability table 3 (312), as shown in the figure, the reliability of the sensing result by the radar 200 and the in-vehicle camera 220 corresponding to “pedestrian” is set to “low” and “high”, respectively. calculate.

  Then, the reliability calculation unit 330 outputs the reliability calculation result (see FIG. 9) of the sensing results obtained by the radar 200 and the in-vehicle camera 220 to the detection threshold value changing unit 340 and inputs from the radar recognition unit 210 and the image recognition unit 230. The sensed result is output to the detection determination unit 350. FIG. 9 is a diagram illustrating an example of the reliability calculation result according to the first embodiment.

  If the environmental condition at the time of sensing result acquisition is “night” and the recognition results from the radar 200 and the in-vehicle camera 220 are “pedestrian” and “guardrail”, the reliability table 1 (310) is stored. Referring to FIG. 10, the reliability of the sensing result by radar 200 and in-vehicle camera 220 corresponding to “night” is calculated as “high” and “low”, respectively. Further, referring to the reliability table 3 (312), as shown in the figure, the reliability of the sensing results obtained by the radar 200 and the in-vehicle camera 220 corresponding to “pedestrian” or “guardrail” is set to “low” and “low”, respectively. Calculate as “high”, “low” and “high”.

  Note that the recognition results input from the radar 200 and the in-vehicle camera 220 are “guardrail”, “post”, “snow wall”, etc., and the sensing results input from the infrastructure information device 150 of the environmental sensor group 100 are the same. , “Guardrail”, “post”, “snow wall”, etc., the reliability of the sensing result by the radar 200 and the in-vehicle camera 220 is calculated as “high”.

  Then, the reliability calculation unit 330 outputs the reliability calculation result (see FIG. 10) of the sensing results obtained by the radar 200 and the in-vehicle camera 220 to the detection threshold value changing unit 340 and inputs from the radar recognition unit 210 and the image recognition unit 230. The sensed result is output to the detection determination unit 350. FIG. 10 is a diagram illustrating an example of the reliability calculation result according to the first embodiment.

  The detection threshold value changing unit 340 has a detection threshold value set in advance for each type of vehicle-mounted sensor in order to determine whether an obstacle is detected based on the reliability calculation result input from the reliability calculation unit 330. Make a change.

  Specifically, the detection threshold value changing unit 340 executes case classification of the reliability calculation result input from the reliability calculation unit 330 based on, for example, a classification standard as illustrated in FIG. For example, when the reliability calculation result as shown in FIG. 7 or FIG. 9 is input from the reliability calculation unit 330, the sensing result by the in-vehicle camera 220 is less reliable than the sensing result by the radar 200. Therefore, the reliability calculation result is classified as Case A. Then, when the reliability calculation result is classified as case A, the detection threshold value changing unit 340 sets the detection threshold value set in advance for the in-vehicle camera 220 at a predetermined rate lower than the initial value. A change instruction is output to the detection determination unit 350.

  Further, for example, when the reliability calculation result as shown in FIG. 8 or 10 is input, the detection threshold value changing unit 340 has the same reliability of the sensing result by the radar 200 and the in-vehicle camera 220 (“low”). ”And“ high ”are the same number and the number of reliability“ low ”≧ the number of reliability“ high ”), which corresponds to case C. Therefore, the reliability calculation result is classified into case C. Then, when the reliability calculation result is classified as case C, the detection threshold value changing unit 340 sets the detection threshold values set in advance for the radar 200 and the in-vehicle camera 220 at a predetermined rate lower than the initial values, respectively. Thus, a change instruction is output to the detection determination unit 350.

  In addition, the detection threshold value changing unit 340 has, for example, the reliability of the sensing results obtained by the radar 200 and the in-vehicle camera 220 are equal, and unlike the case C, the number of reliability “low” <the number of reliability “high”. When it is, it classifies without the corresponding case. The detection threshold changing unit 340 does not output a detection threshold changing instruction to the detection determining unit 350 when the reliability calculation result is classified as no corresponding case. FIG. 11 is a diagram illustrating a case classification reference example of the reliability calculation result according to the first embodiment.

  The detection determination unit 350 includes a detection threshold that is set in advance for each type of in-vehicle sensor, a detection threshold that is changed in accordance with a change instruction input from the detection threshold change unit 340, and a sensing result corresponding to each detection threshold. Obstacle detection is determined using the comparison result with the sensing value.

  Specifically, when the detection determination unit 350 inputs a detection threshold change instruction from the detection threshold change unit 340, the detection threshold for the in-vehicle sensor (radar 200 or in-vehicle camera 220) to be changed is determined from the initial value. Is also set low at a predetermined rate. Whether each of the sensing values acquired by the radar 200 and the in-vehicle camera 220 exceeds a corresponding detection threshold (a detection threshold set in advance for the radar 200 or the in-vehicle camera 220 or a detection threshold that has been changed to be low). Determine whether or not.

  As a result of the determination, if it is determined that the respective sensing values acquired by the radar 200 and the vehicle-mounted camera 220 exceed the corresponding detection threshold values, the detection determination unit 350 detects an obstacle. judge. Then, a determination result indicating that the obstacle indicated by the recognition result by the radar recognition unit 210 and the image recognition unit 230 has been detected is output to the notification control unit 360.

  The notification control unit 360 notifies the driver that an obstacle has been detected (for example, that a long-distance object has been detected) via the display 240 or the speaker 250 according to the determination result input from the detection determination unit 350. .

[Processing by Detection Device (Example 1)]
Subsequently, a process performed by the detection apparatus according to the first embodiment will be described with reference to FIG. FIG. 12 is a diagram illustrating a process flow of the detection device according to the first embodiment.

  As shown in the figure, when the sensing result is acquired by the radar 200, the in-vehicle camera 220, and the environment sensor group 100 (Yes in step S1), the reliability calculation unit 330 displays the environmental condition and the sensing result when the sensing result is acquired. The reliability of the sensing result is calculated by referring to the reliability tables 1 (310) to 3 (312) based on the detected object (step S2).

  Specifically, the reliability calculation unit 330 is based on the environmental condition at the time of sensing result input from the environmental sensor group 100 and the recognition result input from the radar recognition unit 210 and the image recognition unit 230 based on the reliability table 1 ( 310) to reliability table 3 (312), the reliability of the sensing result is calculated.

  After the reliability calculation by the reliability calculation unit 330, the detection threshold value changing unit 340 executes the case classification of the reliability calculation result calculated by the reliability calculation unit 330 based on the classification criteria as shown in FIG. (Step S3). Then, the detection threshold value changing unit 340 outputs an instruction to change the detection threshold value set in advance for each type of vehicle-mounted sensor in order to determine whether or not an obstacle has been detected according to the case classification result (step) S4).

  For example, when the reliability calculation result as shown in FIG. 7 or FIG. 9 is input from the reliability calculation unit 330, the sensing result by the in-vehicle camera 220 is less reliable than the sensing result by the radar 200. Therefore, the reliability calculation result is classified as Case A. Then, when the reliability calculation result is classified as case A, the detection threshold value changing unit 340 sets the detection threshold value set in advance for the in-vehicle camera 220 at a predetermined rate lower than the initial value. A change instruction is output to the detection determination unit 350.

  Further, for example, when the reliability calculation result as shown in FIG. 8 or 10 is input, the detection threshold value changing unit 340 has the same reliability of the sensing result by the radar 200 and the in-vehicle camera 220 (“low”). ”And“ high ”are the same number and the number of reliability“ low ”≧ the number of reliability“ high ”), which corresponds to case C. Therefore, the reliability calculation result is classified into case C. Then, when the reliability calculation result is classified as case C, the detection threshold value changing unit 340 sets the detection threshold values set in advance for the radar 200 and the in-vehicle camera 220 at a predetermined rate lower than the initial values, respectively. Thus, a change instruction is output to the detection determination unit 350.

  The detection determination unit 350 is configured to detect a detection threshold that is set in advance for each type of vehicle-mounted sensor, a detection threshold that is changed in accordance with a change instruction from the detection threshold change unit 340, and a sensing result corresponding to each detection threshold. Obstacle detection is determined using the comparison result with the value (step S5).

  Specifically, when the detection determination unit 350 inputs a detection threshold change instruction from the detection threshold change unit 340, the detection threshold for the in-vehicle sensor (radar 200 or in-vehicle camera 220) to be changed is determined from the initial value. Is also set low at a predetermined rate. Whether each of the sensing values acquired by the radar 200 and the in-vehicle camera 220 exceeds a corresponding detection threshold (a detection threshold set in advance for the radar 200 or the in-vehicle camera 220 or a detection threshold that has been changed to be low). Determine whether or not. And when each sensing value acquired by the radar 200 and the vehicle-mounted camera 220 exceeds the corresponding detection threshold value, it determines with having detected the obstruction.

  If it is determined by the detection determination unit 350 that an obstacle has been detected (Yes at step S5), the notification control unit 360 passes through the display 240 or the speaker 250 according to the determination result of the detection determination unit 350. Is detected to the driver (for example, that a long-distance object has been detected).

  On the other hand, when the sensing result is not acquired (No at Step S1), when it is not determined that the obstacle is detected by the detection determination unit 350 (No at Step S5), and after the notification to the driver is executed (Step S6). In this case, the process returns to step S1 described above. In addition, the process from step S1 to step S6 described above is repeatedly executed when the detection device 300 is activated while the vehicle is traveling.

[Effects of Example 1]
As described above, according to the first embodiment, the reliability of the sensing result can be taken into account when the obstacle is detected using the sensing result obtained by the radar mounted on the vehicle or the in-vehicle camera. Furthermore, by changing the corresponding detection threshold according to the reliability of the sensing result, an obstacle can be detected more positively using a plurality of sensing results (sensing results by a radar or a camera).

  As in the first embodiment, instead of changing the corresponding detection threshold according to the reliability of the sensing result, the sensing result employed for obstacle detection is selected according to the reliability of the sensing result. It may be. Therefore, in the following second embodiment, the configuration and processing of the detection apparatus according to the second embodiment will be described in order, and finally the effects of the first embodiment will be described.

[Configuration of Detection Device (Example 2)]
FIG. 13 is a diagram illustrating the configuration of the detection device according to the first embodiment. As shown in the figure, the detection apparatus 300 according to the second embodiment is different from the first embodiment in that the detection apparatus 300 according to the second embodiment includes a sensing result selection unit 380 instead of the detection threshold value changing unit 340.

  The sensing result selection unit 380 receives from the reliability calculation unit 330 the sensing results obtained by the radar recognition unit 210 and the image recognition unit 230 and the reliability calculation results (see FIGS. 7 to 10) for the sensing results. And the sensing result selection part 380 selects the sensing result employ | adopted for an obstruction detection according to a reliability calculation result.

  Specifically, for example, when a reliability calculation result as illustrated in FIG. 7 or FIG. 9 is input from the reliability calculation unit 330, the sensing result selection unit 380 receives the sensing result from the radar 200 from the in-vehicle camera 220. Since this corresponds to case A having higher reliability than the sensing result (the radar has a higher number of reliability “high”), the reliability calculation result is classified as case A. Then, when classified into case A, the sensing result selection unit 380 selects the sensing result by the radar 200 as the sensing result to be used for obstacle detection, and only the sensing result by the radar 200 is sent to the detection determination unit 350. Output.

  In addition, the sensing result selection unit 380 classifies the reliability calculation result into the case B when the sensing result by the in-vehicle camera 220 corresponds to the case B having higher reliability than the sensing result by the radar 200. Then, the sensing result selection unit 380 selects the sensing result by the in-vehicle camera 220 as the sensing result to be used for obstacle detection when classified into the case B, and detects only the sensing result by the in-vehicle camera 220. Output to 350.

  For example, when a reliability calculation result as illustrated in FIG. 8 or 10 is input to the sensing result selection unit 380, the reliability of the sensing result by the radar 200 and the in-vehicle camera 220 is equal (“low” ”And“ high ”are the same number and the number of reliability“ low ”≧ the number of reliability“ high ”), which corresponds to case C. Therefore, the reliability calculation result is classified into case C.

  In addition, the sensing result selection unit 380 has, for example, the reliability of the sensing results obtained by the radar 200 and the in-vehicle camera 220 are equal, and unlike the case C, the number of reliability “low” <the number of reliability “high”. When it is, it classifies without the corresponding case. Then, the sensing result selection unit 380 selects the sensing result obtained by the radar 200 and the in-vehicle camera 220 as the sensing result to be used for obstacle detection when it is classified as “Case C” or “No applicable case”. The sensing results from the radar 200 and the in-vehicle camera 220 and the corresponding case (case C or no corresponding case) are output to the detection determination unit 350, respectively.

  When only the sensing result by the radar 200 (corresponding to case A) is input from the sensing result selection unit 380, the detection determination unit 350 senses the detection threshold set in advance for the radar 200 and the sensing result by the radar 200. Obstacle detection detection is performed using the comparison result with the value. If it is determined that an obstacle has been detected as a result of the determination, the detection determination unit 350 outputs a detection result indicating that the vehicle has been detected to the notification control unit 360. In other words, from the characteristic of the radar 200 that it is easy to detect the vehicle, in a situation where the reliability of the sensing result by the radar 200 is high (Case A), the notification is made that the vehicle is detected.

  When only the sensing result from the in-vehicle camera 220 (corresponding to case B) is input from the sensing result selection unit 380, the detection determining unit 350 detects the detection threshold set in advance for the in-vehicle camera 220 and the sensing by the in-vehicle camera 220. Obstacle detection detection is executed using the result of comparison with the sensing value. As a result of the determination, if it is determined that an obstacle has been detected, the detection determination unit 350 outputs a detection result indicating that a pedestrian has been detected to the notification control unit 360. In other words, from the characteristic of the in-vehicle camera 220 that it is easy to detect a pedestrian, in a situation where the reliability of the sensing result by the in-vehicle camera 220 is high (Case B), the intention is to make a notification that the pedestrian is detected. .

  The detection determination unit 350 is set in advance for the radar 200 and the in-vehicle camera 220 when the sensing result from the radar 200 and the in-vehicle camera 220 (case C, corresponding to no case) is input from the sensing result selection unit 380. Obstacle detection detection is performed using a comparison result between the detection threshold and the sensing value of the sensing result of the radar 200 and the in-vehicle camera 220.

  As a result of the determination, if it is determined that an obstacle has been detected in “Case C”, the detection determination unit 350 outputs a detection result indicating that an obstacle has been detected to the notification control unit 360. In other words, in consideration of the low reliability of both the radar 200 and the in-vehicle camera 220, notification is performed without limiting the type of obstacle.

  As a result of the determination, if it is determined that an obstacle has been detected as “no corresponding case”, the detection determination unit 350 outputs a detection result indicating that the vehicle and the pedestrian have been detected to the notification control unit 360. In other words, the fact that the reliability of both the radar 200 and the in-vehicle camera 220 is high is taken into account, and notification that the vehicle and the pedestrian are detected is performed.

[Processing by Detection Device (Example 1)]
Next, the processing will be described using the detection apparatus according to the second embodiment with reference to FIG. FIG. 14 is a diagram illustrating a process flow of the detection device according to the second embodiment. The processing by the detection apparatus according to the second embodiment is different from the first embodiment in the points described below.

  That is, the sensing result selection unit 380 executes case classification of the reliability calculation result by the reliability calculation unit 330 (step S3), and selects a sensing result to be used for obstacle detection according to the classification result (step S4). ).

  Specifically, for example, when a reliability calculation result as illustrated in FIG. 7 or FIG. 9 is input from the reliability calculation unit 330, the sensing result selection unit 380 receives the sensing result from the radar 200 from the in-vehicle camera 220. Since this corresponds to case A having higher reliability than the sensing result (the radar has a higher number of reliability “high”), the reliability calculation result is classified as case A. Then, when classified into case A, the sensing result selection unit 380 selects the sensing result by the radar 200 as the sensing result to be used for obstacle detection, and only the sensing result by the radar 200 is sent to the detection determination unit 350. Output.

  In addition, the sensing result selection unit 380 classifies the reliability calculation result into the case B when the sensing result by the in-vehicle camera 220 corresponds to the case B having higher reliability than the sensing result by the radar 200. Then, the sensing result selection unit 380 selects the sensing result by the in-vehicle camera 220 as the sensing result to be used for obstacle detection when classified into the case B, and detects only the sensing result by the in-vehicle camera 220. Output to 350.

  For example, when a reliability calculation result as illustrated in FIG. 8 or 10 is input to the sensing result selection unit 380, the reliability of the sensing result by the radar 200 and the in-vehicle camera 220 is equal (“low” ”And“ high ”are the same number and the number of reliability“ low ”≧ the number of reliability“ high ”), which corresponds to case C. Therefore, the reliability calculation result is classified into case C.

  In addition, the sensing result selection unit 380 has, for example, the reliability of the sensing results obtained by the radar 200 and the in-vehicle camera 220 are equal, and unlike the case C, the number of reliability “low” <the number of reliability “high”. When it is, it classifies without the corresponding case.

  Then, the sensing result selection unit 380 selects the sensing result obtained by the radar 200 and the in-vehicle camera 220 as the sensing result to be used for obstacle detection when it is classified as “Case C” or “No applicable case”. The sensing results from the radar 200 and the in-vehicle camera 220 and the corresponding case (case C or no corresponding case) are output to the detection determination unit 350, respectively.

  When only the sensing result by the radar 200 (corresponding to case A) is input from the sensing result selection unit 380, the detection determination unit 350 senses the detection threshold set in advance for the radar 200 and the sensing result by the radar 200. Obstacle detection determination is executed using the comparison result with the value (step S5).

  As a result of the determination, when it is determined that the obstacle is detected in “Case A” (Yes in Step S5), the detection determination unit 350 outputs a detection result indicating that the vehicle is detected to the notification control unit 360, The notification control unit 360 notifies the driver that the vehicle has been detected (step S6). In other words, from the characteristic of the radar 200 that it is easy to detect the vehicle, in a situation where the reliability of the sensing result by the radar 200 is high (Case A), the notification is made that the vehicle is detected.

  Further, when the detection determination unit 350 determines that an obstacle has been detected in “Case B” as a result of the determination, the detection determination unit 350 notifies the detection result indicating that a pedestrian has been detected. 360, the notification control unit 360 notifies the driver that a pedestrian has been detected (step S6). In other words, from the characteristic of the in-vehicle camera 220 that it is easy to detect a pedestrian, in a situation where the reliability of the sensing result by the in-vehicle camera 220 is high (Case B), the intention is to make a notification that the pedestrian is detected. .

  Further, when the detection determination unit 350 determines that an obstacle has been detected in “Case C” as a result of the determination (Yes in step S5), the detection control unit notifies the detection result that an obstacle has been detected. 360, the notification control unit 360 notifies the driver that an obstacle has been detected (step S6). In other words, in consideration of the low reliability of both the radar 200 and the in-vehicle camera 220, notification is performed without limiting the type of obstacle.

  Further, when it is determined that the obstacle is detected as “no applicable case” as a result of the determination (Yes in step S5), the notification control unit 350 performs notification control of the detection result indicating that the vehicle and the pedestrian are detected. The information is output to the unit 360, and the notification control unit 360 notifies the driver that the vehicle has been detected (step S6). In other words, the fact that the reliability of both the radar 200 and the in-vehicle camera 220 is high is taken into account, and notification that the vehicle and the pedestrian are detected is performed.

[Effects of Example 2]
As described above, according to the second embodiment, the reliability of the sensing result can be taken into account when the obstacle is detected using the sensing result obtained by the radar mounted on the vehicle or the in-vehicle camera. Furthermore, according to the reliability of a sensing result, the sensing result employ | adopted for the detection of an obstruction can be selected, and the alerting | reporting content of a detection result can be changed according to the selected sensing result.

  In addition, the detection apparatus 300 which concerns on Example 1 or Example 2 mentioned above can also be applied to the vehicle control apparatus mounted in a vehicle in order to perform vehicle travel control.

  Now, the first embodiment has been described as an embodiment of the detection device according to the present invention, but the present invention may be implemented in various different forms other than the above-described embodiments. Therefore, in the following, other embodiments included in the present invention will be described.

(1) Device Configuration, etc. Each component of the detection device 300 shown in FIG. 3 or FIG. 13 is functionally conceptual, and does not necessarily need to be physically configured as illustrated. That is, the specific form of distribution / integration of the detection apparatus 300 is not limited to the illustrated one.

  For example, the detection threshold value changing unit 340 and the detection determination unit 350 shown in FIG. 3 are integrated, and the sensing result selection unit 380 and the detection determination unit 350 shown in FIG. It can be configured to be functionally or physically distributed / integrated in an arbitrary unit according to the usage situation. Furthermore, each processing function (see FIG. 12 or FIG. 14 or the like) performed in the detection device 300 is realized by a CPU and a program that is analyzed and executed by the CPU. It can be realized as hardware by wired logic.

(2) Detection processing program Various processes (for example, see FIG. 12 or FIG. 14) of the apparatus described in the above embodiment are executed by a computer system such as a personal computer or a workstation. Can be realized. In the following, an example of a computer that executes a detection processing program having the same function as that of the above-described embodiment will be described with reference to FIG. FIG. 15 is a diagram illustrating a computer that executes a detection processing program.

  As shown in the figure, the computer 400 is configured by connecting an input unit 410, an output unit 420, an HDD 430, a RAM 440 and a CPU 450 via a bus 500.

  Here, the input unit 410 receives input of various data from the user. The output unit 420 displays various information. The HDD 430 stores information necessary for the CPU 450 to execute various processes. The RAM 440 temporarily stores various information. The CPU 450 executes various arithmetic processes.

  As shown in FIG. 15, the HDD 430 stores in advance a detection processing program 431 that exhibits the same function as each processing unit of the detection apparatus shown in the above embodiment, and detection processing data 432. Yes. Note that the detection processing program 431 may be appropriately distributed and stored in a storage unit of another computer that is communicably connected via a network.

  Then, the CPU 450 reads out the detection processing program 431 from the HDD 430 and develops it in the RAM 440, so that the detection processing program 431 functions as a detection processing process 441 as shown in FIG.

  That is, the detection processing process 441 reads the detection processing data 432 and the like from the HDD 430, expands them in an area allocated to itself in the RAM 440, and executes various processes based on the expanded data and the like.

  Note that the detection process 441 includes the environment identification unit 320, the reliability calculation unit 330, the detection threshold value changing unit 340 (sensing result selection unit 380), the detection determination unit 350, and the notification of the detection apparatus 300 illustrated in FIG. This corresponds to each process executed in the control unit 360.

  Note that the above-described detection processing program 431 is not necessarily stored in the HDD 430 from the beginning. For example, a “portable physical medium” such as a flexible disk (FD), a CD-ROM, a DVD disk, a magneto-optical disk, and an IC card inserted into the computer 400, a public line, the Internet, a LAN, a WAN, etc. Each program may be stored in “another computer (or server)” connected to the computer 400 via the computer 400, and the computer 400 may read and execute each program from these.

(3) Detection Method The following detection method is realized by the apparatus described in the above embodiment.

  That is, when sensing results from each in-vehicle sensor are acquired, sensing results are obtained from each in-vehicle sensor that senses obstacles around the vehicle based on the environmental conditions at the time of the sensing result acquisition and the detected object indicated by the sensing result. The reliability of the sensing result by referring to the reliability storage means for storing the reliability of the sensing result by each in-vehicle sensor in association with the environmental condition at the time of detection or the sensing object to be sensed by each in-vehicle sensor. Whether or not an obstacle is detected using the reliability calculation step for calculating the degree (see, for example, steps S1 to S2 in FIG. 12), the reliability calculated by the reliability calculation step, and the sensing result. A detection method including a detection determination step for determination (see, for example, step S5 in FIG. 12) is realized. .

  As described above, the detection device, the detection method, and the vehicle control device according to the present invention are useful when detecting an obstacle existing around a traveling vehicle, and in particular, a radar or an in-vehicle camera mounted on the vehicle. It is suitable for taking into account the reliability of the sensing result when detecting an obstacle using the sensing result of

BRIEF DESCRIPTION OF THE DRAWINGS It is a figure for demonstrating the outline | summary and characteristic of the detection apparatus which concern on Example 1. FIG. BRIEF DESCRIPTION OF THE DRAWINGS It is a figure for demonstrating the outline | summary and characteristic of the detection apparatus which concern on Example 1. FIG. It is a figure which shows the structure of the detection apparatus which concerns on Example 1. FIG. It is a figure which shows the structural example of the reliability table 1 which concerns on Example 1. FIG. It is a figure which shows the structural example of the reliability table 2 which concerns on Example 1. FIG. It is a figure which shows the structural example of the reliability table 3 which concerns on Example 1. FIG. It is a figure which shows the reliability calculation result example which concerns on Example 1. FIG. It is a figure which shows the reliability calculation result example which concerns on Example 1. FIG. It is a figure which shows the reliability calculation result example which concerns on Example 1. FIG. It is a figure which shows the reliability calculation result example which concerns on Example 1. FIG. It is a figure which shows the example of case classification reference | standard of the reliability calculation result which concerns on Example 1. FIG. It is a figure which shows the flow of the process by the detection apparatus which concerns on Example 1. FIG. It is a figure which shows the structure of the detection apparatus which concerns on Example 1. FIG. FIG. 10 is a diagram illustrating a flow of processing by a detection device according to a second embodiment. It is a figure which shows the computer which performs a detection processing program.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Environmental sensor group 110 Illuminance sensor 120 Temperature sensor 130 Raindrop sensor 140 Gyro sensor 150 Infrastructure information device 200 Radar 210 Radar recognition part 220 Car-mounted camera 230 Image recognition part 240 Display 250 Speaker 300 Detection apparatus 310 Reliability table 1
311 Reliability table 2
312 Reliability table 3
320 environment identification unit 330 reliability calculation unit 340 detection threshold change unit 350 detection determination unit 360 notification control unit 380 sensing result selection unit 400 computer 410 input unit 420 output unit 430 HDD (Hard Disk Drive)
431 Detection processing program 432 Detection processing data 440 RAM (Random Access Memory)
441 Detection processing process 450 CPU (Central Processing Unit)
500 buses

Claims (6)

Confidence of sensing results by each in-vehicle sensor by associating it with environmental conditions when sensing results are acquired from each in-vehicle sensor that senses obstacles around the vehicle Reliability storage means for storing each degree,
When the sensing result by each of the in-vehicle sensors is acquired, the reliability storage unit is referred to based on the environmental condition at the time of the sensing result acquisition and the detected object indicated by the sensing result, and the reliability of the sensing result is calculated Reliability calculation means;
Detection determination means for determining whether or not the obstacle has been detected using the reliability calculated by the reliability calculation means and the sensing result;
A detection device comprising: Based on the reliability calculated by the reliability calculation means, a detection threshold set in advance for each type of in-vehicle sensor is determined to be lower than an initial value in order to determine whether or not the obstacle has been detected. Further comprising threshold changing means for
The detection determination unit detects the obstacle using a comparison result between the detection threshold set in advance and the detection threshold changed by the threshold change unit and the sensing result corresponding to each detection threshold. The detection device according to claim 1, wherein it is determined whether or not it has been performed.   The detection determination unit selects a sensing result determined by the reliability calculation unit as being highly reliable from each sensing result acquired from each of the in-vehicle sensors, and uses the selected sensing result. The detection device according to claim 1, wherein it is determined whether or not the obstacle has been detected.   Informing means for informing the vehicle driver according to the type of the in-vehicle sensor that is the acquisition source of the sensing result used for the determination when it is determined by the detection determining means that the obstacle is detected The detection device according to claim 3, further comprising: When the sensing results from each in-vehicle sensor are acquired, the sensing results are obtained from each in-vehicle sensor that senses obstacles around the vehicle based on the environmental conditions at the time of the sensing result acquisition and the detected object indicated by the sensing result. The reliability of the sensing result is referred to by referring to the reliability storage means for storing the reliability of the sensing result by each in-vehicle sensor in association with the environmental condition at the time or the sensing object to be sensed by each in-vehicle sensor. A reliability calculation step for calculating a degree;
A detection determination step of determining whether or not the obstacle is detected using the reliability calculated in the reliability calculation step and the sensing result;
The detection method characterized by including. Confidence of sensing results by each in-vehicle sensor by associating it with the environmental conditions when sensing results are obtained from each in-vehicle sensor that senses obstacles around the vehicle, or the sensing object to be sensed by each in-vehicle sensor Reliability storage means for storing each degree,
When the sensing result by each of the in-vehicle sensors is acquired, the reliability storage unit is referred to based on the environmental condition at the time of the sensing result acquisition and the detected object indicated by the sensing result, and the reliability of the sensing result is calculated. Reliability calculation means;
Detection determination means for determining whether or not the obstacle has been detected using the reliability calculated by the reliability calculation means and the sensing result;
A vehicle control device comprising:

Images