JP2019158762A - Device, method, and program for detecting abnormalities - Google Patents

Abstract

To provide a technique for appropriately detecting an abnormality of a sensor of a mobile body.SOLUTION: An abnormality detector detects an abnormality of a sensor equipped to a mobile body capable of measuring a distance to a mobile body. The abnormality detector includes: an evaluation reference information acquisition unit for acquiring evaluation reference information including distance reference data as distance data to be a reference for determining an abnormality of the sensor, the evaluation reference information being generated on the basis of information collected from a plurality of mobile bodies equipped to the sensor; a data calculation unit for calculating evaluation distance data as distance data to be compared with the distance reference data on the basis of the information acquired by the sensor when the mobile bodies move from a predetermined evaluation position; and a determination unit for determining whether the sensor is abnormal according to the evaluation reference information and the evaluation distance data.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an abnormality detection device, an abnormality detection method, and an abnormality detection system.

  2. Description of the Related Art Conventionally, a technique for improving the reliability of an image processing result by detecting a shift in the installation direction of an imaging unit while traveling is disclosed (see, for example, Patent Document 1).

  The image processing apparatus for a vehicle disclosed in Patent Literature 1 performs image processing for performing various image processing such as recognition processing for a preceding vehicle based on a plurality of imaging units that respectively capture the front of the vehicle and image data from the plurality of imaging units. Means. The image processing unit calculates a distance between the plurality of white lines when the white line recognition unit recognizes a plurality of white lines in the horizontal direction, and outputs the distance information between the white lines, and the calculation An imaging direction abnormality determination unit that determines that the imaging direction of the imaging unit is abnormal when the distance information between the white lines and the predetermined distance information between the white lines are different from each other by a predetermined value or more, and an alarm that notifies the abnormality An imaging direction abnormality alarm unit that outputs a signal to the display means.

JP-A-8-285534

  For example, a captured image captured while the vehicle is traveling may vary depending on various traveling conditions such as the weather when the vehicle travels, the condition of the road on which the vehicle travels, and the speed of the vehicle. In the conventional abnormality determination of the photographing direction of the imaging means, there is a possibility that accurate determination cannot be performed due to the influence of the fluctuation of the traveling condition.

  In view of the above problems, an object of the present invention is to provide a technique capable of appropriately detecting an abnormality of a sensor mounted on a moving body.

  In order to achieve the above object, an abnormality detection apparatus of the present invention is an abnormality detection apparatus that detects an abnormality of a sensor that is mounted on a moving body and enables distance measurement to an object. An evaluation reference information acquisition unit that acquires evaluation reference information that includes distance reference data that is generated based on information collected from a moving body and that serves as a reference for determining abnormality of the sensor; and A distance data calculation unit that calculates distance data for evaluation, which is distance data to be compared with the distance reference data, based on information obtained by the sensor when the evaluation point is moved, and the evaluation reference information and the evaluation And a determination unit that determines whether or not the sensor is abnormal based on the working distance data (first configuration).

  The abnormality detection apparatus of the first configuration further includes an environment data acquisition unit that acquires environment data for evaluation indicating an environment around the moving body when the moving body moves the predetermined evaluation location, and the evaluation The reference information includes environment data associated with the distance reference data, and the determination unit is configured to determine abnormality of the sensor (second configuration) in consideration of the evaluation environment data. Is preferred.

  The abnormality detection device having the first or second configuration further includes an operation data acquisition unit that acquires operation data for evaluation indicating an operation state of the moving body when the moving body moves the predetermined evaluation location. The evaluation reference information includes operation data associated with the distance reference data, and the determination unit determines an abnormality of the sensor in consideration of the evaluation operation data (third configuration). It is preferable that

  The abnormality detection device having any one of the first to third configurations further includes a sensor data acquisition unit that acquires sensor data indicating information on the sensor, and the evaluation reference information is associated with the distance reference data. Preferably, the sensor data is included, and the determination unit is configured to determine abnormality of the sensor (fourth configuration) in consideration of the sensor data.

  In order to achieve the above object, an abnormality detection method of the present invention is a method of detecting an abnormality of a sensor that is mounted on a moving body and enables measurement of a distance to an object using an abnormality detection device, and includes the sensor. An evaluation reference information acquisition step for acquiring evaluation reference information including distance reference data that is generated based on information collected from a plurality of moving bodies and serves as a reference for determining abnormality of the sensor; and the moving body A distance data calculating step of calculating distance data for evaluation, which is distance data to be compared with the distance reference data, based on information obtained by the sensor when moving a predetermined evaluation point, and the evaluation reference information A determination step of determining whether the sensor is abnormal based on the evaluation distance data (fifth configuration).

  In order to achieve the above object, the abnormality detection system of the present invention is an abnormality detection system including an information generation device and an abnormality detection device, and the information generation device is equipped with a sensor capable of measuring a distance to an object. A collection unit that collects information from a plurality of moving bodies that generates information on the basis of information collected by the collection unit and evaluation reference information that includes distance reference data that is distance data that serves as a reference for determining abnormality of the sensor The abnormality detection device is obtained by the sensor when the moving body moves a predetermined evaluation location, and the evaluation reference information acquisition unit that acquires the evaluation reference information. Based on the information, a distance data calculation unit that calculates distance data for evaluation which is distance data to be compared with the distance reference data, and based on the evaluation reference information and the evaluation distance data. Wherein the sensor has a configuration comprising, a determination section for determining whether or not an abnormality (sixth configuration) Te.

  In the abnormality detection system of the sixth configuration, the information generation device further includes an evaluation point generation unit that generates the predetermined evaluation point based on information collected from the mobile body (seventh configuration) It may be.

  The present invention provides a technique capable of appropriately detecting an abnormality of a sensor mounted on a moving body.

The block diagram which shows the structure of the abnormality detection system which concerns on embodiment of this invention The figure which shows an example of an evaluation criteria table The flowchart which shows an example of the production | generation procedure of the evaluation criteria table by an information generation device The flowchart which shows an example of the procedure of the abnormality determination by an abnormality detection apparatus The block diagram which shows the structure of the information generation apparatus of a modification. The flowchart which shows an example of the production | generation procedure of the evaluation location by the information generation apparatus of a modification

  Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the drawings.

  In the following, the case where the moving body is a vehicle will be described as an example. Vehicles widely include vehicles having wheels such as automobiles, trains, and trains. In addition, the mobile body of this invention is not restricted to a vehicle, For example, a ship, an automatic guided device, a robot, etc. may be sufficient.

  In the following description, a case where a sensor mounted on a moving body and capable of measuring a distance to an object is a camera will be described. The camera may be configured to be able to measure the distance to the object and may be a dedicated device for measuring the distance to the object, but may not be a dedicated device for measuring the distance. The camera may be a stereo camera or a monocular camera, for example. The sensor mounted on the moving body and capable of measuring the distance to the object is not limited to the camera, and may be, for example, a millimeter wave radar, LIDAR (Laser Imaging Detection and Ranging), or the like.

<1. Overview of anomaly detection system>
FIG. 1 is a block diagram showing a configuration of an abnormality detection system SYS according to an embodiment of the present invention. As shown in FIG. 1, the abnormality detection system SYS includes an information generation device 3 and an abnormality detection device 4.

  In the present embodiment, the information generation device 3 is a server device provided in the data center 1. The data center 1 is an organization that provides information to the vehicle 2 and collects information from the vehicle 2. The abnormality detection device 4 is mounted on the vehicle 2. However, the abnormality detection device 4 may be included in a server device provided in the data center 1, for example. Further, the abnormality detection device 4 may be configured such that, for example, a part of the configuration is provided in the server device and the remaining configuration is provided in the vehicle 2.

  The information generation device 3 and the abnormality detection device 4 are connected via a network 5 such as the Internet. The connection form between the information generating device 3 and the network 5 may be wireless or wired. The connection form between the abnormality detection device 4 and the network 5 is preferably wireless. Although only one vehicle 2 is shown in FIG. 1, it is preferable that the abnormality detection system SYS includes a plurality of vehicles 2 connected to the information generation device 3 via the network 5.

  The abnormality detection system SYS further includes a camera 6, a GPS (Global Positioning System) sensor 7, and a vehicle speed sensor 8. In the present embodiment, the camera 6, the GPS sensor 7 and the vehicle speed sensor 8 are mounted on the vehicle 2. The camera 6, the GPS sensor 7, and the vehicle speed sensor 8 output information to the abnormality detection device 4.

  The camera 6 is an example of a “sensor mounted on a moving body and capable of measuring a distance to an object” in the present invention. The camera 6 may be a front camera mounted in front of the vehicle 2, for example. The GPS sensor 7 may be configured to be mounted on, for example, a navigation device, a drive recorder, or a device in which these are integrated. The vehicle speed sensor 8 may be a magnetic sensor or an optical sensor that outputs the rotation of a rotor (not shown) provided on the wheel shaft of the vehicle 2 as a pulse signal. The vehicle speed sensor 8 outputs speed information of the vehicle 2 to the abnormality detection device 4 via, for example, a CAN (Controller Area Network) bus.

<2. Details of Information Generator>
The information generation device 3 generates evaluation reference information used when the abnormality detection device 4 determines an abnormality of the sensor (camera 6 in this embodiment). In the present embodiment, the evaluation criterion information is an evaluation criterion table 310.

  As shown in FIG. 1, the information generation device 3 includes a control unit 30 and a storage unit 31. The control unit 30 is a computer including, for example, a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), an input / output port, and the like (not shown). The storage unit 31 is a nonvolatile memory. For example, the control unit 30 performs processing and transmission / reception of information based on a program (not shown) stored in the storage unit 31. Note that the control unit 30 and the storage unit 31 may be configured to transmit and receive information using wired or wireless.

  The control unit 30 includes a collection unit 301, an evaluation criterion information generation unit 302, and a transmission unit 303. The functions of these units 301 to 303 included in the control unit 30 are realized by the CPU performing arithmetic processing according to a program. In other words, the information generation apparatus 3 includes a collection unit 301, an evaluation criterion information generation unit 302, and a transmission unit 303.

  The collection unit 301 collects information from a plurality of moving objects that are equipped with sensors that enable distance measurement to an object. In the present embodiment, the collection unit 301 collects information from a plurality of vehicles 2 on which the camera 6 is mounted. The information collected by the collecting unit 301 from each vehicle 2 includes distance data calculated based on information acquired by the camera 6 when each vehicle 2 travels through a predetermined evaluation location. The predetermined evaluation place is, for example, a place where at least two specific road installations are arranged in the traveling direction of the vehicle 2 with an interval, and a distance between the specific road installations by a sensor (camera 6 in the present embodiment). It is a place where can be measured. The installation on the road is, for example, a white line, a pole, a lighting device, a protective fence, or the like. That is, the distance data is, for example, a distance between two white lines, a distance between two poles, or the like.

  The predetermined evaluation location may be singular or plural. For example, each vehicle 2 is given in advance a location on a map related to a predetermined evaluation location from the data center 1, and collects various information when traveling through the predetermined evaluation location. For the information collection, a function of the abnormality detection device 4 described later in detail may be used.

  In the present embodiment, the information collected from each vehicle 2 by the collection unit 301 includes environmental data and operation data when each vehicle 2 travels through a predetermined evaluation location, and information related to the camera 6 that each vehicle 2 has. Sensor data is also included. The environmental data is data indicating the surrounding environment of the vehicle 2, and may be, for example, data such as a road condition (road pattern) on which the vehicle 2 travels, weather during travel, brightness during travel, and the like. The operation data is data indicating the operation state of the vehicle 2, and may be data such as vehicle speed and acceleration, for example. The sensor data may be data such as the type of the camera 6, the mounting position of the camera 6, and the type of the vehicle 2 to which the camera 6 is mounted.

  Note that at least one of the environmental data, the operation data, and the sensor data may not be included in the information that the collection unit 301 collects from each vehicle 2. Information collected by the collection unit 301 from each vehicle 2 may include target data related to a road installation object to be measured. The target data may be, for example, data on the shape, color, installation position, etc. of the installation object on the road.

  The evaluation reference information generation unit 302 generates evaluation reference information including distance reference data based on the information collected by the collection unit 301. FIG. 2 is a diagram illustrating an example of the evaluation criterion table 310. In this embodiment, the evaluation criterion information generation unit 302 generates an evaluation criterion table 310 as shown in FIG. 2 as evaluation criterion information. In the present embodiment, the evaluation criterion information generation unit 302 generates the evaluation criterion table 310 for each of the predetermined evaluation points described above. If there are a plurality of predetermined evaluation points, a plurality of evaluation criteria tables 310 are also generated. The distance reference data (corresponding to the recognition distance in FIG. 2) is distance data serving as a reference for determining an abnormality of the camera 6 in each vehicle 2. In the present embodiment, one distance value is assigned to each classification described later. That is, the distance reference data is composed of a plurality of distance values (data). In the table 310 shown in FIG. 2, each row corresponds to one classification.

  As shown in FIG. 2, the evaluation criterion table 310 includes sensor data associated with the distance criterion data. In the example shown in FIG. 2, the sensor data includes two items of camera type and camera position. If the type of the camera 6 is different, the classification is different. Moreover, if the mounting position of the camera 6 is different, the classification is different.

  In addition, the evaluation criterion table 310 includes operation data associated with the distance criterion data. In the example shown in FIG. 2, vehicle speed (one item) is included as operation data. That is, if the vehicle speed is different, the classification is different. In addition, although the vehicle speed of each classification may be a single value, it is preferable that the vehicle speed has a certain width in the form of Xkm / Hr to Ykm / Hr.

  In addition, the evaluation standard table 310 includes environment data associated with the distance standard data. In the example shown in FIG. 2, the environmental data includes three items: road pattern, weather, and time. The road pattern is classified based on, for example, the slope of the road and the curvature of the curve. For example, two roads having a difference in inclination exceeding a predetermined difference are classified as different from each other. The gradient and curvature of each classification may be a single value, but preferably has a certain width. The weather is classified according to the weather, such as sunny, cloudy, rainy, and snowy. The time is used as an index indicating the brightness around the vehicle 2. If the time is different, the classification is different. The time classification may be determined as appropriate, such as a classification based on seconds, a classification based on minutes, or a classification based on time.

  In the present embodiment, there are a plurality of data items associated with the distance reference data. Specifically, there are six data items associated with the distance reference data: camera type, camera position, vehicle speed, road pattern, weather, and time. If any one of a plurality of data items has a different content (numerical value or the like), it is classified as a different classification, and a distance value as distance reference data is assigned to each classification. Note that distance values may be the same between different classifications.

  Returning to FIG. 1, the transmission unit 303 performs a transmission process for transmitting information to the abnormality detection device 4. For example, the transmission unit 303 performs a process of transmitting the evaluation criterion table 310 and information that can be derived from the evaluation criterion table 310 to the abnormality detection device 4 according to a request from the abnormality detection device 4.

  FIG. 3 is a flowchart illustrating an example of a procedure for generating the evaluation criterion table 310 by the information generating apparatus 3. First, the collection unit 301 appropriately collects information necessary for generating the evaluation criterion table 310 from each vehicle 2 (step S1). The information generation device 3 distributes a predetermined evaluation location for collecting information for generating the evaluation reference table 310 to each vehicle 2 when, for example, there is a request from a predetermined time or each vehicle 2. Each vehicle 2 stores predetermined evaluation location data in a storage unit (not shown). When each vehicle 2 recognizes that its own vehicle travels a predetermined evaluation location based on information from the GPS sensor 7 or the like, the vehicle 6 captures an image by the camera 6, and a specific road installation object based on the captured image. The distance data is calculated. Each vehicle 2 acquires sensor data indicating information about the camera 6 and environment data and operation data when an image is captured by the camera 6. When each vehicle 2 obtains distance data, sensor data, environmental data, and operation data, each vehicle 2 transmits the data to the information generation device 3. In the present embodiment, in order to obtain such information, an abnormality detection device 4 described later in detail may be used.

  The collection unit 301 stores information collected from each vehicle 2 in the storage unit 31 (step S2). In addition, when the information which becomes the same classification | category is collected in multiple times from the same vehicle 2, it is good also as a structure which memorize | stores any one information in the memory | storage part 31. FIG. The information stored here may be deleted after being used for generating the evaluation reference table 310.

  The evaluation criterion information generation unit 302 checks whether or not a predetermined amount of information necessary for creating the evaluation criterion table 310 is accumulated in the storage unit 31 (step S3). As described above, the evaluation criteria table 310 is generated for each predetermined evaluation location. For example, when a predetermined number or more of information on the same evaluation location is accumulated in the storage unit 31, it may be determined that a predetermined amount of information is accumulated.

  If it is determined that a predetermined amount of information has been accumulated (Yes in step S3), the evaluation criterion information generation unit 302 generates an evaluation criterion table 310 (step S4). In the present embodiment, the evaluation criterion information generation unit 302 collects data groups having the same camera type, camera position, vehicle speed, road pattern, weather, and time as one classification. If all of a plurality of distance data are the same value when grouping as one classification, the value is set as a reference distance value for the classification. However, at least one of the plurality of distance data may be different from the others. In such a case, for example, an average value of a plurality of distance data is calculated, and this average value is set as a reference distance value. However, this is merely an example, and other configurations such as setting the median value of a plurality of distance data as a reference distance value may be employed. When the distance between specific installation objects on the road is known in advance, and there is a calculated distance value that is far from the distance, the average value may be calculated by excluding the distance value that is far away.

  If it is determined that the predetermined amount of information is not accumulated (No in step S3), the evaluation criterion information generation unit 302 does not generate the evaluation criterion table 310. In this case, returning to step S1, information collection for generating the evaluation criterion table 310 is performed.

  Note that it is preferable that the information generation apparatus 3 also collects information about the evaluation part that generated the evaluation criterion table 310 from the vehicle 2 and updates the evaluation criterion table 310 even after the evaluation criterion table 310 is generated. Thereby, the evaluation criterion table 310 based on more information can be generated, and the reliability of the evaluation result using the evaluation criterion table 310 can be improved.

<3. Details of Abnormality Detection Device>
The anomaly detection device 4 detects an anomaly of a sensor that is mounted on a moving body and enables distance measurement to an object. In the present embodiment, the abnormality detection device 4 detects an abnormality of the camera 6 mounted on the vehicle 2. The abnormalities include, for example, abnormal mounting states, abnormal data processing, abnormal data transmission / reception, and the like.

  The abnormality detection device 4 is a computer including a CPU, a ROM, a RAM, a HDD (Hard Disk Drive), an input / output port, and the like (not shown), for example. As shown in FIG. 1, the abnormality detection device 4 includes a measurement timing monitoring unit 41, an image acquisition unit 42, an evaluation reference information acquisition unit 43, an environment data acquisition unit 44, an operation data acquisition unit 45, and sensor data. The acquisition part 46, the distance data calculation part 47, the transmission part 48, and the determination part 49 are provided. The function of each part 41-49 is implement | achieved when CPU performs arithmetic processing according to the program memorize | stored in ROM.

  The measurement timing monitoring unit 41 monitors the timing for starting measurement for detecting an abnormality of the sensor (camera 6 in this embodiment). In the present embodiment, the measurement timing monitoring unit 41 determines that it is the timing to start measurement when the vehicle 2 (hereinafter referred to as the host vehicle) equipped with the abnormality detection device 4 travels in a predetermined evaluation location. To do. The abnormality detection device 4 recognizes a predetermined evaluation location by information distributed from the information generation device 3. The measurement timing monitoring unit 41 determines, for example, whether or not the host vehicle travels a predetermined evaluation location based on map information and GPS information stored in a storage unit such as an HDD.

  The image acquisition unit 42 acquires information captured by the camera 6. The evaluation reference information acquisition unit 43 is generated based on information collected from a plurality of mobile bodies equipped with sensors, and acquires evaluation reference information including distance reference data. In the present embodiment, the evaluation criterion information acquisition unit 43 acquires the evaluation criterion table 310 generated by the information generation device 3.

  The environment data acquisition unit 44 acquires evaluation environment data indicating the surrounding environment of the moving body when the moving body moves in a predetermined evaluation location. In the present embodiment, the environment data acquisition unit 44 acquires evaluation environment data indicating the surrounding environment of the host vehicle when the host vehicle travels through a predetermined evaluation location. In the present embodiment, the environment data acquisition unit 44 acquires a road pattern, weather, and time as information indicating brightness. The road pattern can be acquired from, for example, a GPS sensor and map information. The weather can be acquired from, for example, an external device connected via the network 5. The time can be obtained from a GPS sensor, for example.

  The operation data acquisition unit 45 acquires operation data for evaluation indicating the operation state of the moving body when the moving body moves in a predetermined evaluation location. In the present embodiment, the motion data acquisition unit 45 acquires motion data for evaluation indicating the motion state of the host vehicle when the host vehicle travels through a predetermined evaluation location. In the present embodiment, the environmental data acquisition unit 44 acquires the vehicle speed of the host vehicle. The operation data acquisition unit 45 can acquire the vehicle speed of the host vehicle via a CAN bus (not shown).

  The sensor data acquisition unit 46 acquires sensor data indicating information related to the sensor. In the present embodiment, the sensor data acquisition unit 46 acquires the type of the camera 6 mounted on the host vehicle and the mounting position of the camera 6. Such information is memorize | stored in the memory | storage part of the abnormality detection apparatus 4 at the time of factory shipment of the vehicle 2, for example. These pieces of information may be obtained from the data center 1 by transmitting a vehicle code or the like to the data center 1, for example.

  The distance data calculation unit 47 calculates evaluation distance data based on information obtained by the sensor when the moving body moves in a predetermined evaluation location. The evaluation distance data is distance data to be compared with the distance reference data. In the present embodiment, the distance data calculation unit 47 calculates a distance value between specific on-road objects based on shooting information shot by the camera 6 when the host vehicle travels through a predetermined evaluation location. As described above, the specific installation on the road is, for example, a white line, a pole, a lighting device, a protective fence, or the like. As a method of calculating the evaluation distance data based on the photographed image, a known method such as a method using the principle of triangulation may be used.

  The transmission unit 48 performs transmission processing for transmitting information to the information generating device 3. The transmission unit 48 may transmit, for example, information acquired by the various acquisition units 42 to 46, a calculation result in the distance data calculation unit 47, a determination result in the determination unit 49, and the like to the data center 1.

  The determination unit 49 determines whether the sensor is abnormal based on the evaluation reference information acquired by the evaluation reference information acquisition unit 43 and the evaluation distance data calculated by the distance data calculation unit 47. In the present embodiment, the determination unit 49 determines the abnormality of the camera 6 based on the evaluation reference table 310 and the evaluation distance data. According to this embodiment, the abnormality of the camera 6 of the host vehicle can be determined by comparing with reference data obtained based on information from the plurality of vehicles 2 that are traveling. For this reason, the reliability of the abnormality determination of the camera 6 can be improved.

  The determination unit 49 preferably determines the abnormality of the sensor in consideration of the evaluation environment data acquired by the environment data acquisition unit 44. The determination unit 49 preferably determines the abnormality of the sensor in consideration of the evaluation operation data acquired by the operation data acquisition unit 45. The determination unit 49 preferably determines the abnormality of the sensor in consideration of the sensor data acquired by the sensor data acquisition unit 46.

  In this embodiment, the abnormality determination of the camera 6 is performed in consideration of all of the evaluation environment data, the evaluation operation data, and the sensor data. However, this is an example. For example, in addition to these data, it is good also as a structure by which the above-mentioned target data is considered and abnormality determination is performed. In this case, the evaluation criterion table 310 may include target data associated with the distance criterion data. Further, for example, any one or only two of the evaluation environment data, the evaluation operation data, and the sensor data may be considered in the abnormality determination. Depending on circumstances, the configuration may be such that none of the evaluation environment data, the evaluation operation data, and the sensor data is taken into consideration.

  By performing abnormality determination by considering at least one of the environmental data for evaluation, the operation data for evaluation, and the sensor data, the abnormality determination of the camera 6 is performed by aligning the states between the comparison targets as much as possible. Can do. For this reason, it becomes possible to perform abnormality determination with higher reliability. In addition, since performance evaluation can be performed in consideration of the use conditions of the camera 6 (sensor), data for improving the product can be obtained.

  FIG. 4 is a flowchart illustrating an example of the abnormality determination procedure performed by the abnormality detection device 4. The abnormality detection device 4 starts to operate when, for example, a vehicle power supply (ACC or the like) is turned on. First, the measurement timing monitoring unit 41 monitors whether or not it is time to perform measurement for detecting an abnormality of the camera 6 (step S11). The measurement timing monitoring unit 41 continues monitoring until the measurement timing comes. In the present embodiment, the measurement timing is a timing at which the host vehicle travels through a predetermined evaluation location.

  When the measurement timing monitoring unit 41 detects the measurement timing (Yes in step S11), a captured image and various types of information are acquired (step S12). Specifically, the image acquisition unit 42 acquires a captured image captured by the camera 6 when the host vehicle travels through a predetermined evaluation location. The photographed image includes a specific road installation. The specific road installation is, for example, a white line extending left and right in front of the host vehicle.

  In addition, the environment data acquisition unit 44 acquires environment data for evaluation. More specifically, the environment data acquisition unit 44 acquires a road pattern of a predetermined evaluation location where the host vehicle travels from map information or the like stored in the storage unit. The environment data acquisition unit 44 acquires weather information from the external device via the network 5 when the host vehicle travels through a predetermined evaluation location. The environment data acquisition unit 44 acquires the time at which the host vehicle travels through a predetermined evaluation location from GPS information or the like.

  Further, the motion data acquisition unit 45 acquires motion data for evaluation. More specifically, the operation data acquisition unit 45 acquires the vehicle speed of the host vehicle when the host vehicle travels through a predetermined evaluation location via the CAN bus. Further, the sensor data acquisition unit 46 acquires sensor data. More specifically, the sensor data acquisition unit 46 acquires the type of the camera 6 mounted on the host vehicle and the mounting position of the camera 6 from a storage unit (not shown).

  Next, the distance data calculation unit 47 calculates evaluation distance data (step S13). More specifically, the distance data calculation unit 47 calculates a distance value between specific on-road objects based on the captured image acquired by the image acquisition unit 42. The distance data calculation unit 47 calculates, for example, a distance value between two white lines extending left and right in front of the host vehicle.

  Next, the evaluation criterion information acquisition unit 43 acquires evaluation criterion information (step S14). More specifically, the evaluation criterion information acquisition unit 43 acquires the evaluation criterion table 310 generated for a predetermined evaluation point where the host vehicle has traveled from the information generation device 3 via the network 5. The evaluation criterion information acquisition unit 43 may acquire the entire evaluation criterion table 310 or may acquire a part of the evaluation criterion table 310. For example, the evaluation reference information acquisition unit 43 may be configured to acquire only data relating to the same type of camera as the camera 6 of the host vehicle.

  The evaluation criterion information acquisition unit 43 may be configured to acquire the evaluation criterion information prior to step S13. For example, the evaluation criterion information acquisition unit 43 may acquire evaluation criterion information at the same timing as step S12. In some cases, the evaluation reference table 310 may not be generated for a predetermined evaluation point. In such a case, a configuration that does not determine abnormality may be used.

  Next, the determination part 49 performs abnormality determination (step S15). The determination unit 49 finds a classification that matches the various types of evaluation data acquired from the environment data acquisition unit 44, the operation data acquisition unit 45, and the sensor data acquisition unit 46 from the evaluation reference table 310. The determination unit 49 compares the recognized distance of the found classification with the evaluation distance data (distance value) calculated by the distance data calculation unit 47. When the absolute value of the difference between the two is less than or equal to a predetermined threshold, it is determined that there is no abnormality, and when it exceeds the predetermined threshold, it is determined that there is an abnormality. The predetermined threshold value may be determined based on, for example, results of experiments or simulations.

  Note that the determination unit 49 may not be able to find out from the evaluation criterion table 310 a classification that matches the various types of evaluation data acquired from the acquisition units 44 to 46. In such a case, it may be configured that abnormality determination is not performed because abnormality determination is impossible.

  The result of the abnormality determination may be notified by, for example, a notification device (not shown) included in the host vehicle. The notification by the notification device may be configured to be performed only when an abnormality is detected. The notification device may be, for example, a display device such as a liquid crystal display, an audio output device such as a speaker, or the like. When an abnormality is detected by the abnormality determination, the automatic driving function executed using the sensor (camera 6 in the present embodiment) may be stopped or disabled.

  According to this embodiment, the abnormality determination of the camera 6 can be performed while the host vehicle is traveling. Moreover, according to this embodiment, since abnormality determination is performed by comparison with the reference data acquired from the several vehicle drive | worked on the same driving conditions, the reliability of abnormality determination can be improved. According to this embodiment, since abnormality determination is performed in consideration of a plurality of types of traveling conditions, it is easy to identify the cause of abnormality when an abnormality is detected. According to the present embodiment, since it is not necessary to separately prepare a comparison device for detecting an abnormality, it is possible to easily check the abnormality.

  Note that data acquired or calculated for abnormality determination may be sent to the information generation device 3 by the transmission unit 48. The information generation device 3 may use the data sent by the transmission unit 48 as data for generating or updating the evaluation criterion table 310.

  Further, the configuration in which all the components of the abnormality detection device 4 are mounted on the vehicle 2 has been described above. As described above, for example, the abnormality detection device 4 may have a configuration in which some components are provided in the vehicle 2 and the remaining components are provided in a server device including the information generation device 3. For example, the measurement timing monitoring unit 41, the image acquisition unit 42, the environmental data acquisition unit 44, the operation data acquisition unit 45, the sensor data acquisition unit 46, and the transmission unit 48 are provided in the vehicle 2, and the evaluation reference information acquisition unit 43 and The determination unit 49 may be provided in the server device. That is, the information acquired on the vehicle 2 side may be transmitted to the server device, and the abnormality determination may be performed by the server device. In this case, the determination result may be transmitted to the vehicle 2 side.

<4. Modification>
FIG. 5 is a block diagram illustrating a configuration of an information generation device 3A according to a modification. The information generation apparatus 3 </ b> A according to the modification further includes an evaluation point generation unit 304. The function of the evaluation point generation unit 304 included in the control unit 30A is realized by the CPU executing arithmetic processing according to a program. The evaluation point generation unit 304 generates a predetermined evaluation point based on information collected from the mobile object. In this modification, the evaluation point generation unit 304 generates a predetermined evaluation point based on information collected by the collection unit 301 from the vehicle 2 on which the camera 6 is mounted.

  FIG. 6 is a flowchart illustrating an example of a procedure for generating an evaluation location by the information generation apparatus 3A according to the modification. The evaluation location generator 304 confirms whether or not a predetermined captured image has been acquired from the vehicle 2 on which the camera 6 is mounted (step S21). The predetermined photographed image is an image obtained by photographing a place other than the already set evaluation place. Each vehicle 2 captures an image with the camera 6 based on conditions specified in advance for the information generating device 3A to generate an evaluation location, and transmits the captured image to the information generating device 3A. For example, the information generation device 3A designates a place where information for generating the evaluation reference table 310 can be easily collected from the plurality of vehicles 2 as a shooting location. For example, a straight part on a main road with a large traffic volume is designated as a photographing part.

  The evaluation location generator 304 confirms whether or not there is a distance measurement mark candidate in the captured image acquired to generate the evaluation location (step S22). The mark candidates for distance measurement are, for example, road installations such as white lines, poles, and guide signs that are arranged at intervals in the traveling direction of the vehicle 2. The evaluation location generation unit 304 extracts a distance measurement mark candidate from the acquired image by a method such as pattern matching, for example.

  When the mark candidate for distance measurement is extracted from the captured image (Yes in step S22), the evaluation location generating unit 304 recognizes the location where the captured image is captured as the evaluation location (step S23). When it is recognized as an evaluation location, the information (information including the location of the evaluation location and the type of mark) is notified to each vehicle 2, and the evaluation reference table 310 at the evaluation location is generated according to the flow shown in FIG. The

  After the evaluation location is recognized, the evaluation location generation unit 304 further confirms whether or not to continue generating the evaluation location (step S24). Unless there is a command to end the generation of the evaluation part, the process returns to step S21 and the generation process of the evaluation part is continued. When the mark candidate for distance measurement is not extracted from the captured image (No in step S22), the evaluation location generating unit 304 does not recognize the location where the captured image is captured as an evaluation location. Also in this case, as long as there is no instruction to end the generation of the evaluation portion, the evaluation portion generation unit 304 returns to step S21 and continues the evaluation portion generation process.

  According to this modification, since the number of evaluation points can be automatically increased, it is possible to quickly detect abnormality of the sensor (camera 6 in this example).

<5. Notes>
The configurations of the embodiments and modifications shown in this specification are merely examples of the present invention. The configuration of the embodiment and the modification may be changed as appropriate without departing from the technical idea of the present invention. In addition, a plurality of embodiments and modifications may be implemented in combination within a possible range.

  In the above, all or a part of the functions described as being realized by software by executing the program may be realized by an electrical hardware circuit. In addition, the function described as one block in the above may be realized by cooperation of software and hardware.

  In the above description, the function described as one block is not necessarily realized by a single physical element, but may be realized by distributed physical elements. The functions described as a plurality of blocks may be realized by a single physical element. Further, the function of each block may be realized as a whole by distributing the function of each block to a plurality of devices and exchanging information by communication between these devices.

2 ... Vehicle (moving body)
3, 3A ... Information generation device 4 ... Abnormality detection device 6 ... Camera (sensor)
43 ... Evaluation criteria information acquisition unit 44 ... Environmental data acquisition unit 45 ... Motion data acquisition unit 46 ... Sensor data acquisition unit 47 ... Distance data calculation unit 49 ... Determination unit 301 ... Collection unit 302... Evaluation criterion information generation unit 304... Evaluation point generation unit 310 Evaluation criterion table SYS abnormality detection system

Claims (7)

An anomaly detection device that detects an anomaly of a sensor that is mounted on a moving body and enables distance measurement to an object,
An evaluation reference information acquisition unit that generates evaluation reference information that includes distance reference data that is generated based on information collected from a plurality of moving bodies on which the sensor is mounted and that serves as a reference for determining abnormality of the sensor. When,
A distance data calculation unit that calculates distance data for evaluation, which is distance data to be compared with the distance reference data, based on information obtained by the sensor when the moving body moves in a predetermined evaluation location;
A determination unit that determines whether the sensor is abnormal based on the evaluation reference information and the evaluation distance data;
An abnormality detection device comprising: An environment data acquisition unit that acquires environment data for evaluation indicating an environment around the mobile body when the mobile body moves the predetermined evaluation location;
The evaluation criteria information includes environmental data associated with the distance criteria data,
The abnormality determination device according to claim 1, wherein the determination unit determines abnormality of the sensor in consideration of the environment data for evaluation. An operation data acquisition unit that acquires operation data for evaluation indicating an operation state of the mobile object when the mobile object moves the predetermined evaluation location;
The evaluation reference information includes operation data associated with the distance reference data,
The abnormality detection device according to claim 1, wherein the determination unit determines abnormality of the sensor in consideration of the evaluation operation data. A sensor data acquisition unit that acquires sensor data indicating information about the sensor;
The evaluation reference information includes the sensor data associated with the distance reference data,
The abnormality detection device according to claim 1, wherein the determination unit determines abnormality of the sensor in consideration of the sensor data. A method of detecting an abnormality of a sensor that is mounted on a moving object and enables distance measurement to an object by an abnormality detection device,
An evaluation reference information acquisition step for acquiring evaluation reference information including distance reference data, which is generated based on information collected from a plurality of moving bodies on which the sensor is mounted, and is reference data for determining abnormality of the sensor. When,
A distance data calculation step of calculating distance data for evaluation, which is distance data to be compared with the distance reference data, based on information obtained by the sensor when the movable body moves in a predetermined evaluation location;
A determination step of determining whether or not the sensor is abnormal based on the evaluation reference information and the evaluation distance data;
An abnormality detection method comprising: An anomaly detection system including an information generation device and an anomaly detection device,
The information generation device includes:
A collection unit that collects information from a plurality of moving objects equipped with sensors that enable distance measurement to an object;
Based on the information collected by the collection unit, an evaluation reference information generation unit that generates evaluation reference information including distance reference data that is distance data serving as a reference for determining abnormality of the sensor;
With
The abnormality detection device is:
An evaluation criterion information acquisition unit for acquiring the evaluation criterion information;
A distance data calculation unit that calculates distance data for evaluation, which is distance data to be compared with the distance reference data, based on information obtained by the sensor when the moving body moves in a predetermined evaluation location;
A determination unit that determines whether the sensor is abnormal based on the evaluation reference information and the evaluation distance data;
An anomaly detection system comprising:   The abnormality detection system according to claim 6, wherein the information generation device further includes an evaluation point generation unit that generates the predetermined evaluation point based on information collected from the mobile body.

Images