Calibration verification component, system and method

Abstract

The present application relates to a calibration verification component, system and method, which belongs to the field of sensor technology and can be applied to smart cars in the field of autonomous driving. The calibration verification component includes a first detection object (1), a second detection object (2) and a bracket (3); the first detection object (1) and the second detection object (2) are objects for detection by different types of sensors (7); the first detection object (1) and the second detection object (2) are respectively mounted on the bracket (3). In the embodiment disclosed in the present application, the first detection object (1) or the second detection object (2) can be accurately identified by at least one sensor (7), avoiding the situation in the prior art where different types of sensors (7) have large differences in recognition accuracy for the same target reference object, thereby improving the accuracy of calibration verification.

Description

Technical Field

[0001] The present application relates to the field of sensor technology, and in particular to a calibration verification component, system and method. Background Art

[0002] With the development of social economy, cars have become a necessity in people's daily life. The emergence of autonomous driving function has reduced the traffic accident rate and relieved traffic pressure. Vehicles with autonomous driving function are generally equipped with various sensors, such as laser radar, camera and millimeter wave radar. During the driving process of the vehicle, the sensor collects environmental information and sends it to the vehicle terminal, which controls the vehicle based on the environmental information.

[0003] These sensors installed on vehicles with autonomous driving functions usually need to be calibrated before use. Calibration refers to transferring the measurement data in a sensor's own coordinate system to another sensor's coordinate system through a state transfer function. The accuracy of calibration affects the control accuracy of autonomous driving. At present, the calibration verification method is mainly to select target reference objects in actual scenes, such as vehicles, signs, trees, obstacle balls, etc., and verify the accuracy of the calibration results through the position deviation of the target reference objects captured by different types of sensors.

[0004] The current verification method relies on the recognition accuracy of different types of sensors for target reference objects in actual scenes. Only when the target reference objects are accurately recognized can the verification accuracy of the calibration results be guaranteed. However, the recognition accuracy of different types of sensors for the same target reference object usually varies greatly, which will seriously affect the verification accuracy of the calibration results. For example, when the target reference object is an obstacle ball with a smooth surface, the camera can accurately recognize it, but the laser radar and millimeter wave radar have low recognition capabilities, so the calibration verification results are not accurate. Summary of the invention

[0005] The embodiments of the present application provide a calibration verification component, system and method, which can solve the technical problems existing in the related art. The technical solutions of the calibration verification component, system and method are as follows:

[0006] In the first aspect, an embodiment of the present application provides a calibration verification component, which consists of a first detection object 1, a second detection object 2 and a bracket 3, the first detection object 1 and the second detection object 2 are respectively mounted on the bracket 3, and the first detection object 1 and the second detection object 2 are respectively objects for detection by different types of sensors 7.

[0007] In a possible implementation, the first detection object 1 is a diffuse reflector, and the second detection object 2 is a corner reflector.

[0008] In a possible implementation, the first detection object 1 of the calibration verification component is spherical in shape.

[0009] In a possible implementation, the calibration verification component further includes a pan-tilt platform 4, which is composed of a fixed component 41 and a movable component 42. The fixed component 41 is fixedly connected to the bracket 3, and the first detection object 1 and the second detection object 2 are mounted on the movable component 42.

[0010] In a possible implementation, the calibration verification assembly further includes a first mounting rod 5 and a second mounting rod 6. One end of the first mounting rod 5 is fixedly connected to the movable member 41, and the other end is mounted with the first detection object 1; one end of the second mounting rod 6 is fixedly connected to the movable member 41, and the other end is mounted with the second detection object 2. Furthermore, the detection point of the first detection object 1, the detection point of the second detection object 2, the axis of the first mounting rod 5, and the axis of the second mounting rod 6 are on the same straight line.

[0011] In a possible implementation, the gimbal 4 of the calibration verification component further includes a balancer 43, which is fixedly connected to the movable part 41 of the gimbal 4. The detection plane of the balancer 43 is perpendicular to the straight line where the detection point of the first detection object 1, the detection point of the second detection object 2, the axis of the first mounting rod 5, and the axis of the second mounting rod 6 are located.

[0012] In a possible implementation, a slide groove 61 and a slide seat 62 are provided on the second mounting rod 6 of the calibration verification assembly, and the slide seat 62 is installed in the slide groove 61. The second detection object 2 is installed on the slide seat 62, and the sliding track of the detection point of the second detection object 2 is on the axis of the second mounting rod 6.

[0013] In a possible implementation, the outer wall of the second mounting rod 6 of the calibration verification assembly has a plurality of axially arranged scale lines and scale values ​​corresponding to each scale line, wherein the scale value corresponding to the scale line is the distance between the point corresponding to the scale line on the axis of the second mounting rod 6 and the detection point of the first detection object 1.

[0014] In a possible implementation, the bracket 3 of the calibration verification component may be a tripod.

[0015] In the second aspect, the embodiment of the present application further provides a calibration verification system, which includes a target device 10, a calibration verification device 20, and a calibration verification component 30 as described in the first aspect and its possible implementation. Among them, multiple types of sensors 7 are installed on the target device 10, the calibration verification component 30 is located in the detection area of ​​the multiple types of sensors 7, and the multiple types of sensors 7 and the calibration verification device 20 establish a communication connection.

[0016] In a third aspect, an embodiment of the present application further provides a calibration verification method, which is applied to the calibration verification system as described in the second aspect.

[0017] In a possible method, the multiple types of sensors 7 in the calibration verification system include a first sensor 71 and a second sensor 72, wherein the first sensor 71 and the second sensor 72 are of different types. The calibration verification method is specifically as follows:

[0018] The calibration verification device 20 obtains the detection data of the first sensor 71 and the detection data of the second sensor 72. The calibration verification device 20 determines the first position information of the detection point of the first detection object 1 in the coordinate system of the first sensor 71 based on the detection data of the first sensor 71. The calibration verification device 20 determines the second position information of the detection point of the second detection object 2 in the coordinate system of the second sensor 72 based on the detection data of the second sensor 72.

[0019] Obtain the relative position information of the first detection object 1 and the second detection object 2. Determine the error value of the coordinate conversion function from the second sensor 72 to the first sensor 71 based on the first position information, the second position information and the relative position information.

[0020] In a possible method, the multiple types of sensors 7 in the calibration verification method further include a third sensor 73. The third sensor 73 is of a different type from the first sensor 71 and the second sensor 72. The calibration verification method is specifically as follows:

[0021] The calibration verification device 20 obtains the detection data of the third sensor 73. The calibration verification device 20 determines the third position information of the detection point of the first detection object 1 in the coordinate system of the third sensor 73 based on the detection data of the third sensor 73.

[0022] Based on the first position information and the third position information, an error value of the coordinate conversion function of the third sensor 73 to the first sensor 71 is determined.

[0023] In the above calibration verification method, the first detection object 1 is a spherical diffuse reflector, the second detection object 2 is a corner reflector, the first sensor 71 is a camera, the second sensor 72 is a millimeter wave radar, and the third sensor 73 is a laser radar.

[0024] The technical solution provided by the embodiments of the present application includes at least the following beneficial effects:

[0025] The calibration verification component provided in the embodiment of the present application includes a first detection object 1 and a second detection object 2 for detection by different types of sensors 7. For verification of calibration from the first type of sensor 7 to the second type of sensor 7, a detection object that can be accurately detected by the first type of sensor 7 and a detection object that can be accurately detected by the second type of sensor 7 can be used as the first detection object 1 and the second detection object 2. In this way, both sensors can have a high detection accuracy for the corresponding detection objects, avoiding the situation where different types of sensors 7 have large differences in recognition accuracy for the same target reference object, thereby ensuring the accuracy of calibration verification.

[0026] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0028] Figure 1 is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0029] Figure 2 It is a structural schematic diagram of a single-angle corner reflector shown in an embodiment of the present application;

[0030] Figure 3 is a schematic structural diagram of an octagonal corner reflector shown in an embodiment of the present application;

[0031] Figure 4 is a schematic structural diagram of a spherical diffuse reflector shown in an embodiment of the present application;

[0032] Figure 5 is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0033] Figure 6 is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0034] Figure 7 is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0035] Figure 8 is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0036] Fig. 9is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0037] Fig.10 is a structural schematic diagram of a calibration verification component shown in an embodiment of the present application;

[0038] Fig.11 is a structural schematic diagram of a calibration verification system shown in an embodiment of the present application;

[0039] Fig.12 It is a flow chart of a calibration verification method shown in an embodiment of the present application.

[0040] Legend

[0041] 1. First detection object; 2. Second detection object; 3. Bracket; 4. Pan / tilt; 41. Fixed part; 42. Movable part; 43. Balancer; 5. First mounting rod; 6. Second mounting rod; 61. Slide groove; 62. Slide seat; 7. Sensor; 71. First sensor; 72. Second sensor; 73. Third sensor; 10. Target device; 20. Calibration verification device; 30. Calibration verification assembly. DETAILED DESCRIPTION

[0042] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0043] The embodiment of the present application provides a calibration verification component, which is a detection object combination for calibration verification and may include multiple detection objects. Based on the calibration verification component, the embodiment of the present application also provides a calibration verification system for performing calibration verification processing. Based on the calibration verification system, the embodiment of the present application also provides a calibration verification method. The calibration verification component, the calibration verification system, and the calibration verification method will be introduced one by one below.

[0044] Calibration Verification Kit

[0045] The present application embodiment provides a calibration verification component, such as Figure 1 As shown, the calibration verification assembly includes a first detection object 1, a second detection object 2 and a bracket 3. The first detection object 1 and the second detection object 2 are objects for different types of sensors 7 to detect. The first detection object 1 and the second detection object 2 are mounted on the bracket 3, respectively.

[0046] The first detection object 1 and the bracket 3 can be directly connected by means of threaded connection, welding, clamping, etc., or the first detection object 1 and the bracket 3 can be indirectly connected by means of an intermediate connection member, which can be a connecting rod, a rope, a chain, etc. The second detection object 2 and the bracket 3 can be directly connected by means of threaded connection, welding, clamping, etc., or the second detection object 2 and the bracket 3 can be indirectly connected by means of an intermediate connection member, which can be a connecting rod, a rope, a chain, etc.

[0047] The connection method used by the first detection object 1 and the connection method used by the second detection object 2 may be the same or different. For example, the first detection object 1 is connected to the bracket 3 via a connecting rod, and the second detection object 2 is directly welded to the bracket 3. For another example, the first detection object 1 and the second detection object 2 are both fixedly connected to the bracket 3 via a connecting rod. For another example, the first detection object 1 and the second detection object 2 are both fixedly connected to the bracket 3 via a threaded connection.

[0048] Among them, the second detection object 2 is a corner reflector, which is used to detect the millimeter wave radar. The electromagnetic wave signal emitted by the millimeter wave radar can generate a strong echo signal after being reflected by the corner reflector. The echo signal enhances the recognition ability of the millimeter wave radar for the corner reflector. Corner reflectors are divided into single-angle, quadrangular, hexagonal, octagonal, polygonal and other types according to the number of reflection angles. The corner reflector in a calibration verification component provided in an embodiment of the present application can be selected as follows: Figure 2 The single-angle corner reflector shown in the figure can also be selected as Figure 3 The octagonal corner reflector shown can also be selected from other types of corner reflectors mentioned above. The corner reflector in the calibration verification assembly is made of metal material, and any one of aluminum, iron, stainless steel and other metal materials can be selected.

[0049] The first detection object 1 is a diffuse reflector. The diffuse reflector can be a white spherical object with a diffuse reflection surface, such as Figure 4 As shown. In order to improve the diffuse reflection ability of the white spherical object, white diffuse reflection paint can be sprayed on the surface. Among them, the white diffuse reflection paint can be a white paint doped with glass beads or a paint added with fine metal powder. The spherical object is made of non-metallic materials, such as ceramics, rubber, plastic, wood, etc. The diameter of the spherical object has a variety of optional values, and the diameter can be between 500mm-5000mm.

[0050] In the embodiment of the present application, the detection object can be mounted on the platform, so as to achieve the position adjustment of the detection object. The platform 4 is composed of a fixed part 41 and a movable part 42. The fixed part 41 of the platform 4 is fixedly connected to the bracket 3. The first detection object 1 and the second detection object 2 are respectively mounted at different positions of the movable part 42 of the platform 4, and the movable part 42 is movably connected to the fixed part 41. The spatial position of the first detection object 1 and the second detection object 2 can be achieved by the rotation of the movable part 42.

[0051] A possible structure of the pan / tilt 4 is as follows: Figure 5 As shown, the fixing part 41 of the pan-tilt head 4 is an inverted hemispherical shell with an open top, and the hemispherical outer surface of the fixing part 41 is directly fixedly connected to the bracket 3, and the connection method can be selected from threaded connection, welding, clamping, etc. The movable part 42 of the pan-tilt head 4 is a sphere, and the movable part 42 is connected to the hemispherical inner surface of the fixing part 41 by a spherical pair. Among them, the first detection object 1 and the second detection object 2 are respectively installed at different positions of the movable part 42. When the movable part 42 rotates in the fixing part 41, the positions of the first detection object 1 and the second detection object 2 change, and they maintain a certain relative position relationship during the position change process.

[0052] In a possible structure of the pan / tilt head 4, the fixed part 41 can be a plate-like object with a hole in the middle, and the movable part 42 can be a plate-like object with three or more screw feet installed on the same surface, and the other end of the screw feet is installed on the upper surface of the fixed part 41. The first detection object 1 is installed on the upper surface of the movable part 42, and the second detection object 2 is installed on the lower surface of the movable part 42. The inclination angle between the movable part 42 and the horizontal plane is adjusted by adjusting the height of each screw foot, thereby adjusting the positional relationship between the first detection object 1 and the second detection object 2.

[0053] The lower surface of the fixing member 41 and the bracket 3 may be directly connected by threaded connection, welding, clamping, etc., or indirectly connected by an intermediate connecting member, which may be a connecting rod, a stud bolt, etc. The movable member 42 and the spiral support leg, and the fixing member 41 and the threaded support leg may be directly connected by threaded connection, welding, clamping, etc.

[0054] Among them, the thread legs are distributed in a regular polygonal shape. For example, when the number of thread legs is 3, the distribution is in a regular triangle, when the number of thread legs is 4, the distribution is in a square, and when the number of thread legs is 6, the distribution is in a regular hexagon.

[0055] A possible structure of the gimbal 4 may also be a three-axis gimbal.

[0056] The first detection object 1 and the movable part 42 can be directly connected by means of threaded connection, welding, clamping, etc., or the first detection object 1 and the movable part 42 can be indirectly connected by means of an intermediate connection, which can be a connecting rod, a rope, a chain, etc. The second detection object 2 and the movable part 42 can be directly connected by means of threaded connection, welding, clamping, etc., or the second detection object 2 and the movable part 42 can be indirectly connected by means of an intermediate connection, which can be a connecting rod, a rope, a chain, etc.

[0057] The connection method used by the first detection object 1 and the connection method used by the second detection object 2 may be the same or different. For example, the first detection object 1 is connected to the movable member 42 via a connecting rod, and the second detection object 2 is directly welded to the bracket 3. For another example, the first detection object 1 and the second detection object 2 are both fixedly connected to the movable member 42 via a connecting rod. For another example, the first detection object 1 and the second detection object 2 are both fixedly connected to the movable member 42 via a threaded connection.

[0058] By adjusting the movable member 42 of the pan-tilt platform 4, the positions of the first detection object 1 and the second detection object 2 can be adjusted. Because there are vertical coordinate axes in the coordinate systems of different sensors, it is possible to consider adjusting the first detection object 1 and the second detection object 2 to the same plumb line through the pan-tilt platform 4. In this way, during calibration and verification, only the height difference between the first detection object 1 and the second detection object 2 needs to be measured, the measurement operation is relatively simple, and the error caused by the measurement can be reduced. In addition, the computational complexity of the calibration and verification process can be simplified to a certain extent.

[0059] On the basis of the pan-tilt head 4, two co-linearly arranged connecting rods can be used to connect the first detection object 1 and the second detection object 2, so as to more conveniently adjust the first detection object 1 and the second detection object 2 to the same plumb line. Figure 6 As shown, the calibration verification assembly includes a first mounting rod 5 and a second mounting rod 6. One end of the first mounting rod 5 is fixedly connected to the movable member 42, one end of the second mounting rod 6 is fixedly connected to the movable member 42, the first detection object 1 is mounted on the first mounting rod 5, and the second detection object 2 is mounted on the second mounting rod 6. The detection point of the first detection object 1, the detection point of the second detection object 2, the axis of the first mounting rod 5, and the axis of the second mounting rod 6 are on the same straight line.

[0060] The detection point of the detection object can be a specific point in the detection object pre-specified by the technician, for example, the center of a spherical detection object, the vertex of a single-angle corner reflector (the intersection of three reflective plates), or the geometric center of an octagonal corner reflector.

[0061] Among them, the first mounting rod 5 and the movable part 42 can be directly connected by means of threaded connection, welding, clamping, etc. The first mounting rod 5 and the first detection object 1 can also be directly connected by means of threaded connection, welding, clamping, etc. The second mounting rod 6 and the movable part 42 can be directly connected by means of threaded connection, welding, clamping, etc. The second detection object 2 and the second mounting rod 6 can be directly connected by means of threaded connection, welding, clamping, etc., or they can be indirectly connected by an intermediate connecting member, which can be a slider, etc.

[0062] The first mounting rod 5 and the second mounting rod 6 may be made of non-metallic materials, such as wood, plastic, ceramic, etc., to avoid affecting the detection accuracy of the millimeter wave radar. The first mounting rod 5 and the second mounting rod 6 may be made of the same material or different materials. For example, the first mounting rod 5 and the second mounting rod 6 are both made of plastic, or the first mounting rod 5 is made of wood and the second mounting rod 6 is made of plastic.

[0063] A balancer 43 may be provided on the pan / tilt platform 4 to detect whether the connecting line of the detection point of the first detection object 1 and the detection point of the second detection object 2 are on the same plumb line. Figure 7 As shown, a balancer 43 is fixedly connected to the movable part 42 of the pan / tilt platform 4. The connecting line of the detection point of the first detection object 1 and the detection point of the second detection object 2 is perpendicular to the detection plane of the balancer 43.

[0064] The balance instrument 43 can be a circular level or a tube level with two perpendicular axes. The circular level is a circular container with a circular dividing line engraved in the center of the top cover, and there is a bubble inside the container. When the center of the bubble coincides with the center of the circular dividing line, the detection plane of the circular level is in a horizontal state. The tube level is a tubular container with a top cover engraved with multiple parallel dividing lines spaced 2 mm apart, and the dividing lines are perpendicular to the axis of the tubular container. There is a bubble inside the tubular container. When the midpoints of the bubbles of the two perpendicular tube levels coincide with the midpoints of their respective corresponding dividing lines, the detection plane of the tube level is in a horizontal state. When using the above-mentioned calibration verification component, the operator can adjust the movable part 42 of the pan / tilt 4 so that the detection plane of the balance instrument 43 is in a horizontal state, thereby ensuring that the detection point of the first detection object 1 and the detection point of the second detection object 2 are on the same plumb line.

[0065] In the embodiment provided in the present application, an adjustment mechanism may be provided to adjust the relative position between the second detection object 2 and the first detection object 1. The corresponding structure is as follows Figure 8As shown, the second mounting rod 6 includes a slide groove 61 and a slide seat 62. The slide seat 62 is installed in the slide groove 61, and the second detection object 2 is connected to the slide seat 62. The slide seat 62 drives the second detection object 2 to slide up and down along the axis direction of the second mounting rod 2 in the slide groove 61, and the sliding track of the detection point of the second detection object 2 is on the axis of the second mounting rod 6. The second detection object 2 can be selected as follows Figure 3 The single-angle corner reflector shown can be directly connected to the sliding seat 62 by means of threaded connection, clamping, welding, etc.

[0066] In the embodiment provided in the present application, the outer wall of the second mounting rod 6 has a plurality of axially arranged scale lines and scale values ​​corresponding to each scale line, so as to accurately record the distance between the detection point of the first detection object 1 and the detection point of the second detection object 2. The corresponding structure is as follows Fig. 9 As shown in the figure, the scale value corresponding to each scale line is the distance between the point corresponding to the scale line on the axis of the second mounting rod 6 and the detection point of the first detection object 1. When the above-mentioned single-angle corner reflector slides in the slide groove 61 along with the slide seat 62, the scale value of the scale line corresponding to the bottom surface of the slide seat 62 is regarded as the distance between the detection point of the single-angle corner reflector and the detection point of the first detection object 1.

[0067] In the embodiment provided in the present application, the bracket 3 can be a tripod with a circular cross-section of the legs, and the corresponding structure is as follows: Fig.10 As shown. The tripod legs can be freely extended and retracted along the axis of the legs to adjust the overall height of the calibration verification assembly. The legs can be extended and retracted in 3-5 sections, and the legs can be fixed by any means such as buckle type, screw fastening type, bolt fastening type, etc. The bracket 3 of the calibration verification assembly provided in the present application can be selected with a tripod with any number of legs, and the cross-section of the legs can also be selected in any shape. The structural form of the bracket with any number of legs and any cross-sectional shape is similar to the above-mentioned tripod, so it will not be repeated.

[0068] The above calibration verification components are used as follows:

[0069] First, determine the common detection area of ​​different types of sensors that need to be calibrated and verified, and fix the calibration verification component in the common detection area. Secondly, adjust the height of each leg of the bracket 3 of the calibration verification component to ensure that the detection object of the calibration verification component is within the sensor detection range. Then, adjust the angle between the legs of the bracket 3 of the calibration verification component to ensure that the level bubble of the balancer 43 is at or close to the center position. Finally, adjust the movable part 42 so that the detection plane of the balancer 43 is in the horizontal direction, thereby ensuring that the detection point of the first detection object 1 and the detection point of the second detection object 2 are on the same plumb line. At this point, the sensor can detect the data and perform calibration verification.

[0070] Calibration Verification System

[0071] The present application also provides a calibration verification system. Fig.11 As shown, the calibration verification system includes a target device 10, a calibration verification device 20 and a calibration verification component 30. The target device 10 is equipped with various types of sensors 7, the calibration verification component 30 is located within the detection range of the various types of sensors 7 on the target device 10, and the target device 10 and the calibration verification device 20 are in communication connection.

[0072] The target device 10 may be any product, such as a car, a robot, or a mechanical arm, equipped with various types of sensors 7. The various types of sensors 7 include but are not limited to laser radar, camera, millimeter wave radar, and may include two or three of them.

[0073] The calibration verification device 20 can be a part of the product or an external device, for example, a vehicle-mounted terminal of a car, a terminal of a robot, or an external computer as a terminal. Various types of sensors 7 and the calibration verification device 20 can directly establish communication through WiFi connection, Bluetooth connection, data cable connection, etc. Indirect communication can also be established between the sensor 7 and the calibration verification device 20. For example, the sensor 7 on the car transmits the detection data to the vehicle-mounted terminal, and the vehicle-mounted terminal transmits the detection data to the external computer. Real-time communication can be established between the sensor and the vehicle-mounted terminal, and between the vehicle-mounted terminal and the external computer through any method such as WiFi connection, Bluetooth connection, data cable connection, etc.

[0074] Calibration Verification Method

[0075] The present application also provides a calibration verification method, the process of which is as follows: Fig.12 As shown. The calibration verification method is applied to the above calibration verification system to verify the calibration accuracy between the above multiple types of sensors 7. The multiple types of sensors 7 include a first sensor 71, a second sensor 72, and a third sensor 73, and the first sensor 71, the second sensor 72, and the third sensor 73 are of different types. For example, a laser radar, a camera, and a millimeter wave radar.

[0076] First, verify the calibration of the second sensor 72 to the first sensor 71. The first sensor 71 and the second sensor 72 respectively detect different detection objects in the calibration verification component.

[0077] In step 101 , the calibration verification device 20 obtains detection data of the first sensor 71 and detection data of the second sensor 72 .

[0078] Step 102 : The calibration verification device 20 determines the first position information of the detection point of the first detection object 1 in the coordinate system of the first sensor 71 based on the detection data of the first sensor 71 , which is recorded as (a1, b1, c1).

[0079] Step 103 : The calibration verification device 20 determines the second position information of the detection point of the second detection object 2 in the coordinate system of the second sensor 72 based on the detection data of the second sensor 72 , which is recorded as (a2, b2, c2).

[0080] Step 104, obtaining the relative position information of the detection point of the first detection object 1 and the detection point of the second detection object 2 in the coordinate system of the first sensor 71, and recording a x 、b x 、c x The relative position information indicates the lengths of the projections of the line connecting the detection point of the first detection object 1 and the detection point of the second detection object 2 along the x-, y-, and z-axis directions, respectively.

[0081] Step 105, based on the first position information, the second position information and the relative position information, determine the error value of the coordinate conversion function from the second sensor 72 to the first sensor 71. The specific method for calculating the error value is as follows:

[0082] Method 1:

[0083] Based on the coordinate conversion function from the second sensor 72 to the first sensor 71 , fourth position information of the second position information in the coordinate system of the first sensor 71 is obtained, which is recorded as (a4, b4, c4).

[0084] Subtract the corresponding coordinate components in the first position information and the fourth position information and take the absolute value, which is recorded as Δa1, Δb1, and Δc1. Calculate a respectively x With Δa1, b x With Δb1,c x The difference with Δc1 represents the error value of the coordinate conversion function from the second sensor 72 to the first sensor 71 along the x-axis, y-axis and z-axis directions respectively.

[0085] Alternatively, the sum of the squares of the absolute values ​​of the coordinate components between the first position information and the fourth position information is calculated, and then the square difference of the sum of the squares is calculated, which is recorded as d1. The sum of the squares of the relative position information is calculated, and then the square difference of the sum of the squares is calculated, which is recorded as d. The difference between d and d1 is calculated, and the difference represents the overall error value of the coordinate conversion function from the second sensor 72 to the first sensor 71.

[0086] Method 2:

[0087] Based on the coordinate conversion function from the second sensor 72 to the first sensor 71 , fourth position information of the second position information in the coordinate system of the first sensor 71 is obtained, which is recorded as (a4, b4, c4).

[0088] According to the relative position information of the detection point of the first detection object 1 and the detection point of the second detection object 2, the fourth position information is offset in the coordinate system of the first sensor 71 to obtain fifth position information, which is recorded as (a5, b5, c5).

[0089] Subtract the corresponding coordinate components in the first position information and the fifth position information and take the absolute value, which is recorded as Δa2, Δb2, and Δc2. The absolute value of each coordinate component represents the error value of the coordinate conversion function from the second sensor 72 to the first sensor 71 along the x, y, and z axes respectively.

[0090] Alternatively, the square sum of the absolute values ​​of the coordinate components between the first position information and the fifth position information is calculated, and then the square root of the square sum is calculated, which is recorded as d2. The square difference represents the overall error value of the coordinate conversion function from the second sensor 72 to the first sensor 71.

[0091] Method 3:

[0092] According to the relative position information of the detection point of the first detection object 1 and the detection point of the second detection object 2, the second position information is offset in the coordinate system of the second sensor 72 to obtain sixth position information, which is recorded as (a6, b6, c6).

[0093] Based on the coordinate conversion function from the second sensor 72 to the first sensor 71 , the seventh position information of the sixth position information in the coordinate system of the first sensor 71 is obtained, which is recorded as (a7, b7, c7).

[0094] Subtract the corresponding coordinate components in the first position information and the seventh position information and take the absolute value, which is recorded as Δa3, Δb3, and Δc3. The absolute value of each coordinate component represents the error value of the coordinate conversion function from the second sensor 72 to the first sensor 71 along the x, y, and z axes respectively.

[0095] Alternatively, the square sum of the absolute values ​​of the coordinate components between the first position information and the seventh position information is calculated, and then the square root of the square sum is calculated and recorded as d3. The square difference represents the error value of the coordinate conversion function from the second sensor 72 to the first sensor 71.

[0096] As described above, in the calibration verification of the second sensor 72 to the first sensor 71, the first detection object 1 is a spherical object, and the second detection object 2 is a corner reflector. The first sensor 71 is a camera or a laser radar, and the second sensor 72 is a millimeter wave radar.

[0097] The calibration error from the first sensor 71 to the second sensor 72 is the same as the calibration error from the second sensor 72 to the first sensor 71, so it will not be repeated.

[0098] Second, verify the calibration of the third sensor 73 to the first sensor 71. The first sensor 71 and the third sensor 73 respectively detect the same detection object in the calibration verification component.

[0099] Step 106 : The calibration verification device 20 obtains detection data of the third sensor 73 .

[0100] Step 107 : The calibration verification device 20 determines the third position information of the detection point of the first detection object 1 in the coordinate system of the third sensor 73 based on the detection data of the third sensor 73 , which is recorded as (a3, b3, c3).

[0101] Step 108 , based on the coordinate conversion function from the third sensor 73 to the first sensor 71 , determine the eighth position information of the third position information of the detection point of the first detection object 1 in the coordinate system of the first sensor 71 , recorded as (a8, b8, c8).

[0102] Subtract the corresponding coordinate components in the first position information and the eighth position information and take the absolute value, which is recorded as Δa4, Δb4, and Δc4. The absolute value of each coordinate component represents the error value of the coordinate conversion function from the third sensor 73 to the first sensor 71 along the x, y, and z axes respectively.

[0103] Alternatively, the square sum of the absolute values ​​of the coordinate components between the first position information and the eighth position information is calculated, and then the square difference of the square sum is calculated, which is recorded as d4. The square difference represents the overall error value of the coordinate conversion function from the third sensor 73 to the first sensor 71.

[0104] In the calibration verification of the third sensor 73 to the first sensor 71, the first detection object 1 is used as a spherical object. The first sensor 71 is a camera and the third sensor 73 is a laser radar; or the first sensor 71 is a laser radar and the third sensor 73 is a camera.

[0105] The calibration error from the first sensor 71 to the third sensor 73 is the same as the calibration error from the third sensor 73 to the first sensor 71, so it will not be repeated.

[0106] The following uses the verification of the calibration accuracy of LiDAR, camera, and millimeter-wave radar as an example to explain the use of the calibration verification component and the calibration verification method:

[0107] The first sensor 71 is a laser radar, the second sensor 72 is a millimeter wave radar, and the third sensor 73 is a camera. The laser radar and the camera detect the first detection object 1, which is a white sphere with diffuse reflection, hereinafter referred to as the sphere. The millimeter wave radar detects the second detection object 2, which is a single-angle corner reflector, hereinafter referred to as the corner reflector.

[0108] First, determine the common detection area of ​​the laser radar, camera, and millimeter-wave radar, and fix the calibration verification component in the common detection area. Secondly, adjust the height of each leg of the bracket 3 of the calibration verification component to ensure that the detection object of the calibration verification component is within the detection range of the sensor. Then, adjust the angle between the legs of the bracket 3 of the calibration verification component to ensure that the level bubble of the balancer 43 is at or close to the center position. Finally, adjust the movable part 42 to ensure that the level bubble of the balancer 43 is at the center position. When the level bubble of the balancer 43 is at the center position, it means that the detection plane of the balancer 43 is in the horizontal direction, which means that the detection point of the first detection object 1 and the straight line where the detection point of the second detection object 2 are located are in the vertical direction. At this point, calibration verification can be performed.

[0109] First, verify the calibration of millimeter wave radar to lidar.

[0110] A set of hemispherically distributed spatial coordinate points about the sphere detected by the lidar is obtained, and the robust ball fitting algorithm is used to fit the set of hemispherically distributed spatial coordinate points into a set of spherically distributed spatial coordinate points, thereby obtaining the coordinates of each point on the spherical surface of the entire sphere, and then based on the coordinates of each point on the spherical surface, the coordinates of the center of the sphere in the lidar coordinate system are calculated (algorithm models such as machine learning models can be used), which is the first position information.

[0111] A cluster of millimeter-wave radar target points about the corner reflector detected by the millimeter-wave radar is obtained, and the coordinates of the detection points of the corner reflector in the millimeter-wave radar coordinate system are filtered out according to the placement position information of the calibration verification component, that is, the second position information.

[0112] Based on the coordinate conversion function from millimeter wave radar to laser radar, the second position information of the detection point of the corner reflector in the millimeter wave radar coordinate system is converted to the laser radar coordinate system to obtain the coordinates of the detection point of the corner reflector in the laser radar coordinate system, that is, the third position information.

[0113] The Euclidean distance between the first position information and the third position information is calculated, and the Euclidean distance is compared with the height value from the center of the sphere to the corner reflector detection point to obtain the accuracy of the millimeter wave radar to the laser radar calibration.

[0114] Second, verify the calibration of millimeter wave radar to camera.

[0115] A picture of a sphere detected by a camera is obtained, and the center position of the sphere in the picture is detected using any sphere detection algorithm to obtain the center coordinates of the sphere in the camera coordinate system, i.e., the fourth position information. The sphere detection algorithm includes but is not limited to Hough transform, template matching, and deep learning.

[0116] Based on the second position information of the detection point of the corner reflector in the millimeter wave radar coordinate system, an offset is added upward to obtain the fifth position information corresponding to the second position information, wherein the offset added upward is numerically equal to the height value from the center of the sphere to the detection point of the corner reflector obtained by actual measurement.

[0117] Based on the coordinate conversion function from millimeter wave radar to camera, the fifth position information of the detection point of the corner reflector in the millimeter wave radar coordinate system is converted to the camera coordinate system, and the coordinates of the detection point of the corner reflector in the camera coordinate system after the offset are obtained, that is, the sixth position information.

[0118] Based on the fourth position information and the sixth position information, a reprojection error is determined, where the reprojection error represents an error value from millimeter wave radar to camera calibration.

[0119] Third, verify the camera to lidar calibration.

[0120] Based on the coordinate conversion function from camera to lidar, the fourth position information of the sphere in the camera coordinate system is converted to the lidar coordinate system to obtain the seventh position information.

[0121] Based on the first position information and the seventh position information, a reprojection error is determined, where the reprojection error represents an error value of camera-to-lidar calibration.

[0122] In the above embodiments provided in the present application, the calibration verification between the laser radar, the camera, and the millimeter-wave radar is performed based on the same calibration verification component. When multiple calibration verification components are placed, target matching is performed before calculating the error. Since the placement position is fixed and known, the nearest neighbor detection based on Euclidean distance is used to match the target. For example, when detecting the calibration accuracy of the laser radar to the camera, after projecting multiple laser sphere centers onto the picture, for each projection point, find the nearest picture center point as a match. The accuracy of the above calibration verification can be expressed by the error value obtained by a single calculation, or by the average of multiple error values ​​obtained by multiple calculations, or by any value of multiple error values ​​obtained by multiple calculations.

[0123] The above description is only an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A calibration verification component, characterized in that: The calibration verification assembly comprises a first detection object (1), a second detection object (2) and a bracket (3), wherein: The first detection object (1) and the second detection object (2) are objects for detection by different types of sensors (7) respectively; The first detection object (1) and the second detection object (2) are respectively mounted on the bracket (3); wherein the first detection object (1) is a spherical diffuse reflector, the second detection object (2) is a corner reflector, the first sensor (71) is a camera, the second sensor (72) is a millimeter wave radar, and the third sensor (73) is a laser radar; the first detection object and the second detection object are used to verify the calibration of the second sensor to the third sensor, the calibration of the second sensor to the first sensor, and the calibration of the first sensor to the third sensor, and the first detection object and the second detection object are used to verify the calibration of the second sensor to the third sensor, including: calculating the Euclidean distance between the first position information and the third position information, comparing the Euclidean distance with the height value from the center of the sphere to the detection point of the corner reflector, and obtaining the accuracy of the calibration from the millimeter wave radar to the laser radar, wherein the first position information is obtained by obtaining the calibration value of the laser radar. A group of hemispherically distributed spatial coordinate points about the spherical diffuse reflector are measured, and the robust ball fitting algorithm is used to fit the group of hemispherically distributed spatial coordinate points into a group of spherically distributed spatial coordinate points, so as to obtain the coordinates of each point on the spherical surface of the spherical diffuse reflector, and then based on the coordinates of each point on the spherical surface, the coordinates of the center of the sphere in the laser radar coordinate system are calculated to obtain a cluster of millimeter-wave radar target points of the corner reflector detected by the millimeter-wave radar, and according to the placement position information of the calibration verification component, the coordinates of the detection point of the corner reflector in the millimeter-wave radar coordinate system are filtered out, that is, the second position information, and based on the coordinate conversion function from the millimeter-wave radar to the laser radar, the second position information of the detection point of the corner reflector in the millimeter-wave radar coordinate system is converted to the laser radar coordinate system to obtain the coordinates of the detection point of the corner reflector in the laser radar coordinate system, that is, the third position information.

2. The calibration verification component according to claim 1, characterized in that: The calibration verification component further comprises a pan / tilt platform (4), wherein the pan / tilt platform (4) comprises a fixed part (41) and a movable part (42); The fixing member (41) is fixedly connected to the bracket (3), and the first detection object (1) and the second detection object (2) are respectively mounted on the movable member (42).

3. The calibration verification component according to claim 2, characterized in that: The calibration verification assembly further comprises a first mounting rod (5) and a second mounting rod (6); One end of the first mounting rod (5) is fixedly connected to the movable member (42), and one end of the second mounting rod (6) is fixedly connected to the movable member (42); The first detection object (1) is mounted on the first mounting rod (5), and the second detection object (2) is mounted on the second mounting rod (6); the detection point of the first detection object (1), the detection point of the second detection object (2), the axis of the first mounting rod (5) and the axis of the second mounting rod (6) are on the same straight line.

4. The calibration verification component according to claim 3, characterized in that: The pan / tilt platform (4) further comprises a balancer (43); the balancer (43) is fixedly connected to the movable part (42); The detection plane of the balancing instrument (43) is perpendicular to the straight line.

5. The calibration verification component according to claim 3, characterized in that: The second mounting rod (6) comprises a slide groove (61) and a slide seat (62); the slide seat (62) is mounted in the slide groove (61); The second detection object (2) is mounted on the slide seat (62), and the sliding track of the detection point of the second detection object (2) is on the axis of the second mounting rod (6).

6. The calibration verification assembly according to claim 5, characterized in that: The outer wall of the second mounting rod (6) has a plurality of axially arranged scale lines and a scale value corresponding to each scale line, wherein the scale value corresponding to the scale line is the distance between the point corresponding to the scale line on the axis of the second mounting rod (6) and the detection point of the first detection object (1).

7. The calibration verification assembly according to claim 1, characterized in that: The support (3) is a tripod.

8. A calibration verification system, characterized in that: The calibration verification system comprises a target device (10), a calibration verification device (20) and a calibration verification component (30) as described in any one of claims 1 to 7, wherein multiple types of sensors (7) are installed on the target device (10), the calibration verification component (30) is located within the detection area of ​​the multiple types of sensors (7), and a communication connection is established between the multiple types of sensors (7) and the calibration verification device (20).

9. A calibration verification method, characterized in that: The calibration verification method is applied to the calibration verification system according to claim 8, wherein the multiple types of sensors (7) include a first sensor (71) and a second sensor (72), and the first sensor (71) and the second sensor (72) are different types of sensors (7), and the method comprises: The calibration verification device (20) acquires detection data of the first sensor (71) and detection data of the second sensor (72); the calibration verification device (20) determines first position information of a detection point of the first detection object (1) in a coordinate system of the first sensor (71) based on the detection data of the first sensor (71); The calibration verification device (20) determines second position information of a detection point of the second detection object (2) in a coordinate system of the second sensor (72) based on detection data of the second sensor (72); Relative position information of the first detection object (1) and the second detection object (2) is obtained; and based on the first position information, the second position information and the relative position information, an error value of a coordinate conversion function from the second sensor (72) to the first sensor (71) is determined.

10. The calibration verification method according to claim 9, characterized in that: The plurality of types of sensors (7) further include a third sensor (73), wherein the third sensor (73) and the first sensor (71) and the second sensor (72) are sensors (7) of different types, and the method includes: The calibration verification device (20) obtains detection data of the third sensor (73); The calibration verification device (20) determines third position information of a detection point of the first detection object (1) in a coordinate system of the third sensor (73) based on detection data of the third sensor (73); Based on the first position information and the third position information, an error value of a coordinate conversion function from the third sensor (73) to the first sensor (71) is determined.

Images