Control device, control method, program, and memory medium

Abstract

To provide a control device that can suitably measure peripheral environment even when the detection range of an outside-monitoring sensor fluctuates due to an accident or the like.SOLUTION: A control device 6 includes: acquiring an output signal from a lidar unit 7 that can detect a ground object existing around a vehicle; and calculating a reference angle θtag, which represents an inclination angle to the detection range of the lidar unit 7, in terms of the ground object recognized based on the output signal of the lidar unit 7. The control device 6 also controls the detection range of the lidar unit 7 based on the reference angle θtag, and inclination-angle information that is memorized in a memory unit 2 and represents an angle of the ground object to the road surface.SELECTED DRAWING: Figure 13

Description

[Technical field]

[0001] The present invention relates to a technique for adjusting an illumination unit for object detection. [Background technology]

[0002] Conventionally, there is known a technique for using a measurement result by taking into consideration an error that occurs in a measurement section due to an alignment error, etc. For example, Patent Document 1 discloses a moving body that includes a sensor that measures the position of an object in a measurement area where an object that needs to be measured during the traveling of the vehicle is present, and a control device that corrects an error caused by an error in the mounting angle of the sensor to control the movement of the vehicle. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2011-221957 A Summary of the Invention [Problem to be solved by the invention]

[0004] In autonomous driving, it is necessary to recognize the vehicle's surrounding environment using external sensors such as LIDAR. However, if the alignment of the external sensor is misaligned due to an accident or other reason, the results of the external sensor's recognition of the surrounding environment will contain errors, resulting in a decrease in the reliability of the external sensor.

[0005] The present invention has been made to solve the above-mentioned problems, and its main object is to provide a control device that can appropriately measure the surrounding environment even if the detection range of the external sensor fluctuates due to an accident or the like. [Means for solving the problem]

[0006] The invention described in the claims is a control device comprising a first acquisition unit that acquires an output signal from a detection device capable of detecting features present around a moving body, a second acquisition unit that acquires first information regarding the inclination of the feature recognized based on the output signal of the detection device relative to the detection range of the detection device, and a control unit that controls the detection range of the detection device based on the first information and second information regarding the angle of the feature relative to the road surface stored in a memory unit.

[0007] The invention described in the claims is a control method executed by a control device, comprising a first acquisition step of acquiring an output signal from a detection device capable of detecting features present around a moving body, a second acquisition step of acquiring first information regarding the inclination of the feature recognized based on the output signal of the detection device relative to the detection range of the detection device, and a control step of controlling the detection range of the detection device based on the first information and second information regarding the angle of the feature relative to the road surface stored in a memory unit.

[0008] The invention described in the claims is also characterized in that it is a program executed by a computer, comprising a first acquisition unit that acquires an output signal from a detection device capable of detecting features present around a moving body, a second acquisition unit that acquires first information regarding the inclination of the feature recognized based on the output signal of the detection device relative to the detection range of the detection device, and causes the computer to function as a control unit that controls the detection range of the detection device based on the first information and second information regarding the angle of the feature relative to the road surface stored in a memory unit. [Brief description of the drawings]

[0009] [Figure 1] 1 shows a schematic configuration of a measurement system. [Diagram 2] FIG. 2 is a block diagram illustrating the process of a control device. [Diagram 3] 1 shows a block configuration of a rider unit. [Figure 4] An example of scanner placement is shown. [Diagram 5] 2 shows a schematic configuration example of an optical transmitting / receiving unit. [Figure 6] The three-axis coordinate system detected by the three-axis sensor installed in the scanner housing is shown. [Figure 7] The positional relationship between the scannable range and the actual scanning range is shown. [Figure 8] 13 is a flowchart showing a procedure of an alignment process. [Figure 9] 4 is a flowchart showing a processing procedure during automatic driving. [Figure 10] 13 shows an example of a data structure of a map DB according to the second embodiment. [Figure 11] FIG. 2 is a diagram showing the relationship between features and tilt angles indicated by tilt information of the feature information. [Figure 12] 1 is a diagram showing the correspondence between the actual scanning range of a scanner and a scene overlapping the actual scanning range; [Figure 13] 10 is a flowchart showing a processing procedure of calibration based on a feature according to the second embodiment. [Figure 14] 13 shows a data structure of a map DB according to a third embodiment. [Figure 15] 13 is an example of a system configuration according to a modified example. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] According to a preferred embodiment of the present invention, the control device includes a first acquisition unit that acquires an output signal from a detection device capable of detecting features present around a mobile body, a second acquisition unit that acquires first information on the inclination of the feature recognized based on the output signal of the detection device with respect to the detection range of the detection device, and a control unit that controls the detection range of the detection device based on the first information and second information on the angle of the feature with respect to the road surface stored in a memory unit. According to this aspect, the control device can suitably identify a deviation of the detection range of the detection device based on the first information on the measured inclination of the feature and the second information on the inclination of the feature stored in the memory unit, and can appropriately adjust the detection range.

[0011] In one aspect of the control device, the first information is information on an angle between a line indicating the detection range of the detection device and a predetermined portion indicating the contour of the feature recognized based on the output signal, and the second information is information on an inclination angle of the predetermined portion constituting the contour of the feature with respect to the road surface. According to this aspect, the control device can identify a deviation of the detection range of the detection device based on the inclination of the predetermined portion indicating the contour of the feature, and appropriately adjust the detection range.

[0012] In another aspect of the control device, the first information is information on an angle between a line indicating the detection range of the detection device and a line indicating the symmetry axis of the contour of the feature recognized based on the output signal, and the second information is information on an inclination angle of the line with respect to the road surface. According to this aspect, the control device can identify a deviation of the detection range of the detection device based on the inclination of the line passing through the center of gravity of the feature, and appropriately adjust the detection range.

[0013] In another aspect of the control device, the control unit controls the detection range of the detection device based on first information on the inclination of a feature that is perpendicular to the road surface, among features recognized based on the output signal of the detection device, with respect to the detection range, and second information on the angle of the feature with respect to the road surface. In this aspect, the control device identifies a deviation of the detection range of the detection device for features whose inclination is relatively easy to measure, and can therefore more accurately adjust the detection range.

[0014] In another aspect of the control device, the control unit controls the detection range of the detection device based on first information on the inclination of a feature that does not vary depending on the environment, among features recognized based on the output signal of the detection device, with respect to the detection range, and second information on the angle of the feature with respect to the road surface. In this aspect, the control device identifies a deviation of the detection range of the detection device for features that are unlikely to cause measurement errors when measuring the inclination, and can therefore more accurately adjust the detection range.

[0015] According to another preferred embodiment of the present invention, a control method executed by a control device includes a first acquisition step of acquiring an output signal from a detection device capable of detecting features present around a moving body, a second acquisition step of acquiring first information on an inclination of the feature recognized based on the output signal of the detection device with respect to the detection range of the detection device, and a control step of controlling the detection range of the detection device based on the first information and second information on an angle of the feature with respect to a road surface stored in a storage unit. By executing this control method, the control device can preferably identify a deviation of the detection range of the detection device and appropriately adjust the detection range.

[0016] According to yet another preferred embodiment of the present invention, a program executed by a computer causes the computer to function as a first acquisition unit that acquires an output signal from a detection device capable of detecting features present around a moving body, a second acquisition unit that acquires first information on the inclination of the feature recognized based on the output signal of the detection device with respect to the detection range of the detection device, and a control unit that controls the detection range of the detection device based on the first information and second information on the angle of the feature with respect to the road surface stored in a storage unit. By executing this program, the computer can preferably identify a deviation in the detection range of the detection device and accurately adjust the detection range. Preferably, the program is stored in a storage medium. EXAMPLES

[0017] First to third preferred embodiments of the present invention will now be described with reference to the drawings.

[0018] <First Example> [Overall configuration] 1 is a schematic diagram of a measurement system 100 according to a first embodiment. The measurement system 100 is a system that performs measurements for automatic driving of a vehicle (not shown), and mainly includes an input unit 1, a storage unit 2, a sensor unit 3, a notification unit 4, a communication unit 5, and a control device 6. The control device 6 and other elements are configured to be capable of data communication based on a predetermined communication protocol.

[0019] The input unit 1 is a button, a touch panel, a remote controller, a voice input device, or the like that is operated by a user, and receives various inputs such as switching between automatic and manual driving, and whether or not calibration, which will be described later, needs to be performed.

[0020] The storage unit 2 stores a program executed by the control device 6 and information required for the control device 6 to execute a predetermined process. In this embodiment, the storage unit 2 stores a map DB 20. The map DB 20 may be used for navigation during manual driving of the vehicle, or may be used for driving control of the vehicle during automatic driving of the vehicle. The map DB 20 includes road data, facility information, and feature information related to features to be detected by the lidar unit 7 described later. The storage unit 2 also stores alignment information "IA" that records the arrangement angle of the lidar unit 7 to the vehicle after alignment described later. Note that "alignment" refers to adjustments made to the lidar unit 7 when the lidar unit 7 is manufactured or when the lidar unit 7 is attached to the vehicle.

[0021] The sensor unit 3 is composed of an internal sensor for detecting the state of the vehicle and an external sensor for recognizing the surrounding environment of the vehicle. The sensor unit 3 includes a Lidar (Light Detection and Ranging, or Laser Illuminated Detection and Ranging) unit 7, an attitude sensor 8, and an impact sensor 9.

[0022] As described later, the rider unit 7 has a plurality of scanning units (scanners) installed in different orientations. An example of the configuration of the rider unit 7 will be described later with reference to Figs. 3 to 5. The rider unit 7 functions as a detection device having a predetermined detection range. The attitude sensor 8 is, for example, a three-axis acceleration sensor or a strain gauge, and is provided as a sensor for detecting the attitude of the rider unit 7. The attitude sensor 8 may be provided in the rider unit 7, or may be further provided in the vehicle as a sensor for detecting the attitude of the vehicle. The impact sensor 9 is a sensor that detects an impact on the vehicle, and may be an acceleration sensor or a sensor that generates an opening signal for an airbag based on an impact. The impact sensor 9 may be provided in the rider unit 7.

[0023] The notification unit 4 is, for example, a display, a speaker, or the like that performs output under the control of the control device 6. The communication unit 5 performs data communication with an external device under the control of the control device 6.

[0024] The control device 6 includes a CPU that executes a program and controls the entire measurement system 100. In this embodiment, when the control device 6 determines that the alignment of the rider unit 7 has shifted based on the output of the sensor unit 3 such as the attitude sensor 8, the control device 6 performs calibration of the rider unit 7 or switching to manual driving. Here, the term "calibration" refers to an adjustment performed on the rider unit 7 after the rider unit 7 is attached to the vehicle. The control device 6 may be an ECU (Electronic Control Unit) that automatically controls the driving of the vehicle, or may be a CPU of an on-board device that transmits a control signal to the ECU to instruct the ECU to switch between automatic driving and manual driving. In another example, the control device 6 may be configured as a part of the rider unit 7. The control device 6 may also perform highly accurate vehicle position estimation based on the output of the sensor unit 3 and the map DB 20.

[0025] [Function block] 2 is a block diagram showing an outline of the processing of the control device 6 in the first embodiment. As shown in FIG.

[0026] The detection unit 61 acquires a detection signal from the attitude sensor 8 provided in the rider unit 7 as measurement information indicating the current arrangement angle of each scanner of the rider unit 7 from the attitude sensor 8. Then, the detection unit 61 detects an alignment deviation of the rider unit 7 by referring to alignment information IA indicating the arrangement angle of each scanner at the time of alignment stored in the storage unit 2 and comparing it with the above-mentioned measurement information. In addition, the detection unit 61 acquires a detection signal generated by the impact sensor 9 when vibration of a predetermined degree or more occurs in the vehicle from the impact sensor 9 as impact information on the vehicle. Then, when the detection unit 61 detects an alignment deviation of the rider unit 7 based on the measurement information of the attitude sensor 8 or when the detection unit 61 receives impact information from the impact sensor 9, the detection unit 61 transmits information regarding the direction and amount (angle) of the detected deviation of the rider unit 7 to the control unit 62.

[0027] In this manner, in this embodiment, when the rider unit 7 adjusted by alignment becomes misaligned due to an accident, the control device 6 calibrates the rider unit 7 to suppress deviations in the measurement range of the rider unit 7 and maintain an optimal state in which automatic driving can continue.

[0028] [Example of rider unit configuration] Next, a configuration example of the lidar unit 7 will be described. FIG. 3 shows a block diagram of the lidar unit 7. The lidar unit 7 is a TOF (Time Of Flight) type lidar, and performs distance measurement and detection of objects present around the vehicle. The lidar unit 7 is used, for example, as a part of an advanced driving assistance system for the purpose of assisting in recognition of the surrounding environment of the vehicle. The lidar unit 7 mainly has a plurality of scanners (L1 to L4, ...), a plurality of optical transmitting and receiving units (TR1 to TR4, ...), and a signal processing unit SP. In the following description, when the scanners (L1 to L4, ...) are not distinguished from each other, they are simply referred to as "scanner L", and when the optical transmitting and receiving units (TR1 to TR4, ...) are not distinguished from each other, they are simply referred to as "optical transmitting and receiving unit TR".

[0029] The scanner L emits a laser pulse (hereinafter also referred to as a "transmitted light pulse") within a range of a predetermined horizontal angle and vertical angle. The scanner L emits a transmitted light pulse for each segment obtained by dividing the above-mentioned horizontal angle by an equal angle. The optical transmitting / receiving unit TR then receives the reflected light (hereinafter also referred to as a "received light pulse") of the transmitted light pulse within a predetermined period after the emission of the transmitted light pulse, and outputs a signal (hereinafter also referred to as a "segment signal Sseg") related to the received light intensity for each segment to the signal processing unit SP. The signal processing unit SP outputs point cloud information indicating a set of the distance to each point of the object irradiated with the transmitted light pulse and the angle of the object based on the segment signal Sseg for each segment received from the optical transmitting / receiving unit TR.

[0030] In this embodiment, the scanner L and the optical transmitter-receiver TR are accommodated in a scanner accommodation section 50 (50a to 50c, ...). The scanner accommodation section 50 accommodates one or more pairs of the scanner L and the optical transmitter-receiver TR. In the example of FIG. 3, the scanner accommodation section 50a accommodates the scanner L1 and the optical transmitter-receiver TR1, the scanner accommodation section 50b accommodates the scanners L2 and L3 and the optical transmitter-receiver TR2 and TR3, and the scanner accommodation section 50c accommodates the scanner L4 and the optical transmitter-receiver TR4. Each of the scanner accommodation sections 50 is provided with a posture sensor 8 (8a to 8c, ...) for detecting the posture of the scanner accommodation section 50 and an adjustment mechanism 10 (10a to 10c, ...) for adjusting the orientation of the scanner accommodation section 50.

[0031] The attitude sensors 8 (8a to 8c, ...) are sensors used to detect misalignment of each scanner housing section 50, and transmit a detection signal to the control device 6. The adjustment mechanism 10 includes, for example, an actuator, and adjusts the orientation of the scanner housing section 50 based on a control signal received from the control device 6. Note that the attitude sensor 8 and adjustment mechanism 10 shown in FIG. 3 may be provided for each scanner L instead of for each scanner housing section 50.

[0032] FIG. 4 shows an example of the arrangement of the individual scanners L of the lidar unit 7. In FIG. 4, hatched sectoral areas indicate scanning ranges by transmission light pulses emitted by the corresponding scanners L. In the example of FIG. 4, the vehicle is provided with 12 scanners L (L1 to L12), the scanners L1 to L4 are provided so that the scanning range is the front of the vehicle, and the scanners L5 and L6 are provided so that the scanning range is the left side of the vehicle. The scanners L7 and L8 are provided so that the scanning range is the right side of the vehicle, and the scanners L9 to L12 are provided so that the scanning range is the rear of the vehicle.

[0033] Fig. 5 shows a schematic configuration example of the optical transmitter / receiver TR. As shown in Fig. 5, the optical transmitter / receiver TR mainly includes a synchronization control unit 11, an LD driver 12, a laser diode 13, a motor control unit 15, a light receiving element 16, a current-voltage conversion circuit (transimpedance amplifier) ​​17, an A / D converter 18, a segmentator 19, and a crystal oscillator 20.

[0034] The crystal oscillator 20 outputs a pulsed clock signal "S1" to the synchronization control unit 11 and the A / D converter 18. The synchronization control unit 11 outputs a pulsed trigger signal "S2" to the LD driver 12. The synchronization control unit 11 also outputs a segment extraction signal "S3" to the segmentator 19, which determines the timing at which the segmentator 19 extracts the output of the A / D converter 18, as described below.

[0035] The LD driver 12 supplies a pulse current to the laser diode 13 in synchronization with a trigger signal S2 input from the synchronization control unit 11. The laser diode 13 is, for example, an infrared pulse laser, and emits an optical pulse based on the pulse current supplied from the LD driver 12.

[0036] The scanner L is configured as a scanner including, for example, a transmitting and receiving optical system, and scans the transmitting light pulse emitted by the laser diode 13, and guides the receiving light pulse, which is the return light reflected by the object on which the emitted transmitting light pulse is irradiated, to the light receiving element 16. In this embodiment, the scanner L includes a motor for rotating. In this way, the scanner L functions as an irradiation unit that irradiates electromagnetic waves.

[0037] The light receiving element 16 is, for example, an avalanche photodiode, and generates a weak current corresponding to the light quantity of the reflected light from the object guided by the scanner L, i.e., the received light pulse. The light receiving element 16 supplies the generated weak current to a current-voltage conversion circuit 17. The current-voltage conversion circuit 17 amplifies the weak current supplied from the light receiving element 16 and converts it into a voltage signal, and inputs the converted voltage signal to an A / D converter 18.

[0038] The A / D converter 18 converts the voltage signal supplied from the current-voltage conversion circuit 17 into a digital signal based on the clock signal S1 supplied from the crystal oscillator 20, and supplies the converted digital signal to the segmentator 19. The segmentator 19 generates a digital signal that is the output of the A / D converter 18 during the period in which the segment extraction signal S3 is asserted, as a segment signal Sseg. The segmentator 19 supplies the generated segment signal Sseg to the signal processing unit SP.

[0039] The signal processor SP generates point cloud information indicating the distance and angle of the object for each optical transmitter / receiver TR based on the segment signal Sseg transmitted from each optical transmitter / receiver TR. Specifically, the signal processor SP detects peaks from the waveform of the segment signal Sseg and estimates the amplitude and delay time corresponding to the detected peak. Then, the signal processor SP generates a set of distance information corresponding to the delay time of a peak whose estimated amplitude is equal to or greater than a predetermined threshold among the peaks of the waveform indicated by the segment signal Sseg and angle information corresponding to the target segment as information of each point constituting the point cloud information.

[0040] [Alignment process] Next, an alignment process, which is an adjustment process performed on the rider unit 7 when the rider unit 7 is manufactured or when the rider unit 7 is attached to a vehicle, will be described.

[0041] The point cloud information generated based on the received optical pulses received by each optical transmitter / receiver TR via each scanner L is information expressed in a local coordinate system based on the orientation and position of each scanner L, and depends on the relative position (specifically, the placement position and placement angle) of each scanner L with respect to the vehicle. Therefore, after aligning each scanner L with the vehicle, the measurement system 100 generates alignment information IA indicating the placement position and placement angle of each scanner L with respect to the vehicle after alignment, and stores it in the memory unit 2.

[0042] As a result, for example, when the vehicle is traveling, the control device 6 or the signal processing unit SP can convert the coordinate system of the point cloud information obtained for each scanner L into a common coordinate system based on the vehicle by referring to the alignment information IA stored in the memory unit 2, and can also convert it into an absolute coordinate system based on latitude, longitude, and altitude by using the position information and orientation information of the vehicle. Also, in this embodiment, as described later, the control device 6 refers to the alignment information IA and preferably detects alignment deviation of the scanner housing unit 50.

[0043] [Accident detection processing] When it is estimated that a vehicle accident will occur, the control device 6 determines that calibration of the rider unit 7 is necessary. Here, the vehicle accident detection process will be described.

[0044] (1) Detection based on reference angle If the alignment angle (also called the "reference angle") of the scanner housing unit 50 or scanner L relative to the vehicle indicated by the alignment information IA stored in the memory unit 2 differs from the current alignment angle (also called the "latest measured angle") of the scanner housing unit 50 or scanner L relative to the vehicle measured by the attitude sensor 8, the control unit 6 determines that an alignment shift has occurred in the scanner housing unit 50 or scanner L due to the occurrence of an accident.

[0045] Here, when calculating the latest measured angle, the control device 6 performs a process of canceling the inclination of the vehicle itself. Specifically, the control device 6 calculates the latest measured angle excluding the influence of the inclination of the vehicle itself by correcting the angle measured by the attitude sensor 8 installed in the scanner housing 50 by the vehicle inclination angle measured by the attitude sensor 8 or the like installed in the vehicle. The control device 6 may recognize the inclination angle of the vehicle itself based on the inclination information of the road data registered in the map DB 20 and calculate the latest measured angle. In this case, the control device 6 acquires the inclination information of the road to which the current position belongs from the map DB 20 based on the vehicle position information recognized by the output of the sensor unit 3.

[0046] Fig. 6(A) shows a three-axis (X, Y, Z) coordinate system detected by the attitude sensor 8 installed in the scanner housing section 50 when no misalignment occurs. In the example of Fig. 6(A), the vehicle is placed on a horizontal surface, and the attitude sensor 8 detects acceleration in each of the X, Y, and Z axes. Here, the "X axis" indicates the front direction of the scanner housing section 50 (i.e., the laser emission direction), the "Y axis" indicates the left and right side directions of the scanner housing section 50, and the "Z axis" indicates the height direction of the scanner housing section 50. When no misalignment occurs, the X and Y axes are approximately perpendicular to the gravitational acceleration, and the Z axis is approximately parallel to the gravitational acceleration.

[0047] Fig. 6(B) is a diagram showing a change in the orientation of the scanner housing part 50 when misalignment of the scanner housing part 50 in Fig. 6(A) occurs due to an impact caused by an accident, etc. In Fig. 6(B), the solid XYZ coordinate axes indicate the three coordinate axes detected by the attitude sensor 8 installed on the tilted scanner housing part 50, and the dashed XYZ coordinate axes indicate the three coordinate axes before the scanner housing part 50 shown in Fig. 6(A) is tilted.

[0048] In the example of FIG. 6(B), the scanner housing 50 is tilted from the state of FIG. 6(A), and the measurement values ​​of the three axes of the attitude sensor 8 are different by an amount corresponding to the X-axis deviation (see arrow A1), Y-axis deviation (see arrow A2), and Z-axis deviation (see arrow A3) before and after the scanner housing 50 is tilted. In this way, when the scanner housing 50 is misaligned due to an accident or the like, a deviation (rotation deviation) of the arrangement angle of the scanner housing 50 occurs, and the measurement value of the attitude sensor 8 changes before and after the alignment deviation according to the deviation. Taking the above into consideration, in this embodiment, the control device 6 compares the reference angle indicated by the alignment information IA with the latest measured angle based on the output of the attitude sensor 8 for each scanner housing 50 for each rotation direction (for example, the roll direction around the X axis, the pitch direction around the Y axis, and the yaw direction around the Z axis), and if there is a rotation direction in which a deviation of a predetermined angle or more has occurred, it is estimated that the scanner housing 50 has been misaligned due to an accident or the like.

[0049] It should be noted that, in the case of an alignment shift that does not involve a change in the arrangement angle (i.e., a position shift along the X-axis, Y-axis, and / or Z-axis), the measurement value of the attitude sensor 8 does not change before and after the shift. On the other hand, in the case of a shift in the scanner housing unit 50 due to an impact from an accident or the like, it is expected that a shift in the arrangement angle will inevitably occur in addition to a position shift. Therefore, in this embodiment, the control device 6 detects a shift in the arrangement angle of the scanner housing unit 50 to determine whether calibration is required.

[0050] (2) Detection by other methods The control device 6 may also, instead of the alignment misalignment detection process based on the above-mentioned alignment information IA, if an abnormality is detected using any of the methods (second and third methods) described below, assume that an alignment misalignment of the scanner accommodating section 50 has occurred due to an accident and determine that the calibration process described below should be performed.

[0051] In the second method, when the control device 6 determines based on the output of the shock sensor 9 provided in the vehicle that an impact of a predetermined level or more has occurred to the vehicle, it determines that there is a high possibility that the scanner housing part 50 has become misaligned due to the accident, and that a calibration process, which will be described later, is necessary. For example, as described in FIG. 2, when the control device 6 receives shock information from the shock sensor 9 indicating that an impact of a predetermined level or more has occurred, it determines that an impact has occurred to the vehicle. This also enables the control device 6 to preferably detect the occurrence of misalignment of the scanner housing part 50.

[0052] In the third method, when an error occurs in the object recognition process, the control device 6 determines that there is a high possibility that the scanner housing 50 has become misaligned due to an accident, and that a calibration process (described later) is necessary. For example, when the road surface recognized by combining the point cloud information generated based on the received optical pulses of each scanner L is distorted, that is, when the inclination angle differs for each road surface indicated by the point cloud information for each scanner L, the control device 6 determines that at least one of the scanners L has become misaligned.

[0053] In another example based on the third method, the control device 6 compares point cloud information for overlapping scanning ranges of scanners L, and if there is a shift in the position of the point cloud indicated by the compared point cloud information, determines that an alignment shift has occurred in at least one of the scanners L with overlapping scanning ranges.

[0054] In yet another example based on the third method, when the accuracy of the vehicle's position information is sufficiently high, the control device 6 converts the position information of a specific feature registered in the map DB 20 and the point cloud information including the feature in the scanning range into the same coordinate system based on the vehicle's position information, and compares them. Then, when there is a deviation between the position indicated by the position information of the feature and the position of the feature indicated by the point cloud information, the control device 6 determines that an alignment deviation has occurred in the scanner L that scanned the feature.

[0055] In yet another example based on the third method, the control device 6 compares the position of an object recognized based on an image output by a camera included in the sensor unit 3 with the position of the same object recognized based on point cloud information output by the lidar unit 7. If the position of the compared object is shifted by a predetermined amount or more, the control device 6 determines that there is a high possibility that the alignment of the scanner housing unit 50 has been shifted due to an accident, and that calibration processing is necessary. In this case, the control device 6 may measure the three-dimensional position of the object based on the output of multiple cameras, and compare the measured three-dimensional position with the three-dimensional position based on the point cloud information output by the lidar unit 7.

[0056] [Calibration process] Next, a calibration process that is executed when the alignment of the rider unit 7 is shifted due to a vehicle accident will be described.

[0057] As a calibration process, the control device 6 performs at least one of electronic adjustment to change the actual scanning range, which will be described later, or physical adjustment by controlling the adjustment mechanism 10 provided in each scanner housing part 50. Note that, as a pre-processing of the calibration process, the control device 6 specifies the alignment deviation (i.e., the direction and amount of deviation) of each scanner housing part 50 by calculating, for each direction, the difference between the reference angle of each scanner housing part 50 indicated by the alignment information IA stored in the memory unit 2 and the latest measured angle indicated by the measurement information output by the attitude sensor 8 of each scanner housing part 50.

[0058] Here, a specific example of the above-mentioned electronic adjustment will be described. Fig. 7(A) shows the correspondence between the scannable range "SR" of a certain scanner L and the actual scanning range "FOV" on a virtual irradiation plane perpendicular to the emission direction of the scanner L before the calibration process is performed. Here, the scannable range SR refers to the range where scanning by the transmitted light pulse is possible, and the actual scanning range FOV refers to the range where scanning by the transmitted light pulse is actually performed. Note that the arrow in the actual scanning range FOV shows an example of the scanning direction.

[0059] As shown in FIG. 7A, the actual scanning range FOV is moved within the scannable range SR based on the direction and amount of deviation from the scannable range.

[0060] FIG. 7B shows the correspondence between the scannable range SR and the actual scanning range FOV after the control device 6 adjusts the alignment as the electronic adjustment, which is set to a range smaller than the scanning range SR of the scanner L based on electronic adjustment. In the example of FIG. 7B, the control device 6 transmits a control signal instructing a change in the parameters of the internal signal of the optical transmitting / receiving unit TR based on the direction and amount of misalignment caused in the scanner housing unit 50 that houses the scanner L. As a result, the control device 6 moves the actual scanning range FOV by a predetermined distance in the upper left direction (see arrow 55). In this case, the control device 6 stores in advance in the storage unit 2, for example, a map or the like for each scanner L indicating the movement direction and movement width of the actual scanning range FOV according to the misalignment of each scanner housing unit 50, and generates the above-mentioned control signal based on the movement direction and movement width of the actual scanning range FOV determined by referring to the map. As a result, the control device 6 suitably adjusts the current scanning range of the lidar unit 7 so that it becomes equivalent to the scanning range of the lidar unit 7 immediately after the alignment process (that is, before the occurrence of alignment deviation).

[0061] In this way, by moving the actual scanning range FOV based on electronic adjustment, the control device 6 can effectively suppress a decrease in the accuracy of the object recognition processing using the lidar unit 7 and continue automatic driving control even if an alignment shift occurs due to an accident or the like.

[0062] Furthermore, instead of the electronic adjustment described above, the control device 6 may perform physical adjustment to move the actual scanning range FOV by controlling the adjustment mechanism 10 of the scanner housing part 50. In this case, for example, each scanner housing part 50 is rotatable in the roll direction, pitch direction, and yaw direction, and the adjustment mechanism 10 adjusts the angle of the corresponding scanner housing part 50 in any rotation direction based on a control signal supplied from the control device 6.

[0063] Therefore, when performing the above-mentioned physical adjustment, the control device 6 generates a control signal to be sent to the adjustment mechanism 10 based on the identified direction and amount of deviation of the scanner housing part 50 so that the actual scanning range FOV is substantially the same before and after the alignment deviation. In this case, for example, a map showing the relationship between the direction and amount of deviation of the scanner housing part 50 and the direction (for example, directions around the X, Y, and Z axes in FIG. 6) and amount of movement of the scanner housing part 50 is stored in advance in the storage unit 2. Then, the control device 6 generates a control signal for the adjustment mechanism 10 of the scanner housing part 50 where alignment deviation has occurred, by referring to the above-mentioned map.

[0064] Preferably, when the control device 6 determines that the actual scanning range FOV cannot be adjusted to be equivalent to the scanning range of the lidar unit 7 immediately after the alignment process (i.e., before the alignment shift occurs) using only the above-mentioned electronic adjustment, the control device 6 adjusts the position of the scanner accommodating section 50 based on the control of the adjustment mechanism 10.

[0065] In addition, the control device 6 does not perform the calibration process when the deviation of the scanner housing 50 is large and the actual scanning range FOV cannot be adjusted to the same range before and after the alignment deviation by either the electronic adjustment or the physical adjustment described above. In this case, instead of the calibration process, the control device 6 causes the notification unit 4 to output an output that prompts or notifies the user to switch to manual driving, or causes the notification unit 4 to output a warning that an error has occurred. In this case, for example, when the control device 6 detects a user input that instructs switching to manual driving, or after a predetermined time has elapsed since the above-mentioned warning, the control device 6 switches the driving mode of the vehicle from automatic driving to manual driving. As a result, the control device 6 can prevent the automatic driving from continuing in a situation where the accuracy of the recognition process using the rider unit 7 is low, and can appropriately ensure safety.

[0066] [Processing flow] Next, the alignment process executed before the shipment of the vehicle and the process executed during autonomous driving after the shipment of the vehicle will be described with reference to the flowcharts of FIGS.

[0067] (1) Alignment Processing Fig. 8 is a flowchart showing the procedure of the alignment process of the lidar unit 7. The measurement system 100 executes the process of the flowchart shown in Fig. 8 when installing each scanner L on the vehicle. Note that the alignment process is performed in a state where the vehicle is placed on a horizontal surface.

[0068] First, the scanner housings 50 housing the scanners L are fitted into the vehicle at their respective predetermined positions, and the alignment of each scanner housing 50 is adjusted (step S101). In this case, for example, the control device 6 adjusts the angle of at least one of the pitch, yaw, and roll directions of the scanner housing 50 by transmitting a control signal to the adjustment mechanism 10 of the scanner housing 50 that requires position adjustment, based on an input to the input unit 1, etc. In another example, the alignment of each scanner housing 50 may be adjusted based on manual work.

[0069] Then, measurement information output by the attitude sensor 8 provided in each scanner housing part 50 after the alignment adjustment in step S101 is stored in the memory unit 2 as alignment information IA indicating the reference angle of each scanner housing part 50 (step S103).

[0070] (2) Processing during autonomous driving Fig. 9 is an example of a flowchart showing a process flow executed by the control device 6 during automatic driving. The control device 6 repeatedly executes the process of the flowchart in Fig. 9 according to a predetermined cycle, for example.

[0071] First, the control device 6 determines whether or not an object recognition error has occurred (step S201). In this case, the control device 6 determines whether or not there is a deviation or the like in the point cloud information obtained for each scanner L based on the third method described in the [Accident Detection Processing] section. Then, if an object recognition error has occurred (step S201; Yes), the control device 6 proceeds to step S204.

[0072] On the other hand, if an object recognition error has not occurred (step S201; No), the control device 6 determines whether or not an impact that is estimated to be an accident has been detected (step S202). In this case, the control device 6 determines whether or not the value indicated by the impact information output by the impact sensor 9 temporarily exceeds a predetermined threshold value based on the second method described in the [Accident detection processing] section. Then, if the control device 6 detects an impact that is estimated to be an accident (step S202; Yes), the control device 6 proceeds to step S204.

[0073] On the other hand, if the above-mentioned impact is not detected (step S202; No), the control device 6 judges whether or not an angle deviation is detected in which the deviation amount between the reference angle indicated by the alignment information IA and the latest measured angle detected by the attitude sensor 8 is equal to or greater than a predetermined amount (step S203). If the control device 6 detects the above-mentioned angle deviation (step S203; Yes), the control device 6 proceeds to the process of step S204. On the other hand, if the above-mentioned angle deviation is not detected (step S203; No), it is determined that no alignment deviation has occurred in the rider unit 7, and the process of the flowchart is terminated.

[0074] Next, when the control device 6 detects an abnormality in any of steps S201 to S203, it recognizes the direction and amount of the angle shift between the reference angle and the latest measured angle (step S204). Then, the control device 6 determines whether or not the actual scanning range FOV can be adjusted to compensate for the above-mentioned angle shift by a calibration process (step S205). For example, the control device 6 stores information on the range of the angle shift that can be adjusted by electronic adjustment or physical adjustment described in the [Calibration Process] section in advance in the storage unit 2, and determines whether or not the actual scanning range FOV can be adjusted by the calibration process by referring to the information on the above-mentioned range.

[0075] Then, when the control device 6 determines that the actual scanning range FOV can be adjusted so as to compensate for the above-mentioned angular deviation by the calibration process (step S205; Yes), the control device 6 outputs a guide regarding the execution of the calibration process by the notification unit 4 (step S206). For example, the control device 6 outputs a message that the automatic operation can be continued by executing the calibration process, and also outputs an input prompting the user to select whether or not to execute the calibration process by the notification unit 4. Then, when the control device 6 detects an input indicating that the calibration process should be executed (step S207; Yes), the control device 6 executes the calibration process (step S208). In this case, for example, when the control device 6 can compensate for the above-mentioned angular deviation by adjusting the actual scanning range FOV by electronic adjustment, the control device 6 adjusts the actual scanning range FOV by electronic adjustment, and when the control device 6 cannot compensate for the above-mentioned angular deviation only by electronic adjustment, the control device 6 adjusts the actual scanning range FOV by performing a physical adjustment of the scanner housing unit 50 based on the control of the adjustment mechanism 10. Then, when the calibration process is successful (step S209; Yes), the control device 6 ends the processing of the flowchart. In this case, the control device 6 continues to execute the control of the automatic driving.

[0076] On the other hand, if the calibration process cannot adjust the actual scanning range FOV to compensate for the above-mentioned angle deviation (step S205; No), or if there is no input to execute the calibration process (step S207; No), or if the calibration process is not successful (step S209; No), the control device 6 displays a predetermined warning and / or switches to manual driving (step S210). In this case, the control device 6 outputs a guidance to the effect that automatic driving cannot be continued or a guidance to prompt manual switching to manual driving, or performs automatic switching to manual driving after the above guidance. As a result, when the reliability of the rider unit 7 is degraded, the control device 6 can prompt a smooth switch to manual driving and appropriately suppress a decrease in safety.

[0077] As described above, the control device 6 according to the first embodiment functionally includes a detection unit 61 and a control unit 62. The detection unit 61 detects a change in the arrangement angle of the scanner L that irradiates electromagnetic waves toward the outside of the vehicle based on the alignment information IA generated during alignment and the measurement information related to the lidar unit 7 received from the attitude sensor 8. The control unit 62 controls the irradiation direction of the scanner L based on the detection result of the detection unit 61. As a result, even if an alignment deviation occurs in the lidar unit 7 due to an accident or the like, the control device 6 can calibrate the lidar unit 7 to preferably maintain a state in which automatic driving can be continued.

[0078] <Second Example> The second embodiment differs from the first embodiment in that the control device 6 performs calibration based on the detection result of the deviation of the actual scanning range FOV based on the inclination of the detected feature, instead of or in addition to performing calibration based on the deviation between the reference angle indicated by the alignment information IA and the latest measured angle measured by the attitude sensor 8. Hereinafter, the same components as those in the first embodiment are appropriately denoted by the same reference numerals, and their explanation will be omitted.

[0079] Fig. 10 shows an example of the data structure of the map DB 20 according to the second embodiment. As shown in Fig. 10, the map DB 20 includes, in addition to road data and the like, feature information on features around the roads. The feature information is information registered for each feature, and includes type information, position information, and slope information. By using a map DB including such feature information, it becomes possible to perform vehicle navigation during manual driving and advanced vehicle driving control during automatic driving.

[0080] Here, the type information is information that indicates the type of feature, and is referred to, for example, in a process of determining whether or not a feature around the vehicle is suitable as a reference feature (also called a "reference feature") when performing calibration. The position information is information that indicates the absolute position of the feature, and is referred to in a process of identifying the position of the feature relative to the vehicle. The inclination information is information that indicates the inclination angle of the feature from the road surface.

[0081] Here, the inclination information may indicate the angle that the contour of the side (including the front and back) of the feature forms with the road surface, or may indicate the angle that the line indicating the axis of symmetry of the contour of the feature (the line passing through the center of gravity of the feature) forms with the road surface. This will be described with reference to FIG.

[0082] Figure 11(A) is a diagram showing the relationship between feature A, which has a rectangular outline when observed from the road, and the inclination angle θ indicated by the inclination information of the feature information. In Figure 11(A), feature A has a prism or cylinder shape, and the inclination angle θ indicated by the inclination information of the feature information corresponding to feature A is the angle (here, about 90 degrees) that the outline of the side (including the front and back) of feature A forms with the road surface.

[0083] FIG. 11(B) is a diagram showing the relationship between feature B, which has an outline that tapers toward the top when observed from above the road, and the inclination angle θ indicated by the inclination information of the feature information. In FIG. 11(B), the outline of feature B has a shape with a broken line 70 as an axis of symmetry. The broken line 70 is a line that passes through the center of gravity of the outline of feature B, and is a line that is approximately parallel to the outline that shows the side of feature B. The inclination angle θ indicated by the inclination information of the feature information corresponding to feature B indicates the angle that the broken line 70 forms with respect to the road surface (here, about 90 degrees). Thus, in the example of FIG. 11(B), information on the inclination of feature B as a whole is recorded in the feature information as inclination information. In this case, a predetermined flag or the like indicating that the inclination information indicates the inclination angle of the line that passes through the center of gravity of the feature may be added to the feature information.

[0084] FIG. 11C is a diagram showing the relationship between feature C when observed from the road and the inclination angles θ1, θ2 indicated by the inclination information of the feature information. In FIG. 11C, the feature C has contour lines 71, 72 forming the side surfaces that form different angles θ1, θ2 with respect to the road surface. In this case, the feature information corresponding to feature C includes inclination information indicating the inclination angles θ1, θ2 corresponding to the contour lines 71, 72. In this case, the feature information may have a data structure capable of identifying the correspondence between the contour lines detected by the lidar unit 7 and the inclination angles recorded in the feature information. For example, the inclination angles may be recorded in the feature information in order from the inclination angle of the contour line closest to the road.

[0085] The control device 6 refers to the feature information including the tilt information based on any of the formats described in Fig. 11(A) to (C), and identifies a line (also called "feature reference line Lo") that approximately connects points that form the tilt angle indicated by the tilt information of the feature information from the point cloud that forms the outline of the reference feature detected by the LIDAR unit 7. Next, the control device 6 calculates the angle between the identified feature reference line Lo and a line (also called "actual scanning boundary line Lf") that forms any side of the boundary of the rectangular actual scanning range FOV. Then, the control device 6 detects an alignment deviation of the LIDAR unit 7 based on the angle (also called "reference angle θtag") between the feature reference line Lo and the actual scanning boundary line Lf and the tilt angle indicated by the corresponding tilt information. In this way, the information of the reference angle θtag functions as the first information, and the tilt information functions as the second information.

[0086] Next, a specific example of a method for detecting an alignment error based on the reference angle θtag and the tilt angle indicated by the tilt information will be described with reference to FIG.

[0087] Fig. 12(A) is a diagram showing the correspondence between the actual scanning range FOV of a scanner L before misalignment occurs and the scenery overlapping the actual scanning range FOV. In the example of Fig. 12(A), when no misalignment occurs, the lateral direction of the actual scanning range FOV of the target scanner L is parallel to the road surface (i.e., the horizon).

[0088] In this case, first, the control device 6 extracts a point cloud forming the contour of the feature 60 from the point cloud information generated by the lidar unit 7. Then, the control device 6 projects the point cloud forming the contour of the feature 60 onto the same plane as the actual scanning range FOV, and identifies the point cloud forming the contour of the side of the feature 60 from the point cloud on the plane, thereby recognizing the feature reference line Lo. In this case, for example, the control device 6 may assume that the actual scanning range FOV is a range on a virtual irradiation surface that is a predetermined distance away from the vehicle and on which the transmitted light pulse is irradiated. In addition, the control device 6 determines the actual scanning boundary line Lf (here, the bottom side of the actual scanning range FOV) from the boundary line of the actual scanning range FOV, and calculates the reference angle θtag based on the determined actual scanning boundary line Lf and the feature reference line Lo.

[0089] Then, the control device 6 calculates the angle difference between the calculated reference angle θtag (here, 90 degrees) and the tilt angle (here, 90 degrees) indicated by the tilt information of the feature information corresponding to the feature 60. In this case, since the above-mentioned angle difference is 0 degrees, the control device 6 determines that no angle deviation occurs in the scanner L corresponding to the actual scanning range FOV of the target in the example of Fig. 12(A). Therefore, in this case, the control device 6 determines that calibration is not required for the scanner L of the target.

[0090] FIG. 12B is a diagram showing the correspondence between the actual scanning range FOV of the scanner L in which misalignment occurs and the scenery overlapping the actual scanning range FOV.

[0091] In this case, the control device 6 executes the same process as in Fig. 12(A) to identify the actual scanning boundary line Lf and the feature reference line Lo, and calculates the reference angle θtag between the actual scanning boundary line Lf and the feature reference line Lo. Then, the control device 6 calculates the angle difference (here, 10 degrees) between the calculated reference angle θtag (here, 100 degrees) and the tilt angle (here, 90 degrees) of the feature 60 indicated by the referenced tilt information. In this case, the control device 6 determines that an alignment deviation has occurred in the target scanner L since the above-mentioned angle difference is equal to or greater than a predetermined difference. Therefore, in this case, the control device 6 determines that it is necessary to perform calibration on the target scanner L.

[0092] Fig. 12C is a diagram showing the correspondence relationship between the actual scanning range FOV after the calibration and the scenery overlapping the actual scanning range FOV. In Fig. 12C, a dashed line 61 indicates the position of the actual scanning boundary line Lf before the calibration.

[0093] Based on the above-mentioned angle difference, the control device 6 determines that the scanner L corresponding to the target actual scanning range FOV is shifted by 10 degrees in the roll direction (clockwise). Therefore, in this case, the control device 6 rotates the actual scanning range FOV by 10 degrees in the roll direction (counterclockwise) based on the calibration method described in the [Calibration Processing] of the first embodiment. As a result, as shown in FIG. 12(C), the longitudinal direction of the actual scanning range FOV and the road surface become approximately parallel, and they have the same corresponding relationship as before the alignment shift of the target scanner L occurs (i.e., before the impact of the vehicle occurs).

[0094] In this way, the control device 6 can determine whether calibration is necessary and perform calibration in an appropriate manner by calculating the reference angle θtag between the actual scanning boundary line Lf and the feature reference line Lo. In addition, in a state where there is no alignment deviation, the longitudinal direction of the actual scanning range FOV is not limited to being parallel to the road surface, and may have a predetermined inclination angle with respect to the road surface. In this case, for example, the control device 6 stores information on the inclination angle in advance, and performs calibration by further taking the above-mentioned inclination angle into consideration with respect to the angle difference between the reference angle θtag and the inclination angle indicated by the inclination information.

[0095] Also, when the inclination information of the reference feature indicates the inclination angle of a line passing through the center of gravity of the feature as in the example of FIG. 11(B), the control device 6, for example, obtains the center of gravity of the point cloud when the point cloud of the reference feature is projected onto the same plane as the actual scanning range FOV, and calculates the feature reference line Lo (for example, a symmetric line) passing through the obtained center of gravity based on a predetermined calculation method. Also, when the inclination information of the reference feature records the inclination angles of each of a plurality of contours (see FIG. 11(C)), the control device 6 calculates the feature reference line Lo based on at least one inclination angle indicated by the inclination information and point cloud information representing the contour portion corresponding to the inclination angle. In this case, the control device 6 may calculate the feature reference line Lo based on point cloud information of a contour corresponding to an inclination angle close to 90 degrees, for example, or may calculate the feature reference line Lo based on point cloud information of a contour that has been detected as point cloud information.

[0096] Fig. 13 is a flowchart showing the procedure of calibration based on a feature according to the second embodiment. The control device 6 repeatedly executes the process of the flowchart shown in Fig. 13, for example, in a predetermined cycle. The control device 6 may execute the process of the flowchart shown in Fig. 13 in parallel with the process of the flowchart of Fig. 9 described in the first embodiment, or may execute it instead of the process of the flowchart of Fig. 9.

[0097] First, the control device 6 acquires feature information of features existing around the vehicle (step S301). In this case, for example, the control device 6 refers to the vehicle position information generated based on the position information of the feature information and the output of the sensor unit 3, and extracts from the map DB 20 the feature information of features existing within the actual scanning range FOV of any of the scanners L within the maximum distance measurement distance from the vehicle position.

[0098] Next, the control device 6 judges whether or not there is a feature suitable for use as a reference feature (step S302). In this case, for example, the control device 6 refers to type information included in the feature information acquired in step S301 and judges whether or not there is a fixed object such as a building whose outline does not change due to weather conditions such as wind or other external environments. In another example, the control device 6 judges a feature (e.g., a building, a traffic light pole, a utility pole, etc.) that is estimated to be configured perpendicular to the road surface as a feature suitable for use as a reference feature. In this example, the control device 6 may estimate whether or not the feature is configured perpendicular to the road surface based on type information of the feature information, or may make the above-mentioned judgment based on whether or not the inclination angle of the outline of the feature indicated by the inclination information included in the feature information is about 90 degrees.

[0099] Then, if there is a feature suitable for use as a reference feature (step S302; Yes), the control device 6 regards the feature as a reference feature and executes the processes of steps S303 to S306 below. On the other hand, if there is no feature suitable for use as a reference feature (step S302; No), the control device 6 ends the process of the flowchart.

[0100] Next, the control device 6 calculates, for each of the reference features, a feature reference line Lo that forms a line passing through the contour or center of gravity of the feature based on the point cloud information of each feature (step S303). In this case, the control device 6 extracts a point cloud for each reference feature from the point cloud information obtained from the lidar unit 7 based on, for example, the position information of the feature, and calculates a feature reference line Lo that forms the contour, etc. of the reference feature for each reference feature.

[0101] Then, the control device 6 specifies an actual scanning boundary line Lf based on the actual scanning range FOV of the scanner L that includes each reference feature in its detection range for the calculated feature reference line Lo for each reference feature, and calculates the reference angle θtag (step S304).The control device 6 then determines whether or not there is a reference feature for which the angle difference between the tilt angle indicated by the tilt information and the reference angle θtag is equal to or greater than a predetermined angle difference (step S305).The above-mentioned predetermined angle is determined in advance, for example, taking into consideration the necessity for calibration based on the angle difference and the calculation error of the reference angle θtag.

[0102] Then, when there is a reference feature where the angle difference between the tilt angle indicated by the tilt information and the reference angle θtag is equal to or greater than a predetermined angle difference (step S305; Yes), the control device 6 determines that there is a deviation in the actual scanning range FOV used to calculate the reference angle θtag, and performs calibration based on the above-mentioned angle difference for the scanner L corresponding to the actual scanning range FOV (step S306). In this case, the control device 6 rotates the actual scanning range FOV of the target by electronic adjustment or physical adjustment under the control of the adjustment mechanism 10 so as to eliminate the above-mentioned angle difference. Note that, similar to the processing of the flowchart of FIG. 9 described in the first embodiment, the control device 6 may provide guidance regarding the execution of calibration before the execution of calibration, and perform calibration when it detects an input indicating that calibration should be performed.

[0103] On the other hand, if there is no reference feature for which the angle difference between the tilt angle indicated by the tilt information and the reference angle θtag is equal to or greater than the predetermined angle (step S305; No), the control device 6 determines that no alignment shift of the scanner L is detected and that there is no need to perform calibration. Therefore, in this case, the control device 6 ends the processing of the flowchart without performing calibration.

[0104] As described above, the control device 6 according to the second embodiment acquires an output signal from the lidar unit 7 capable of detecting features present around the vehicle, and calculates the reference angle θtag indicating the inclination angle of the feature recognized based on the output signal of the lidar unit 7 with respect to the detection range of the lidar unit 7. The control device 6 then controls the detection range of the lidar unit 7 based on the reference angle θtag and the inclination information stored in the storage unit 2 and indicating the angle of the feature with respect to the road surface. This also allows the control device 6 to calibrate the lidar unit 7 when an alignment deviation occurs in the lidar unit 7, as in the first embodiment, to favorably maintain a state in which automatic driving can be continued.

[0105] <Third Example> The third embodiment differs from the second embodiment in that the feature information included in the map DB 20 includes information indicating whether calibration can be performed using the target feature as a reference feature (also referred to as "calibration feasibility information Ic"). In the third embodiment, the control device 6 determines the suitability of each feature as a reference feature by referring to the calibration feasibility information Ic.

[0106] Fig. 14 shows the data structure of the map DB 20 stored in the storage unit 2 in the third embodiment. As shown in Fig. 14, the map DB 20 has feature information, which further includes calibration feasibility information Ic in addition to the above-mentioned type information, position information, and tilt information. Here, the calibration feasibility information Ic is flag information (adjustment feasibility information) indicating either permission information that permits the execution of calibration using a corresponding feature as a reference feature, or non-permission information that does not permit the execution of calibration.

[0107] Examples of features for which calibration feasibility information Ic indicating that calibration is permitted to be performed is added to the feature information include fixed objects such as buildings whose contours do not vary depending on the external environment, features that are configured approximately perpendicular to the road surface, and features whose contours have little variation in inclination (i.e. features with contours that are nearly straight). In this way, calibration feasibility information Ic indicating that calibration is permitted to be performed is added to feature information of features that are suitable for calculating the feature reference line Lo (i.e. features that reduce the calculation error of the feature reference line Lo). Examples of features for which calibration feasibility information Ic indicating that calibration is not permitted to be performed is added to the feature information include features whose contour shapes are likely to vary depending on the external environment, features that are configured at an angle to the road surface, and features whose contours have a lot of variation in inclination (i.e. features with contours that are curved or have other complex shapes). In addition, the calibration feasibility information Ic is not limited to being included in the feature information of all features, but may be added only to the feature information of features for which calibration using the reference feature as a reference feature is permitted to be performed, or may be added only to the feature information of features for which calibration using the reference feature as a reference feature is not permitted to be performed.

[0108] The storage unit 2 stores feature information having a data structure as shown in FIG. 14 and supplies it to the control device 6. The control device 6 then executes the flowchart shown in FIG. 13 based on the feature information supplied from the storage unit 2. In this case, in step S302, the control device 6 refers to the calibration feasibility information Ic included in the feature information for each feature to determine whether or not each feature is suitable as a reference feature. The control device 6 then regards a feature that has been determined to be suitable as a reference feature based on the calibration feasibility information Ic as the reference feature, and executes the processes of steps S303 to S306. This allows the control device 6 to accurately select a reference feature, and more accurately determine whether or not calibration can be performed and execute calibration.

[0109] As described above, the storage unit 2 of the measurement system 100 according to the third embodiment stores feature information having a data structure including calibration feasibility information Ic indicating whether the lidar unit 7 arranged on the vehicle is permitted or not permitted to calibrate the detection range of the lidar unit 7 based on the feature information. This allows the control device 6 to accurately select a feature to be used as a reference when calibrating the detection range of the lidar unit 7.

[0110] <Modification> Next, a description will be given of preferred modifications to the first to third embodiments. The following modifications may be applied to the above-mentioned embodiments in any combination.

[0111] (Variation 1) In the first to third embodiments, a part or all of the processing of the control device 6 may be executed by the signal processor SP of the rider unit 7. For example, the signal processor SP may execute the processing of the flowchart of FIG. 9 instead of the control device 6. In this case, the signal processor SP executes the processing of steps S201 to S204 of FIG. 9 by referring to the alignment information IA and the like from the memory unit 2 or the memory built in the rider unit 7 and acquiring the output value of the attitude sensor 8 provided in each scanner housing unit 50 and the like. In addition, in the calibration processing of step S208, the signal processor SP transmits a control signal to the adjustment mechanism 10 and / or the light transmitting / receiving unit TR to adjust the scanning range according to the detected angle deviation. Similarly, the signal processor SP may execute the processing of the flowchart of FIG. 13 instead of the control device 6. In this modification, the signal processor SP may exchange signals with each component of the measurement system 100, such as the input unit 1, the memory unit 2, and the attitude sensor 8, without going through the control device 6. In addition, when the signal processor SP executes all the processing of the control device 6, the measurement system 100 may not be provided with the control device 6.

[0112] (Variation 2) In the first to third embodiments, in the example of FIG. 4 etc., a plurality of scanners L are provided for the vehicle, but this is not limiting, and only one scanner L may be provided for the vehicle. In this case, the scanner L scans the transmission light pulse in a range of, for example, 360 degrees, thereby irradiating the transmission light pulse to an object present in any direction with respect to the vehicle. Even in this case, the control device 6 can suitably continue the automatic driving by detecting the alignment deviation of the scanner L provided for the vehicle based on any of the first to third embodiments and performing a calibration process as necessary. In this way, it is sufficient that at least one scanner L is provided for the vehicle.

[0113] (Variation 3) 6, in the first embodiment, the control device 6 detects the angle change of each scanner L of the lidar unit 7 about the X, Y and Z axes relative to the vehicle based on the output of the attitude sensor 8. In addition to this, the control device 6 may detect the position change of the scanner L along the X, Y and Z axes relative to the vehicle.

[0114] In this case, for example, each scanner housing section 50 or scanner L is provided with a sensor (position detection sensor) that detects a position change, and the adjustment mechanism 10 has a function of moving the scanner housing section 50 or scanner L in parallel. In the alignment process shown in FIG. 8, the control device 6 stores information on the arrangement position and arrangement angle of each scanner L after alignment adjustment as alignment information IA in the storage unit 2. Then, the control device 6 executes the process of the flowchart in FIG. 9 based on the alignment information IA stored in the storage unit 2. In this case, in step S203, the control device 6 detects the presence or absence of alignment deviation regarding the arrangement position and arrangement angle of the scanner housing section 50 or scanner L based on the output of the attitude sensor 8 and the output of the above-mentioned position detection sensor. Thereafter, the control device 6 executes the processes of steps S204 to S210 as in the embodiment, thereby executing calibration regarding the alignment deviation of the scanner L, displaying a predetermined warning, or switching to manual operation. In this way, even if the placement position of the scanner L is changed, it is possible to continue measuring the surrounding environment preferably by performing calibration, or to switch to manual operation from a safety standpoint.

[0115] (Variation 4) In the first to third embodiments, the measurement system 100 may detect misalignment and perform calibration processing, etc., on an external sensor that irradiates electromagnetic waves other than the rider unit 7, instead of detecting misalignment of the rider unit 7. In this way, when it is estimated that a misalignment has occurred in the installation position of an external sensor necessary for autonomous driving, the measurement system 100 performs calibration processing, etc. of the external sensor to compensate for the misalignment, based on the same processing procedure as the flowchart of Fig. 9 or Fig. 13. This makes it possible to suitably suppress changes in the detection range of the external sensor before and after the occurrence of an accident, and to ensure safety by encouraging switching to manual driving when the calibration processing cannot be performed.

[0116] (Variation 5) In the first embodiment, the control device 6 may generate or update the alignment information IA at a predetermined timing other than immediately after the alignment adjustment. For example, when the control device 6 determines that the road on which the vehicle is traveling is a flat road, it generates alignment information IA based on the output signal of the attitude sensor 8 provided in each scanner housing section 50 and stores it in the memory unit 2. In this case, the control device 6 may determine whether or not the road on which the vehicle is traveling is a flat road based on the output of an acceleration sensor that detects the inclination of the vehicle, or may determine whether or not the road on which the vehicle is traveling is a flat road based on the current position information and road inclination information of the road data included in the map DB 20.

[0117] (Variation 6) In the first to third embodiments, the measurement system 100 may receive information required for processing from a server device that stores information equivalent to the map DB 20, instead of including the map DB 20.

[0118] FIG. 15 shows a configuration example according to this modification. In the example of FIG. 15, the server device 21 includes a delivery map DB 22 having a data structure similar to that of the map DB 20 shown in FIG. 14, and transmits data such as feature information including calibration feasibility information Ic to the measurement system 100 mounted on each vehicle. In this case, the measurement system 100 may be a communication device such as an on-board device, or may be a system built into the vehicle. Then, in the case of the third embodiment, for example, the measurement system 100 performs calibration based on the flowchart of FIG. 13 based on the feature information received from the server device 21. In this case, the measurement system 100 can accurately determine whether the features around the vehicle are suitable as reference features based on the calibration feasibility information Ic, and can perform calibration necessity determination and calibration. [Explanation of symbols]

[0119] 1 Input section 2 Storage section 3 Sensor section 4. Notification Department 5. Communications Department 6. Control device 100 Measurement System

Claims

[Claim 1] A first acquisition unit that acquires an output signal from a detection device that can detect features present around the moving object; a second acquisition unit that acquires first information regarding an inclination of a feature recognized based on an output signal of the detection device with respect to a detection range of the detection device; a control unit that controls a detection range of the detection device based on the first information and second information related to an angle of the feature with respect to a road surface stored in a storage unit; A control device comprising:

Images