Lidar Sensor Calibration

Abstract

Calibrate the lidar sensors on the vehicle. The method includes positioning a vehicle away from a target including a planar mirror and features surrounding the mirror. The vehicle is positioned and oriented relative to the mirror such that the optical axis of the lidar sensor is nominally parallel to the optical axis of the mirror and the target is nominally centered in the field of view of the lidar sensor. The method further includes using the lidar sensor to acquire a three-dimensional image of the target including images of the target features and a mirror image of the vehicle formed by the mirror. The method further includes determining deviations from an expected alignment of the lidar sensor with respect to the vehicle by analyzing the images of the features and the mirror image of the vehicle in the three-dimensional image of the target.

Description

[Technical Field]

[0001]

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 915,563, entitled "DYNAMIC CALIBRATION OF LIDAR IMAGING SENSORS," filed October 15, 2019, U.S. Patent Application No. 17 / 069,727, entitled "CALIBRATION OF LIDAR SENSORS," filed October 13, 2020, and U.S. Patent Application No. 17 / 069,733, entitled "DYNAMIC CALIBRATION OF LIDAR SENSORS," filed October 13, 2020, the contents of which are incorporated by reference in their entireties. [Background technology]

[0002]

[0002] Three-dimensional sensors can be applied in autonomous vehicles, drones, robotics, security applications, and the like. A lidar (LiDAR) sensor is a type of three-dimensional sensor that can achieve high angular resolution suitable for such applications. A lidar sensor can include one or more laser sources for emitting laser pulses and one or more detectors for detecting reflected laser pulses. The lidar sensor measures the time it takes each laser pulse to travel from the lidar sensor to an object within the sensor's field of view, then bounce off the object and return to the lidar sensor. Based on the time-of-flight of the laser pulse, the lidar sensor determines how far the object is from the lidar sensor. By scanning across a scene, a three-dimensional image of the scene can be acquired.

[0003] For accurate measurements, the orientation of the optical axis of the lidar sensor may need to be calibrated relative to some mechanical reference point, such as a mounting hole on the lidar sensor's case. Additionally, when mounted on a vehicle, the position and orientation of the lidar sensor may need to be calibrated relative to the vehicle. Such calibration may occur, for example, at the manufacturer's plant. In the event of a collision or other mechanical disturbance to the lidar sensor, the calibration relative to either the case or the vehicle may change. Thus, to ensure safe and accurate long-term operation of the lidar sensor, it may be desirable to be able to detect loss of calibration accuracy and correct the calibration. Summary of the Invention

[0004] According to some embodiments, a method for calibrating a lidar sensor mounted on a vehicle includes positioning the vehicle at a distance from a target. The target includes a planar mirror and features surrounding the mirror. The optical axis of the mirror is substantially horizontal. The vehicle is positioned and oriented relative to the mirror such that the optical axis of the lidar sensor is nominally parallel to the optical axis of the mirror and the target is nominally centered in the field of view of the lidar sensor. The method further includes acquiring a three-dimensional image of the target using the lidar sensor. The three-dimensional image of the target includes images of the features of the target and a mirror image of the vehicle formed by the mirror. The method further includes determining a deviation from an expected alignment of the lidar sensor with the vehicle by analyzing the images of the features in the three-dimensional image of the target and the mirror image of the vehicle.

[0005] According to some embodiments, a method for calibrating a lidar sensor mounted on a vehicle includes storing a reference three-dimensional image acquired by the lidar sensor while the lidar sensor is at an expected alignment with the vehicle. The reference three-dimensional image includes a first image of a fixed feature on the vehicle. The method further includes acquiring a three-dimensional image using the lidar sensor, the three-dimensional image including a second image of the fixed feature, and determining a deviation from the expected alignment of the lidar sensor with the vehicle by comparing the second image of the fixed feature in the three-dimensional image with the first image of the fixed feature in the reference three-dimensional image.

[0006] According to some embodiments, a method for calibrating a lidar sensor mounted on a vehicle includes acquiring one or more three-dimensional images using the lidar sensor while the vehicle is traveling along a road having fixed road features, each of the one or more three-dimensional images including an image of the road features. The method further includes analyzing a spatial relationship between the image of the road features in the one or more three-dimensional images and an orientation of a field of view of the lidar sensor, and determining a deviation from an expected positional alignment of the lidar sensor with respect to the vehicle based on the spatial relationship between the image of the road features and the field of view of the lidar sensor. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 illustrates an exemplary lidar sensor for three-dimensional imaging in accordance with some embodiments. [Figure 2A] FIG. 1 is a diagram that schematically illustrates a lidar sensor mounted on a vehicle, in accordance with some embodiments. [Figure 2B] FIG. 2 illustrates a vehicle coordinate system, according to some embodiments. [Figure 2C] FIG. 1 illustrates a lidar coordinate system, according to some embodiments. [Figure 3A] 1A-1C are diagrams that schematically illustrate mounting mechanisms that can be used to mount a lidar sensor on a vehicle, in accordance with some embodiments. [Figure 3B] 1A-1C are diagrams that schematically illustrate mounting mechanisms that can be used to mount a lidar sensor on a vehicle, in accordance with some embodiments. [Figure 4A] FIG. 1 illustrates an example calibration setup for a lidar sensor mounted on a vehicle, in accordance with some embodiments. [Figure 4B] FIG. 1 illustrates an example calibration setup for a lidar sensor mounted on a vehicle, in accordance with some embodiments. [Figure 4C] FIG. 1 illustrates an example calibration setup for a lidar sensor mounted on a vehicle, in accordance with some embodiments. [Figure 5A] 1A-1C are diagrams illustrating schematics of what a lidar sensor sees under certain alignment conditions, in accordance with some embodiments. [Figure 5B] 1A-1C are diagrams illustrating schematic diagrams of what a lidar sensor sees under different alignment conditions, in accordance with some embodiments. [Figure 5C] 1A-1C are diagrams illustrating schematic diagrams of what a lidar sensor sees under different alignment conditions, in accordance with some embodiments. [Figure 6A] 1A-1C are diagrams illustrating schematics of what a lidar sensor sees under certain alignment conditions, in accordance with some embodiments. [Figure 6B] 1A-1C are diagrams illustrating schematic diagrams of what a lidar sensor sees under different alignment conditions, in accordance with some embodiments. [Figure 6C] 1A-1C are diagrams illustrating schematic diagrams of what a lidar sensor sees under different alignment conditions, in accordance with some embodiments. [Figure 6D] 1A-1C are diagrams illustrating schematic diagrams of what a lidar sensor sees under different alignment conditions, in accordance with some embodiments. [Figure 7] 1 is a simplified flowchart illustrating a method for calibrating a lidar sensor mounted on a vehicle using a target with an embedded mirror, according to some embodiments. [Figure 8]FIG. 1 illustrates a lidar sensor mounted behind the windshield of a vehicle, including a mask, in accordance with some embodiments. [Figure 9A] 9 is a schematic side view of a windshield having the mask shown in FIG. 8. FIG. [Figure 9B] 9 is a schematic side view of a windshield having the mask shown in FIG. 8. FIG. [Figure 10A] 9A-9C are diagrams illustrating schematic examples of the effect of the mask shown in FIG. 8 on the field of view of a lidar sensor under certain alignment conditions, in accordance with some embodiments. [Figure 10B] 9A-9C are diagrams illustrating another example of the effect of the mask shown in FIG. 8 on the field of view of the lidar sensor under different alignment conditions, in accordance with some embodiments. [Figure 10C] 9A-9C are diagrams illustrating another example of the effect of the mask shown in FIG. 8 on the field of view of the lidar sensor under different alignment conditions, in accordance with some embodiments. [Figure 10D] 1A-1C are diagrams that schematically illustrate example images of particular features of a vehicle acquired by a lidar sensor, in accordance with some embodiments. [Figure 10E] FIG. 2 is another diagram that schematically illustrates an example image of a particular feature of a vehicle acquired by a lidar sensor, in accordance with some embodiments. [Figure 10F] FIG. 2 is another diagram that schematically illustrates an example image of a particular feature of a vehicle acquired by a lidar sensor, in accordance with some embodiments. [Figure 11] 1 is a simplified flowchart illustrating a method for calibrating a lidar sensor mounted on a vehicle using on-vehicle features, according to some embodiments. [Figure 12A] FIG. 1 illustrates a method for dynamic calibration of a vehicle-mounted lidar sensor using lane markings on a road, according to some embodiments. [Figure 12B] FIG. 1 illustrates a method for dynamic calibration of a vehicle-mounted lidar sensor using lane markings on a road, according to some embodiments. [Figure 13A]FIG. 1 illustrates a method for dynamic calibration of a vehicle-mounted lidar sensor using lane markings on a road, according to some embodiments. [Figure 13B] FIG. 1 illustrates a method for dynamic calibration of a vehicle-mounted lidar sensor using lane markings on a road, according to some embodiments. [Figure 13C] FIG. 1 illustrates a method for dynamic calibration of a vehicle-mounted lidar sensor using lane markings on a road, according to some embodiments. [Figure 14] 1 is a simplified flowchart illustrating a method for calibrating a lidar sensor mounted on a vehicle using road features, according to some embodiments. DETAILED DESCRIPTION OF THE INVENTION

[0008]

[0022] According to some embodiments, a method for calibrating a lidar sensor mounted on a vehicle is provided. Calibration may not require returning the vehicle to a manufacturer's plant or repair shop. Calibration can be performed periodically or continuously, even while the vehicle is parked or in operation.

[0009]

[0023] FIG. 1 illustrates an exemplary lidar sensor 100 for three-dimensional imaging in accordance with some embodiments. The lidar sensor 100 includes an emission lens 130 and a receiving lens 140. The lidar sensor 100 includes a light source 110a disposed substantially at the back focal plane of the emission lens 130. The light source 110a operates to emit light pulses 120 from respective emission positions within the back focal plane of the emission lens 130. The emission lens 130 is configured to collimate and direct the light pulses 120 toward an object 150 located in front of the lidar sensor 100. For a given emission position of the light source 110a, the collimated light pulses 120′ are directed at a corresponding angle toward the object 150.

[0010]

[0024] A portion 122 of the collimated light pulse 120' is reflected from the object 150 toward the receiving lens 140. The receiving lens 140 is configured to focus the portion 122' of the light pulse reflected from the object 150 at a corresponding detection location within the focal plane of the receiving lens 140. The lidar sensor 100 further includes a detector 160a disposed substantially in the focal plane of the receiving lens 140. The detector 160a is configured to receive and detect the portion 122' of the light pulse 120 reflected from the object at the corresponding detection location. The corresponding detection location of the detector 160a is optically conjugate with the respective emission location of the light source 110a.

[0011]

[0025] The light pulses 120 may be of short duration, for example, a pulse width of 10 ns. The lidar sensor 100 further includes a processor 190 coupled to the light source 110a and the detector 160a. The processor 190 is configured to determine a time-of-flight (TOF) of the light pulses 120 from emission to detection. Because the light pulses 120 travel at the speed of light, the distance between the lidar sensor 100 and the object 150 can be determined based on the determined time-of-flight.

[0012]

[0026] One method for scanning the laser beam 120′ across the FOV is to move the light source 110a laterally relative to the output lens 130 within the back focal plane of the output lens 130. For example, the light source 110a may be raster scanned across multiple output positions in the back focal plane of the output lens 130, as shown in FIG. 1 . The light source 110a may emit multiple light pulses at the multiple output positions. Each light pulse emitted at each output position is collimated by the output lens 130, directed at the object 150 at a respective angle, and strikes a corresponding point on the surface of the object 150. Thus, as the light source 110a is raster scanned within a particular region within the back focal plane of the output lens 130, a corresponding object region on the object 150 is scanned. The detector 160a may be raster scanned to be positioned at multiple corresponding detection positions within the focal plane of the receiving lens 140, as shown in FIG. 1 . The scanning of detector 160a is typically synchronized with the scanning of light source 110a, so that detector 160a and light source 110a are optically conjugate to each other at any given time.

[0013]

[0027] By determining the time of flight of each light pulse emitted at each emission position, the distance from the lidar sensor 100 to each corresponding point on the surface of the object 150 can be determined. In some embodiments, the processor 190 is coupled to a position encoder that detects the position of the light source 110a at each emission position. Based on the emission position, the angle of the collimated light pulse 120′ can be determined. The X and Y coordinates of the corresponding point on the surface of the object 150 can be determined based on the angle and the distance to the lidar sensor 100. Thus, a three-dimensional image of the object 150 can be constructed based on the measured distances from the lidar sensor 100 to various points on the surface of the object 150. In some embodiments, the three-dimensional image can be represented as a point cloud, i.e., a set of X, Y, and Z coordinates of points on the surface of the object 150.

[0014]

[0028] In some embodiments, the intensity of the returned light pulse 122' is measured and used to adjust the power of subsequent light pulses from the same emission point to prevent detector saturation, improve eye safety, or reduce overall power consumption. The power of the light pulse can be varied by changing the duration of the light pulse, the voltage or current applied to the laser, or the charge stored on the capacitor used to power the laser. In the latter case, the charge stored on the capacitor can be varied by changing the charging time, charging voltage, or charging current to the capacitor. In some embodiments, the reflectance, determined by the intensity of the detected pulse, can also be used to add another dimension to the image. For example, the image may include X, Y, and Z coordinates as well as reflectance (or brightness).

[0015]

[0029] The angle of view (AFOV) of the lidar sensor 100 may be estimated based on the scanning range of the light source 110a and the focal length of the output lens 130 as follows:

number

[0016]

[0030] The light source 110a can be configured to emit light pulses in the ultraviolet, visible, or near-infrared wavelength range. The energy of each light pulse can be on the order of microjoules, which is typically considered eye-safe for repetition rates in the kHz range. For light sources operating at wavelengths above approximately 1500 nm, higher energy levels can be used because the eye does not focus on those wavelengths. The detector 160a can include a silicon avalanche photodiode, a photomultiplier tube, a PIN diode, or other semiconductor sensor.

[0017]

[0031] When a lidar sensor, such as the lidar sensor 100 shown in FIG. 1, is used for obstacle detection in an autonomous vehicle, the position and orientation of the lidar sensor relative to the vehicle may need to be precisely known in order to accurately determine the location of an obstacle relative to the vehicle. FIG. 2A schematically illustrates a lidar sensor 210 mounted on a vehicle 220. For example, the lidar sensor 210 may be mounted at the top center behind the windshield of the vehicle 220. The lidar sensor 210 may be characterized by an optical axis 250 (e.g., the optical axis of the exit lens or the receiving lens). The vehicle 220 may have a longitudinal axis 240. It may be advantageous to align the optical axis 250 of the lidar sensor 210 with the longitudinal axis 240 of the vehicle 220 so that the lidar sensor 210 looks straight ahead in the direction the vehicle 220 is traveling.

[0018]

[0032] Assume that the lidar sensor 210 is initially calibrated according to this alignment condition. If the lidar sensor becomes misoriented (e.g., turning left) due to some mechanical disturbance, the optical axis 250 of the lidar sensor 210 will no longer be aligned with the longitudinal axis 240 of the vehicle 220, as shown in FIG. 2A . The misalignment may result in an inaccurate measurement of the position of an obstacle (e.g., a person 260) relative to the vehicle 220. Thus, recalibration of the lidar sensor 210 may be required.

[0019]

[0033] FIG. 2B illustrates a vehicle coordinate system. The vehicle coordinate system may have three translational degrees of freedom, which may be represented, for example, by X, Y, and Z coordinates. The vehicle coordinate system may have three rotational degrees of freedom, which may be represented, for example, by roll, pitch, and yaw angles about the x, y, and z axes, respectively. For example, the x-axis may be along the longitudinal axis of the vehicle 220, the z-axis may be along the vertical direction, and the y-axis may be along the lateral direction. The position and orientation of the lidar sensor 210 relative to the vehicle 220 may be characterized by (x, y, z, roll, pitch, yaw) coordinates in the vehicle coordinate system.

[0020]

[0034] 2C shows a lidar coordinate system. The lidar coordinate system may also have three translational degrees of freedom (e.g., represented by X, Y, and Z coordinates) and three rotational degrees of freedom (e.g., represented by roll, pitch, and yaw angles about the X, Y, and Z axes, respectively). For example, the X axis may be along the optical axis of the lidar sensor 210, the Z axis may be along the nominal vertical direction, and the Y axis may be along a direction orthogonal to the X and Z axes.

[0021]

[0035] The raw point cloud data acquired by the lidar sensor 210 may be in a lidar coordinate system. To determine the position of an obstacle relative to the vehicle 220, the point cloud data can be transformed to the vehicle coordinate system if the position and orientation of the lidar sensor 210 in the vehicle coordinate system is known. The transformation from the lidar coordinate system to the vehicle coordinate system may be referred to herein as calibrating the lidar sensor 210 relative to the vehicle 220.

[0022]

[0036] According to some embodiments, the lidar sensor 210 can be mounted on the vehicle 220 so that the lidar sensor 210 is nominally aligned with respect to the vehicle 220. For example, the lidar sensor 210 can be mounted on the vehicle 220 so that the x-axis of the lidar coordinate system (e.g., along the optical axis of the lidar sensor 210) is nominally aligned with the x-axis of the vehicle coordinate system (e.g., along the longitudinal axis of the vehicle), the y-axis of the lidar coordinate system is nominally aligned with the y-axis of the vehicle coordinate system, and the z-axis of the lidar coordinate system is nominally aligned with the z-axis of the vehicle coordinate system. Thus, the roll, pitch, and yaw angles in the vehicle coordinate system are all approximately zero. Calibration may be required to correct any remaining deviations from the nominal alignment to sufficient accuracy. For example, it may be desirable to calibrate the lidar sensor 210 to a translational accuracy of 2 cm along each of the x-, y-, and z-axes, and a rotational accuracy of 0.1 degrees for each of the roll, pitch, and yaw angles.

[0023]

[0037] 3A and 3B schematically illustrate mounting mechanisms that can be used to mount a lidar sensor to a vehicle, according to some embodiments. The lidar sensor may have an outer housing 310 with three mounting holes 320 (e.g., on the top surface of the outer housing 310), as shown in FIG. 3A. A bracket 330 may have three holes 340 that align with the mounting holes 320 in the outer housing 310, thereby allowing the lidar sensor to be mounted to the bracket 330 in a fixed orientation, as shown in FIG. 3B. The bracket 330 may be mounted to the vehicle (e.g., on the roof of the passenger compartment) using similar mounting holes (not shown) for proper alignment.

[0024]

[0038] The lidar sensor can be pre-calibrated at the manufacturer's plant to correct for any residual misalignment of the lidar sensor relative to the vehicle. Over time, during vehicle operation, the lidar sensor's optics can shift relative to the housing 310, or the lidar sensor's housing 310 can shift relative to the mounting bracket 330 and / or the vehicle. This can occur, for example, due to tearing and wear of the lidar sensor's internal mechanisms, vehicle collisions or vibrations, tire misalignment, or aging of the vehicle's suspension. Thus, over time, the calibration can become inaccurate, requiring a new calibration. According to various embodiments, lidar calibration can be performed periodically using a target with an embedded mirror, or periodically or continuously using fixed features on the vehicle or fixed road features, as described in more detail below.

[0025]

[0039] A. Calibrating a Lidar Sensor Using a Target with an Embedded Mirror 4A-4C illustrate an exemplary calibration setup for a lidar sensor mounted on a vehicle using a target with an embedded mirror, according to some embodiments. The lidar sensor 410 is shown mounted behind the windshield of the vehicle 420 in this example. However, this is not required. For example, the lidar sensor 410 could be mounted in other locations on the vehicle 420, such as the front bumper, rear window, or rear bumper. A target 440 is positioned a distance D in front of the lidar sensor 410. The target 440 includes a flat mirror 430 and an identifiable feature 490 surrounding the mirror 430.

[0026]

[0040] Referring to FIG. 4A , the target 440 can be mounted on a stand (not shown) so that the mirror 430 is nominally vertical. Thus, the optical axis 432 of the mirror 430, which is perpendicular to the surface of the mirror 430, can be nominally horizontal. The vehicle 420 is positioned and oriented relative to the mirror 430 so that the optical axis 412 of the lidar sensor is nominally parallel to the optical axis 432 of the mirror 430, and the target 440 is nominally centered in the field of view of the lidar sensor 410. If the lidar sensor 410 is mounted behind the windshield of the vehicle 420 (e.g., as shown in FIG. 4A ), the vehicle 420 can be positioned so that the longitudinal axis 422 of the vehicle 420 is nominally parallel to the optical axis 432 of the mirror 430.

[0027]

[0041] Referring to FIG. 4B, the lidar sensor 410 can acquire a three-dimensional image of the target 440 by scanning across the target 440. The three-dimensional image of the target can include images of features 490 surrounding the mirror 430 and a mirror image of the vehicle 420 formed by the mirror 430. For example, the lidar sensor 410 can emit a laser pulse 450 (or 450a) directed toward the mirror 430. The reflected laser pulse 450b can be directed toward a particular portion of the vehicle (e.g., the license plate on the front bumper), then reflected toward the mirror 430 (shown as 450c in FIG. 4B) and back to the lidar sensor 410 (shown as 450d in FIG. 4B). Thus, as shown in FIG. 4C, the lidar sensor 410 can effectively acquire a three-dimensional mirror image of the vehicle 420.

[0028]

[0042] According to some embodiments, the distance D between the target 440 and the vehicle 420 can be large enough to meet the desired accuracy. For example, the distance D can be approximately 3 m. The size of the mirror 430 can be large enough to view the entire vehicle 420, but this is not required. For example, the size of the mirror 430 can be approximately 1 m high and approximately 2 m wide. The size of the target 440 can be approximately 2 m high and approximately 3 m wide (e.g., the edge of the target 440 having the feature 490 can have a width of 0.5 m).

[0029]

[0043] 5A shows a schematic representation of what the lidar sensor 410 sees when the vehicle 420 is properly aligned with the target 440 (e.g., the longitudinal axis of the vehicle 420 is perpendicular to the surface of the mirror 430) and the lidar sensor 410 is properly aligned with the vehicle 420 (e.g., the optical axis of the lidar sensor 410 is parallel to the longitudinal axis of the vehicle 420). We also assume that the vehicle 420 is laterally centered relative to the width of the mirror 430 and that the lidar sensor 410 is laterally centered relative to the vehicle 420. In such a case, the mirror image 420' of the vehicle 420 seen by the lidar sensor 410 may appear symmetrical and have no roll, pitch, or yaw angles. The target 440 also appears symmetrical with respect to the lidar sensor's field of view 414 (e.g., the lateral margin d from the left edge of the target 440 to the left boundary of the field of view 414 is approximately the same as the lateral margin d from the right edge of the target 440 to the right boundary of the field of view 414), and may have no roll, pitch, or yaw angles.

[0030]

[0044] FIG. 5B schematically illustrates what the lidar sensor 410 sees when the lidar sensor 410 is correctly aligned with the vehicle 420 but the vehicle 420 is misaligned with the target 440 (e.g., the longitudinal axis of the vehicle 420 has a finite yaw angle with respect to the optical axis of the mirror 430). In such a case, the target 440 is no longer centered in the lidar's field of view 414 (e.g., shifted laterally to the right). The amount of shift (e.g., the difference between the new margin D and the nominal margin d) may be related to the amount of misalignment. Thus, it may be possible to determine the amount of misalignment of the vehicle 420 relative to the target 440 (e.g., yaw error) based on the amount of shift. Additionally, the mirror image 420′ of the vehicle 420 seen by the lidar sensor 410 also appears to have a finite yaw angle (note that the amount of rotation is somewhat exaggerated in FIG. 5B). Thus, the lidar sensor 410 can also determine the amount of misalignment of the vehicle 420 relative to the target 440 by measuring the distance of certain features of the vehicle 420. For example, as shown in FIG. 5B, the lidar sensor 410 can determine that one headlamp (e.g., the left headlamp) is farther from the lidar sensor 410 than the other headlamp (e.g., the right headlamp), thereby allowing the yaw angle of the vehicle 420 to be calculated.

[0031]

[0045] FIG. 5C schematically illustrates what the lidar sensor 410 sees when the vehicle 420 is correctly aligned with respect to the target 440 (e.g., the longitudinal axis of the vehicle 420 is perpendicular to the surface of the mirror 430), but the lidar sensor 410 is incorrectly aligned with respect to the vehicle 420 (e.g., the lidar sensor 410 is looking left instead of straight ahead). In such a case, the mirror image 420′ of the vehicle 420 may appear symmetrical, but the target 440 may be shifted laterally (e.g., to the right) relative to the lidar's field of view 414. Based on the amount of shift (e.g., the difference between the new margin D and the nominal margin d), the lidar sensor 410 can determine the amount of misalignment of the lidar sensor 410 (e.g., yaw error).

[0032]

[0046] FIG. 6A schematically illustrates what the lidar sensor 410 sees when the lidar sensor 410 is correctly aligned with the vehicle 420, but the vehicle 420 has a pitch error relative to the target 440. In such a case, the target 440 is no longer centered in the lidar's field of view 414 (e.g., shifted vertically downward). The amount of vertical shift may be related to the amount of pitch error. Thus, it may be possible to determine the amount of pitch error of the vehicle 420 relative to the target 440 based on the amount of vertical shift. In addition, the mirror image 420' of the vehicle 420 seen by the lidar sensor 410 also appears to have a finite pitch angle (note that the amount of rotation is somewhat exaggerated in FIG. 6A). Thus, the lidar sensor 410 can also determine the amount of pitch error of the vehicle 420 relative to the target 440 by measuring the distance of certain features of the vehicle 420. For example, as shown in FIG. 6A, the lidar sensor 410 can measure that the roof of the vehicle 420 is tilted and calculate the pitch error based on the amount of tilt.

[0033]

[0047] 6B shows a schematic of what the lidar sensor 410 sees when the vehicle 420 is correctly aligned with respect to the target 440, but the lidar sensor 410 has a pitch error relative to the vehicle 420 (e.g., the lidar sensor 410 is looking upward instead of straight ahead). In such a case, the mirror image 420' of the vehicle 420 may appear to have no pitch angle, but the target 440 may be shifted vertically (e.g., downward) relative to the lidar's field of view 414. Based on the amount of vertical shift, the lidar sensor 410 can determine the amount of pitch error of the lidar sensor 410.

[0034]

[0048] 6C shows a schematic of what the lidar sensor 410 sees when the lidar sensor 410 is correctly aligned with the vehicle 420, but the vehicle 420 has a roll error relative to the target 440. In such a case, the mirror image 420' of the vehicle 420 seen by the lidar sensor 410 may appear to have no roll angle, while the target 440 may appear to have a finite roll angle relative to the lidar field of view 414. Based on the amount of roll angle of the target 440 relative to the lidar sensor field of view 414, it may be possible to determine the amount of roll error of the vehicle 420 relative to the target 440.

[0035]

[0049] 6D shows a schematic of what the lidar sensor 410 sees when the vehicle 420 is correctly aligned with respect to the target 440, but the lidar sensor 410 has a roll error relative to the vehicle 420. In such a case, both the mirror image 420' of the vehicle 420 and the target 440 may appear to have a finite roll angle relative to the lidar field of view 414. Based on the amount of roll rotation, the lidar sensor 410 can determine the amount of roll error of the lidar sensor 410.

[0036]

[0050] Thus, the lidar sensor 410 (or a computing unit of the vehicle 420) can determine deviations from an expected alignment of the lidar sensor 410 with respect to the vehicle 420 (e.g., an initial alignment performed at the manufacturer's plant) by analyzing three-dimensional images acquired by the lidar sensor 410, including images of the features 490 on the target and a mirror image 420′ of the vehicle 420. The deviations from the expected alignment can include yaw error, roll error, pitch error, and translation error (e.g., δx, δy, δz).

[0037]

[0051] According to some embodiments, determining the deviation of the lidar sensor 410 from an expected alignment with the vehicle 420 may include the following steps: The position and orientation of the lidar sensor 410 with respect to the target 440 may be determined based on an image of the features 490 on the target 440. The position and orientation of the lidar sensor 410 with respect to a mirror image 420' of the vehicle 420 may be determined based on the mirror image 420' of the vehicle. A transformation from the lidar coordinate system to the vehicle coordinate system may then be determined based on (i) the position and orientation of the lidar sensor 410 with respect to the target 440 and (ii) the position and orientation of the lidar sensor 410 with respect to the mirror image 420' of the vehicle 420.

[0038]

[0052] In some embodiments, the computing unit can store a reference image. For example, the reference image can be an image acquired by the lidar sensor 410 immediately after initial alignment is performed at the manufacturer's plant, or it can be a simulated image of the expected alignment. During recalibration, the computing unit can compare the three-dimensional image acquired by the lidar sensor 410 to the reference image and perform multivariate minimization (e.g., using gradient descent or other algorithms) to determine a transformation matrix that most closely matches the acquired three-dimensional image with the reference image. Deviations from the expected alignment (e.g., yaw error, roll error, pitch error, δx, δy, and δz) can then be derived from the transformation matrix.

[0039]

[0053] The following exemplary method may be used to determine the relationship of the lidar sensor 410 to the vehicle 420 in six degrees of freedom according to various embodiments. Other methods and techniques may also be used by those skilled in the art. In the description of the following exemplary embodiments, the following notation and terminology are used: t represents a matrix describing the relationship of the lidar sensor 410 to the target 440, and C t represents a matrix describing the relationship of the vehicle 420 to the target 440, and L Crepresents a matrix describing the relationship of the lidar sensor 410 to the vehicle 420 (to correct for miscalibration of the lidar sensor 410), M represents the mirror transformation matrix, and L mC represents a matrix that describes the relationship of the lidar sensor 410 to the vehicle 420 as it looks in the mirror 430.

[0040]

[0054] In some embodiments, the lidar sensor 410 can establish its positional (x, y, z) relationship to the target features 490 around the mirror 430 by triangulating the distance from at least three target features 490 (e.g., similar to how a GPS receiver triangulates its position relative to GPS satellites). The pitch, roll, and yaw rotational relationships can be determined by measuring the positions of at least three target features 490 (which can be the same target features used for x, y, and z). Thus, the lidar sensor 410 can establish its positional (x, y, z) relationship to the target features 490 around the mirror 430. t A matrix can be established that describes the relationship of the lidar sensor 410 to the matrix L t is a 4x4 matrix that defines the x, y, and z positions, as well as the pitch, roll, and yaw.

[0041]

[0055] The relationship of vehicle 420 to target 440 can then be established by a similar procedure, namely, by measuring the distance to a particular feature of vehicle 420 seen in mirror 430 to triangulate its position relative to target 440, and by measuring the position of the feature on vehicle 420 relative to mirror 430 to determine the pitch, roll, and yaw of vehicle 420. Thus, target 440C t A matrix can be established that describes the relationship of the vehicle 420 to the matrix C. t is a 4x4 matrix that defines the x, y, and z positions, as well as the pitch, roll, and yaw. The relationship of the lidar sensor 410 to the vehicle 420 is then expressed as L C =(C t -1 )·L t It can be defined by:

[0042]

[0056] According to some embodiments, the lidar sensor 410 calculates the matrix L t Its position relative to the target 440 can be determined as described above to establish the matrix L. mC The relationship of the vehicle's mirror image 420′ to the lidar sensor 410 as seen through the mirror 430 can also be determined to establish θ. The relationship of the lidar sensor 410 to the vehicle 420 can then be determined by matrix multiplication as follows: L C =L t M(L t -1 )L mC M

[0043]

[0057] According to some embodiments, the image 420′ of the vehicle seen in the mirror 430 and the image of the target 440 acquired by the lidar sensor 410 are compared to a stored reference image. The reference image can be either a simulated image or an image taken during factory calibration. A minimization technique (e.g., using a gradient descent algorithm) can then be used to determine transformation parameters (δx, δy, δz, pitch error, roll error, and yaw error) that minimize the difference between the currently acquired image and the stored reference image. In this process, the currently acquired image can be transformed to match the stored reference image, or the stored reference image can be transformed to match the currently acquired image. The transformation parameters can represent the difference between the current lidar position relative to the vehicle 420 and the ideal (or factory-calibrated) lidar position.

[0044]

[0058] Once the relationship of the lidar 410 to the vehicle 420 is determined, any discrepancy in this relationship with the current lidar calibration can be used to correct the lidar calibration.

[0045]

[0059] In some embodiments, one or more distance sensors can be used to determine the position and orientation of the vehicle 420 relative to the target 440 (e.g., in a world coordinate system). For example, by placing two or three distance sensors in the pavement beneath the vehicle 420, the pitch, roll, and z coordinate (height) of the vehicle can be determined. By using six distance sensors, all degrees of freedom (x, y, z, pitch, roll, yaw) can be determined. The distance sensors can be ultrasonic or laser sensors.

[0046]

[0060] An example of using distance sensors to determine the position and orientation of the vehicle 420 is shown in FIG. 5A. In this example, four distance sensors 590 can be placed at appropriate locations around the vehicle 420. For example, each distance sensor 590 can be placed near a respective wheel of the vehicle 420. The distance sensors 590 can estimate the distance from the wheel by sending an ultrasonic or laser pulse toward the wheel and measuring the return pulse. Thus, the y-coordinate and yaw angle of the vehicle 420 can be determined. In some embodiments, only two distance sensors 590 on one side of the vehicle (either the driver's side or the passenger's side) may be required. Another distance sensor 592 can be placed in front of the vehicle 420 to determine the x-coordinate of the vehicle 420.

[0047]

[0061] According to some embodiments, additional corrections or calibrations can be performed. For example, the effects of windshield distortion can be measured and corrected. Windshield distortion correction may be required whenever the windshield is replaced.

[0048]

[0062] FIG. 7 shows a simplified flowchart illustrating a method 700 for calibrating a lidar sensor mounted on a vehicle using a target with an embedded mirror, according to some embodiments.

[0049]

[0063] The method 700 includes, at 702, positioning a vehicle at a distance from a target. The target includes a planar mirror and features surrounding the mirror. The optical axis of the mirror is approximately horizontal. The vehicle is positioned and oriented relative to the mirror such that the optical axis of the lidar sensor is nominally parallel to the optical axis of the mirror and the target is nominally centered in the field of view of the lidar sensor.

[0050]

[0064] The method 700 further includes acquiring a three-dimensional image of the target using the lidar sensor at 704. The three-dimensional image of the target includes images of the target's features and a mirror image of the vehicle formed by the mirror.

[0051]

[0065] The method 700 further includes, at 706, determining a deviation from an expected alignment of the lidar sensor with respect to the vehicle by analyzing the image of the feature in the three-dimensional image of the target and a mirror image of the vehicle.

[0052]

[0066] In some embodiments, the method 700 further includes, at 708, recalibrating the lidar sensor relative to the vehicle based on deviations from an expected alignment of the lidar sensor relative to the vehicle.

[0053]

[0067] In some embodiments, method 700 further includes determining that a deviation from the expected alignment of the lidar sensor exceeds a threshold, and issuing an alert in response to determining that the deviation from the expected alignment of the lidar sensor exceeds the threshold.

[0054]

[0068] In some embodiments, the field of view of the lidar sensor is less than 180 degrees horizontally.

[0055]

[0069] In some embodiments, determining a deviation from an expected alignment of the lidar sensor with respect to the vehicle may include determining a position and orientation of the lidar sensor with respect to the target based on the image of the feature; determining a position and orientation of the lidar sensor with respect to the mirror image of the vehicle based on a mirror image of the vehicle; and determining a transformation from the lidar coordinate system to the vehicle coordinate system based on (i) the position and orientation of the lidar sensor with respect to the target and (ii) the position and orientation of the lidar sensor with respect to the mirror image of the vehicle.

[0056]

[0070] In some embodiments, determining a deviation from an expected alignment of the lidar sensor with respect to the vehicle can include storing a reference matrix for an expected relationship between the lidar sensor and the vehicle, determining a matrix for a current relationship between the lidar sensor and the vehicle, and determining a deviation from an expected alignment of the lidar sensor with respect to the vehicle by comparing the matrix to the reference matrix. In some embodiments, method 700 further includes recalibrating the lidar sensor with respect to the vehicle based on the matrix for the current relationship between the lidar sensor and the vehicle.

[0057]

[0071] It should be understood that the specific steps illustrated in FIG. 7 provide a particular method for calibrating a lidar sensor according to some embodiments. Other orders of steps may be performed according to alternative embodiments. For example, alternative embodiments of the present invention may perform the steps outlined above in a different order. Additionally, individual steps illustrated in FIG. 7 may include multiple sub-steps that may be performed in various orders as appropriate for the individual step. Furthermore, additional steps may be added or some steps may be removed depending on the particular application. Those skilled in the art will recognize many variations, modifications, and alternatives.

[0058]

[0072] B. Dynamic Calibration of Lidar Sensors Using On-Vehicle Features It may be desirable to have a method for periodically or continuously checking the calibration of a lidar sensor for a vehicle, even while the vehicle is parked or in operation. According to some embodiments, the lidar sensor calibration can be performed using characteristics of the vehicle so that the calibration can occur during normal operation of the vehicle. Such a method is referred to herein as dynamic calibration.

[0059]

[0073] FIG. 8 illustrates an example of dynamic calibration of a lidar sensor using a windshield-mounted mask, according to some embodiments. As shown, a lidar sensor 810 (not visible in FIG. 8 ) is mounted behind a windshield 840 of a vehicle 850. A mask 820 is mounted to the windshield 840 in an area directly forward of the lidar sensor 810. Assuming the operating wavelength of the lidar sensor 810 is in the IR wavelength range, the mask 820 can be configured to block infrared (IR) light (e.g., be optically opaque to IR light). In some embodiments, the mask 820 can be annular in shape, surrounding an IR-transmitting portion 830 of the windshield 840. The geometry and size of the IR-blocking mask 820 can be fabricated to block the periphery of the edge of the field of view of the lidar sensor 810.

[0060]

[0074] 9A and 9B schematically illustrate side views of a windshield 840 with an IR-blocking mask 820, according to some embodiments. An IR-transparent portion 830 of the windshield 840 is surrounded by the IR-blocking mask 820. The IR-transparent portion 830 is slightly smaller than the field of view 910 of the lidar sensor 810. Referring to FIG. 9A , under proper alignment (e.g., expected alignment), the lidar sensor 810 can be positioned behind the windshield 840 so that its field of view 910 is centered on the IR-transparent portion 830. Thus, the edges of the field of view 910 are blocked on all four sides (the side view in FIG. 9A shows the top and bottom).

[0061]

[0075] 9B shows an example where the lidar sensor 810 is misaligned (e.g., tilted upward) from its correct alignment. As a result, the IR transparent portion 830 is no longer centered on the field of view 910 of the lidar sensor 810. Thus, a larger portion of its field of view 910 may be blocked above than below.

[0062]

[0076] 10A-10C schematically illustrate some examples of the effect of the IR-blocking mask 820 shown in FIG. 8 under various alignment conditions, according to some embodiments. FIG. 10A illustrates an example in which the lidar sensor 810 is properly aligned. The IR-transparent portion 830 of the windshield 840 is centered on the field of view 910 of the lidar sensor 810. Thus, the edges of the field of view 910 blocked by the IR-blocking mask 820 (the gray area) have approximately the same width on all four sides.

[0063]

[0077] 10B illustrates an example in which the lidar sensor 810 is misaligned to the right from its correct alignment (e.g., the lidar sensor 810 has a yaw error). As a result, the IR-transparent portion 830 of the windshield 840 is misaligned to the left relative to the field of view 910 of the lidar sensor 810. Thus, a larger portion of the field of view 910 is blocked by the IR-blocking mask 820 on the right side than on the left side.

[0064]

[0078] 10C illustrates an example in which the lidar sensor 810 is rotated from its correct alignment (e.g., the lidar sensor 810 has a roll error). As a result, the IR-transparent portion 830 of the windshield 840 is rotated relative to the field of view 910 of the lidar sensor 810.

[0065]

[0079] According to some embodiments, the relative position and orientation of the IR-transparent portion 830 of the windshield 840 with respect to the field of view 910 of the lidar sensor 810 can be used to calibrate the lidar sensor 810. For example, the computing unit can store a reference image. The reference can be acquired by the lidar sensor 810 while the lidar sensor is in the correct alignment position, or it can be acquired by simulation. For example, the reference image can be acquired by the lidar sensor immediately after the lidar sensor is pre-calibrated at a manufacturing facility. When the lidar sensor 810 is in normal operation (when the vehicle is parked or moving), the computing unit can periodically or continuously compare the current lidar image with the reference image. Based on the comparison, a deviation of the lidar sensor 810 from the correct alignment position can be generated.

[0066]

[0080] In some embodiments, multivariate minimization can be performed to determine a transformation matrix that best aligns the IR-transparent portion of the windshield 840 in the current lidar image with the IR-transparent portion 830 in the reference image. Deviations from the correct alignment (e.g., yaw error, roll error, pitch error, δx, δy, and δz) can then be derived from the transformation matrix. According to various embodiments, the lidar sensor 810 can automatically recalibrate itself or can alert the vehicle in response to determining that the deviation from the correct alignment exceeds a threshold.

[0067]

[0081] Additionally or alternatively, images of several fixed features on the vehicle acquired by the lidar sensor 810 can also be used to calibrate the lidar sensor 810. FIGS. 10D-10F show several example images that can be acquired by the lidar sensor 810. The images show a portion of a vehicle's hood 1020 and a decorative feature 1030 (e.g., a manufacturer's logo) attached to the hood 1020. In FIG. 10D, the decorative feature 1030 is approximately centered laterally within the field of view 910 of the lidar sensor 810. In FIG. 10E, the decorative feature 1030 is shifted to the left of the field of view 910, indicating that the lidar sensor 810 may be pointing toward the right (e.g., having a yaw error). In FIG. 10F, the decorative feature 1030 is rotated, indicating that the lidar sensor 810 may be rotated (e.g., having a roll error). Note that the images of the hood 1020 in FIGS. 10E and 10F are also shifted and / or rotated accordingly. Other exemplary vehicle features that can be used for calibration include a cover over a grill-mounted lidar sensor, headlamp or taillamp features (e.g., if the lidar sensor is installed inside a headlamp or taillamp), etc.

[0068]

[0082] According to some embodiments, the computing unit can store a reference image acquired by the lidar sensor while the lidar sensor is in correct alignment (e.g., expected alignment) with the vehicle. The reference image can include a first image of fixed features on the vehicle. When the lidar sensor 810 is in normal operation (e.g., when the vehicle is parked or moving), the computing unit can periodically or continuously compare a current lidar image to the reference image. The current lidar image includes a second image of fixed features on the vehicle. Deviation from correct alignment can be determined by comparing the position and orientation of the fixed features in the second image with the position and orientation in the first image.

[0069]

[0083] FIG. 11 shows a simplified flowchart illustrating a method 1100 for calibrating a lidar sensor mounted on a vehicle using on-vehicle features, according to some embodiments.

[0070]

[0084] The method 1100 includes, at 1102, storing a reference three-dimensional image acquired by the lidar sensor while the lidar sensor is in an expected alignment with the vehicle. The reference three-dimensional image includes a first image of a fixed feature on the vehicle.

[0071]

[0085] The method 1100 further includes, at 1104, acquiring a three-dimensional image using the lidar sensor, the three-dimensional image including a second image of the fixed feature.

[0072]

[0086] The method 1100 further includes, at 1106, determining a deviation from an expected alignment of the lidar sensor with the vehicle by comparing a second image of the fixed feature in the three-dimensional image with the first image of the fixed feature in the reference three-dimensional image.

[0073]

[0087] In some embodiments, the method 1100 further includes, at 1108, recalibrating the lidar sensor based on a deviation from an expected alignment of the lidar sensor with respect to the vehicle.

[0074]

[0088] In some embodiments, the method 1100 further includes determining a transformation to be applied to a second image of a fixed feature in the three-dimensional image to match the first image of the fixed feature in the reference three-dimensional image, and recalibrating the lidar sensor based on the transformation.

[0075]

[0089] In some embodiments, the method 1100 further includes determining that a deviation from the expected alignment of the lidar sensor exceeds a threshold, and issuing an alert in response to determining that the deviation from the expected alignment of the lidar sensor exceeds the threshold.

[0076]

[0090] In some embodiments, the reference three-dimensional image is acquired by a lidar sensor after the lidar sensor has been pre-calibrated at a manufacturing facility.

[0077]

[0091] In some embodiments, the fixed feature comprises a portion of the hood of the vehicle or an object attached to the hood.

[0078]

[0092] In some embodiments, the lidar sensor is positioned behind the vehicle windshield, and the fixed feature includes a mask attached to a region of the windshield directly in front of the lidar sensor. The mask is configured to block light at an operating wavelength of the lidar sensor and is shaped to block a portion of the field of view of the lidar sensor. The mask can have an outer boundary and an inner boundary, the inner boundary being sized such that the mask encroaches on the periphery of the field of view of the lidar sensor.

[0079]

[0093] It should be understood that the specific steps illustrated in FIG. 11 provide a particular method for calibrating a lidar sensor according to some embodiments. Other orders of steps may be performed according to alternative embodiments. For example, alternative embodiments of the present invention may perform the steps outlined above in a different order. Additionally, individual steps illustrated in FIG. 11 may include multiple sub-steps that may be performed in various orders as appropriate for the individual step. Furthermore, additional steps may be added or some steps may be removed depending on the particular application. Those skilled in the art will recognize many variations, modifications, and alternatives.

[0080]

[0094] C. Dynamic Calibration of Lidar Sensors Using Road Features According to some embodiments, a method for dynamic calibration of a vehicle-mounted lidar sensor can use road features as the vehicle travels along a relatively straight section of road. An example is shown in FIGS. 12A and 12B . A lidar sensor 1210 is shown mounted behind the windshield of a vehicle 1220 (the lidar sensor 1210 can be mounted in other locations, such as the front bumper). The vehicle 1220 is traveling along a straight section of road 1230. Painted lane markings 1240 can be used to dynamically calibrate the position and orientation of the lidar sensor 1210 relative to the vehicle 1220 based on lidar images acquired while the vehicle is traveling along the road 1230. Using painted lane markings for calibration can be advantageous because lane markings are present on nearly all roads. Furthermore, the distance between pairs of lane markings 1240 is typically a standard distance. Lane markings using retroreflective paint can appear clearly in the lidar images.

[0081]

[0095] 12A, when the lidar sensor 1210 is properly aligned with the vehicle 1220 (e.g., the lidar sensor 1210 is looking in the same direction that the vehicle 1220 is traveling), the pair of lane markings 1240 may appear to pass equally on either side of the vehicle path 1250. When the lidar sensor 1210 is misaligned with the vehicle 1220 (e.g., the lidar sensor 1210 is looking to the left relative to the longitudinal axis of the vehicle 1220), as shown in FIG. 12B, the pair of lane markings 1240 may appear to pass the vehicle 1220 asymmetrically with respect to the vehicle path 1250. For example, the vehicle path 1250 may appear to move closer to the driver's side lane marking 1240a than to the passenger's side lane marking 1240b. Thus, by analyzing lidar images of lane markings 1240 relative to vehicle path 1250, the amount of positional misalignment (e.g., yaw error) of lidar sensor 1210 relative to vehicle 1220 can be estimated.

[0082]

[0096] 13A-13C show some further examples of using lane configurations for dynamic calibration of a vehicle-mounted lidar sensor, according to some embodiments. As shown in FIG. 13A, if the positional alignment of the lidar sensor 1210 has a pitch error, the lane markings 1240 may appear to be tilted relative to the vehicle path (assuming the road is relatively level).

[0083]

[0097] As shown in FIG. 13B, if the positional alignment of the lidar sensor 1210 has a roll error, the lane marking 1240a on one side (e.g., the driver's side) of the vehicle 1220 may appear higher than the lane marking 1240b on the other side (e.g., the passenger's side) of the vehicle 1220.

[0084]

[0098] 13C, if the lidar sensor 1210 alignment has a Z error (vertically), the height of the lane markings 1240 may be offset from the vehicle path. Translation errors in two other orthogonal directions (e.g., X and Y errors) can also be detected by monitoring the movement of objects such as lane markings (relative to their expected movement) as the vehicle 1220 moves forward.

[0085]

[0099] Thus, by analyzing the spatial relationship between images of lane markings 1240 (or other road features) and the vehicle path, various rotational and translational misalignments of the lidar sensor 1210 relative to the vehicle 1220 can be detected and estimated. If the lidar sensor is mounted on the side or rear of the vehicle, a similar calibration procedure can be used, with appropriate modifications and mathematical transformations to account for the different viewing angles.

[0086]

[0100] According to some embodiments, measurements can be repeated multiple times and the results can be averaged to account for, for example, road irregularities, poorly painted lane markings, curves, potholes, etc. Outliers can be discarded. Outliers can be due, for example, to lane changing and other driving irregularities, obstruction of the view of lane markings by other vehicles, etc.

[0087]

[0101] According to some embodiments, other information can be used to improve the quality and reliability of the calibration data. For example, data from vehicle steering sensors, global navigation satellite system (e.g., GPS) data, and inertial measurement unit (IMU) data can be used to ensure the vehicle is not turning. Map data can be used to select good road sections for calibration, where the road is straight, level, and lane markings are new and properly spaced. Other road features, such as curbs, guardrails, and road signs, can also be used as inputs to the calibration algorithm.

[0088]

[0102] FIG. 14 shows a simplified flowchart illustrating a method 1400 for calibrating a lidar sensor mounted on a vehicle using road features, according to some embodiments.

[0089]

[0103] The method 1400 includes, at 1402, acquiring one or more three-dimensional images using a lidar sensor while the vehicle is traveling on a road having fixed road features, each of the one or more three-dimensional images including an image of the road features.

[0090]

[0104] The method 1400 further includes, at 1404, analyzing a spatial relationship between the images of road features in the one or more three-dimensional images and the orientation of the field of view of the lidar sensor.

[0091]

[0105] The method 1400 further includes, at 1406, determining a deviation from an expected alignment of the lidar sensor relative to the vehicle based on the spatial relationship between the image of the road features and the field of view of the lidar sensor.

[0092]

[0106] In some embodiments, the method 1400 further includes, at 1408, recalibrating the lidar sensor based on a deviation from an expected alignment of the lidar sensor with respect to the vehicle.

[0093]

[0107] In some embodiments, the method 1400 further includes determining that a deviation from an expected alignment of the lidar sensor exceeds a threshold, and issuing an alert in response to determining that the deviation from the expected alignment of the lidar sensor exceeds the threshold.

[0094]

[0108] In some embodiments, the road feature includes one or more pairs of lane markings on either side of the vehicle. Analyzing the spatial relationship can include determining a pitch angle between a pair of lane markings of the one or more pairs of lane markings and a field of view of the lidar sensor, and determining a deviation from expected alignment of the lidar sensor can include determining a pitch error of the lidar sensor based on the pitch angle. In some embodiments, the one or more pairs of lane markings can include a first lane marking on a driver's side of the vehicle and a second lane marking on a passenger's side of the vehicle, and analyzing the spatial relationship can include determining a difference in elevation between the first lane marking and the second lane marking, and determining a deviation from expected alignment of the lidar sensor can include determining a roll error of the lidar sensor based on the difference in elevation.

[0095]

[0109] In some embodiments, the one or more three-dimensional images may include multiple three-dimensional images acquired by a lidar sensor over a time interval as the vehicle travels on a road over a distance. The road may be substantially straight and level over the distance. In some embodiments, method 1400 further includes determining a path of the vehicle over the distance and comparing the path of the vehicle to paths of road features from the multiple three-dimensional images. In some embodiments, the one or more pairs of lane markings may include a first lane marking on a driver's side of the vehicle and a second lane marking on a passenger's side of the vehicle, and analyzing the spatial relationship may include determining a lateral asymmetry between the first lane marking from the vehicle's path and the second lane marking from the vehicle's path, and determining a deviation from an expected alignment of the lidar sensor may include determining a yaw error of the lidar sensor based on the lateral asymmetry. In some embodiments, analyzing the spatial relationship may include determining an elevation difference between a pair of lane markings and the path of the vehicle, and determining a deviation from an expected alignment of the lidar sensor may include determining a vertical error of the lidar sensor based on the elevation difference.

[0096]

[0110] It should be understood that the specific steps illustrated in FIG. 14 provide a particular method for calibrating a lidar sensor according to some embodiments. Other orders of steps may be performed according to alternative embodiments. For example, alternative embodiments of the present invention may perform the steps outlined above in a different order. Additionally, individual steps illustrated in FIG. 14 may include multiple sub-steps that may be performed in various orders as appropriate for the individual step. Furthermore, additional steps may be added or some steps may be removed depending on the particular application. Those skilled in the art will recognize many variations, modifications, and alternatives.

[0097]

[0111] It should also be understood that the examples and embodiments described herein are for illustrative purposes only, and that various modifications or changes therein may be suggested to those skilled in the art and are to be included within the spirit and scope of this application and the scope of the appended claims.

Claims

1. 1. A method for calibrating a lidar sensor mounted on a vehicle, the method comprising: positioning the vehicle away from a target, the target comprises a flat mirror and features surrounding the flat mirror, the optical axis of the flat mirror being substantially horizontal; the vehicle is positioned and oriented relative to the flat mirror such that the optical axis of the lidar sensor is nominally parallel to the optical axis of the flat mirror and the target is nominally centered in the field of view of the lidar sensor; acquiring a three-dimensional image of the target using the lidar sensor, the three-dimensional image of the target including an image of the features of the target and a mirror image of the vehicle formed by the plane mirror; determining deviations from an expected alignment of the lidar sensor with respect to the vehicle by analyzing the image of the feature in the three-dimensional image of the target and the mirror image of the vehicle; A method comprising:

2. recalibrating the lidar sensor relative to the vehicle based on the deviation from the expected alignment of the lidar sensor relative to the vehicle. The method of claim 1 further comprising:

3. determining that the deviation from the expected alignment of the lidar sensor exceeds a threshold; issuing an alert in response to determining that the deviation from the expected alignment of the lidar sensor exceeds the threshold; The method of claim 1 further comprising:

4. The method of claim 1 , wherein the field of view of the lidar sensor is less than 180 degrees horizontally.

5. determining the deviation from the expected alignment of the lidar sensor with respect to the vehicle; determining a position and orientation of the lidar sensor relative to the target based on the image of the feature; determining a position and orientation of the lidar sensor relative to the mirror image of the vehicle based on the mirror image of the vehicle; determining a transformation from a lidar coordinate system to a vehicle coordinate system based on (i) the position and orientation of the lidar sensor relative to the target and (ii) the position and orientation of the lidar sensor relative to the mirror image of the vehicle; The method of claim 1 , comprising:

6. 6. The method of claim 5, wherein the rider coordinate system has three translational and three rotational degrees of freedom, and the vehicle coordinate system has three translational and three rotational degrees of freedom.

7. the three translational degrees of freedom of the lidar coordinate system are along three orthogonal axes including an x-axis, a y-axis, and a z-axis; and the three rotational degrees of freedom of the lidar coordinate system include a roll rotation about the x-axis, a pitch rotation about the y-axis, and a yaw rotation about the z-axis; the three translational degrees of freedom of the vehicle coordinate system are along three orthogonal axes including an X-axis, a Y-axis, and a Z-axis, and the three rotational degrees of freedom of the vehicle coordinate system include a roll rotation about the X-axis, a pitch rotation about the Y-axis, and a yaw rotation about the Z-axis; The method of claim 6.

8. determining a yaw angle of the vehicle relative to the optical axis of the plane mirror using two or more distance sensors positioned adjacent to the vehicle; The method of claim 7 further comprising:

9. using the two or more distance sensors to determine the lateral position of the vehicle along an axis orthogonal to the optical axis and normal axis of the plane mirror. The method of claim 8 further comprising:

10. The method of claim 8 , wherein each of the two or more distance sensors comprises an ultrasonic sensor or a laser sensor.

11. using six range sensors to determine the yaw, roll, pitch, and translational position of the vehicle along three orthogonal axes in the coordinate system of the plane mirror; The method of claim 7 further comprising:

12. The method of claim 11 , wherein each of the six distance sensors comprises an ultrasonic sensor or a laser sensor.

13. determining the deviation from the expected alignment of the lidar sensor with respect to the vehicle; storing a reference matrix relating to an expected relationship between the lidar sensor and the vehicle; determining a matrix relating to a current relationship between the lidar sensor and the vehicle; determining the deviation from the expected alignment of the lidar sensor with respect to the vehicle by comparing the matrix with a reference matrix; The method of claim 1 , comprising:

14. recalibrating the lidar sensor relative to the vehicle based on the matrix relating to the current relationship between the lidar sensor and the vehicle.

14. The method of claim 13, further comprising:

15. 1. A method for calibrating a lidar sensor mounted on a vehicle, the method comprising: storing a reference three-dimensional image acquired by the LIDAR sensor while the LIDAR sensor is in expected alignment with the vehicle, the reference three-dimensional image including a first image of a fixed feature on the vehicle; acquiring a three-dimensional image using the lidar sensor, the three-dimensional image including a second image of the fixed feature; determining a deviation from the expected alignment of the lidar sensor with respect to the vehicle by comparing the second image of the fixed feature in the three-dimensional image with the first image of the fixed feature in the reference three-dimensional image; A method comprising:

16. recalibrating the lidar sensor based on the deviation from the expected alignment of the lidar sensor with respect to the vehicle.

16. The method of claim 15, further comprising:

17. determining a transformation to be applied to the second image of the fixed feature in the three-dimensional image to match the first image of the fixed feature in the reference three-dimensional image; recalibrating the lidar sensor based on the transformation; 16. The method of claim 15, further comprising:

18. determining that the deviation from the expected alignment of the lidar sensor exceeds a threshold; issuing an alert in response to determining that the deviation from the expected alignment of the lidar sensor exceeds the threshold; 16. The method of claim 15, further comprising:

19. The method of claim 15 , wherein the reference three-dimensional image is acquired by the lidar sensor after the lidar sensor is pre-calibrated at a manufacturing facility.

20. 16. The method of claim 15, wherein the deviations from the expected positional alignment of the lidar sensor include one or more of yaw deviations, roll deviations, pitch deviations, and translational deviations along three orthogonal axes.

21. The method of claim 15 , wherein the fixed feature comprises a portion of the hood of the vehicle or an object attached to the hood.

22. the lidar sensor is located behind a windshield of the vehicle; the fixed feature includes a mask attached to an area of ​​the windshield directly in front of the lidar sensor, the mask configured to block light at an operating wavelength of the lidar sensor and shaped to block a portion of the field of view of the lidar sensor; 16. The method of claim 15.

23. 23. The method of claim 22, wherein the mask has an outer boundary and an inner boundary, the inner boundary being sized such that the mask encroaches on the periphery of the field of view of the lidar sensor.

24. 1. A method for calibrating a lidar sensor mounted on a vehicle, the method comprising: acquiring one or more three-dimensional images using the lidar sensor while the vehicle is traveling on a road having fixed road features, each of the one or more three-dimensional images including an image of the road features; analyzing a spatial relationship between the image of the road feature in the one or more three-dimensional images and a field of view orientation of the lidar sensor; determining a deviation from an expected alignment of the lidar sensor relative to the vehicle based on the spatial relationship between the image of the road features and the field of view of the lidar sensor; A method comprising:

25. recalibrating the lidar sensor based on the deviation from the expected alignment of the lidar sensor with respect to the vehicle.

25. The method of claim 24, further comprising:

26. determining that the deviation from the expected alignment of the lidar sensor exceeds a threshold; issuing an alert in response to determining that the deviation from the expected alignment of the lidar sensor exceeds the threshold; 25. The method of claim 24, further comprising:

27. The method of claim 24 , wherein the road features include one or more pairs of lane markings on either side of the vehicle.

28. analyzing the spatial relationship includes determining a pitch angle between a pair of lane markings of the one or more pairs of lane markings and the field of view of the lidar sensor; determining the deviation from the expected alignment of the lidar sensor includes determining a pitch error of the lidar sensor based on the pitch angle.

28. The method of claim 27.

29. the one or more pairs of lane markings include a first lane marking on a driver's side of the vehicle and a second lane marking on a passenger's side of the vehicle; analyzing the spatial relationship includes determining an elevation difference between the first lane marking and the second lane marking; determining the deviation from the expected alignment of the lidar sensor includes determining a roll error of the lidar sensor based on the elevation difference.

28. The method of claim 27.

30. 28. The method of claim 27, wherein the one or more three-dimensional images comprise a plurality of three-dimensional images acquired by the lidar sensor over a time interval as the vehicle travels on the road over a distance.

31. 31. The method of claim 30, wherein the road is substantially straight and level over the distance.

32. determining a route for the vehicle over the distance; comparing the path of the vehicle with paths of the road features from the plurality of three-dimensional images; 31. The method of claim 30, further comprising:

33. the one or more pairs of lane markings include a first lane marking on a driver's side of the vehicle and a second lane marking on a passenger's side of the vehicle; analyzing the spatial relationship includes determining a lateral asymmetry between the first lane marking from the path of the vehicle and the second lane marking from the path of the vehicle; determining the deviation from the expected alignment of the lidar sensor includes determining a yaw error of the lidar sensor based on the lateral asymmetry; 33. The method of claim 32.

34. analyzing the spatial relationship includes determining an elevation difference between a pair of lane markings and the path of the vehicle; determining the deviation from the expected alignment of the lidar sensor includes determining a vertical error of the lidar sensor based on the elevation difference; 33. The method of claim 32.

Images