3D scanner

Abstract

Embodiments of the present disclosure provide a three-dimensional scanner. The three-dimensional scanner comprises: a mounting support comprising a base, a camera mounting rack and a radar mounting rack arranged on the base, a side of the camera mounting rack facing the radar mounting rack being provided with a mounting portion; a rotating holder rotatably connected to the base and adapted to drive the base to rotate about a rotation axis; a plurality of image acquisition devices arranged along a circumference of the mounting portion and adapted to acquire image data of an object within a corresponding camera field of view, each camera field of view being centered on a corresponding camera normal, the camera normals of the plurality of image acquisition devices being in a same normal plane and intersecting at a normal intersection point, the rotation axis passing through the normal intersection point; and a laser radar arranged on a side of the radar mounting rack facing the camera mounting rack and adapted to acquire point cloud data of an object within a radar field of view, the radar field of view being centered on a radar normal.

Description

TECHNICAL FIELD

[0001] Example embodiments of the present disclosure generally relate to a three-dimensional scanner. BACKGROUND

[0002] A three-dimensional scanner is a scientific instrument that can detect and analyze shape and appearance data of objects or environments in the real world using three-dimensional scanning technology. Three-dimensional scanning technology can achieve non-contact measurement of objects, converting real-world three-dimensional information into digital signals that can be recognized and directly processed by computers. The data collected by a three-dimensional scanner is often used to perform three-dimensional reconstruction calculations to create a digital model of the actual object in a virtual world. One of the uses of a three-dimensional scanner is to create a point cloud of the geometric surface of an object, which can be used to interpolate the surface shape of the object. The denser the point cloud, the more accurate the model created. If the scanner can obtain the surface color of the object, it can further paste a material map on the reconstructed surface, i.e., so-called material imprinting.

[0003] A three-dimensional scanner can include a camera and a laser radar to form an optical detection part. The camera is used to obtain image data, and the laser radar is used to obtain depth information. Three-dimensional scanners are also increasingly used in the measurement and scanning of building structures. For example, when scanning an indoor environment, a three-dimensional scanner is usually placed at a fixed position inside the building to be scanned, and its optical detection part is driven by a driving component to rotate one revolution to complete the scanning of the room structure. SUMMARY

[0004] In one aspect of the present disclosure, a three-dimensional scanner is provided. The three-dimensional scanner includes: a mounting bracket including a base, a camera mounting rack and a radar mounting rack disposed on the base, a mounting portion disposed on a side of the camera mounting rack facing the radar mounting rack; a rotating holder rotatably connected to the base and adapted to drive the base to rotate about a rotation axis; a plurality of image acquisition devices arranged along a circumference of the mounting portion and adapted to acquire image data of an object within a corresponding camera field of view, each camera field of view being centered on a corresponding camera normal, the camera normals of the plurality of image acquisition devices being in the same normal plane and intersecting at a normal intersection point, the rotation axis passing through the normal intersection point; and a laser radar disposed on a side of the radar mounting rack facing the camera mounting rack and adapted to acquire point cloud data of an object within a radar field of view, the radar field of view being centered on a radar normal.

[0005] The layout of the image acquisition device of the 3D scanner according to the embodiments of this disclosure is more reasonable, so that the image acquisition device only needs to rotate half a revolution around the longitudinal axis to obtain a panoramic image of the entire space when acquiring image data, which significantly improves scanning efficiency. In addition, by making the rotation axis pass through the intersection of the normals, the parallax that may occur during the rotation of the image acquisition device around the rotation axis can be effectively reduced or eliminated, thereby reducing the workload of subsequent data processing and improving the model creation effect. Furthermore, by arranging the LiDAR on the side of the LiDAR mounting bracket facing the camera mounting bracket, the optical center of the LiDAR can be made as close as possible to the intersection of the normals, thereby reducing the calibration error of the depth information and better optimizing the image stitching effect based on the depth information.

[0006] In some embodiments, the normal plane containing the camera normals of the multiple image acquisition devices is a plane, and the radar normal is perpendicular to the normal plane. This arrangement reduces the influence of the image acquisition devices and the lidar on each other's field of view, thereby facilitating a compact arrangement of the image acquisition devices and the lidar, and thus promoting the miniaturization of the 3D scanner.

[0007] In some embodiments, the radar normal passes through the intersection of normals. This arrangement increases the overlap between the camera's field of view and the radar's field of view, effectively reducing or eliminating parallax between the image acquisition device and the lidar, thereby improving data processing efficiency, reducing the workload of image data and point cloud data processing, and thus obtaining a better spatial model.

[0008] In some embodiments, the normal plane containing the camera normals of the multiple image acquisition devices is a conical surface, and the radar normal is collinear with the center line of the conical surface. This arrangement facilitates a compact arrangement of the image acquisition devices and the lidar, promotes field-of-view overlap between the image acquisition devices and the lidar, and facilitates subsequent parameter calibration, thereby improving the quality of data fusion.

[0009] In some embodiments, the radar field of view and the camera field of view partially overlap. This arrangement allows the image acquisition device and the lidar to be calibrated using image data and point cloud data acquired at a fixed location during 3D scanner calibration. This calibration is then used to calculate calibration parameters, thereby improving parameter accuracy and ultimately enhancing the fusion effect of image data and point cloud data.

[0010] In some embodiments, multiple image acquisition devices are circumferentially arranged within a range of greater than 180° in the circumferential direction of the mounting portion. This arrangement significantly reduces the number of acquisition angles required, thereby reducing errors caused by multiple image stitching and improving scanning efficiency.

[0011] In some embodiments, the camera mount also includes a main body portion connected to the base, with the mounting portion protruding from the main body portion toward the radar mount and recessed relative to the main body portion in the radial direction. This provides better protection for the lens modules of multiple image acquisition devices against impacts or damage. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. In the drawings, the same or similar reference numerals denote the same or similar elements, wherein:

[0013] Figure 1 A front view of a 3D scanner according to an embodiment of the present disclosure is shown;

[0014] Figure 2 A left-side view of a 3D scanner according to an embodiment of the present disclosure is shown;

[0015] Figure 3 A 3D scanner according to an embodiment of the present disclosure is shown along... Figure 1 A schematic cross-sectional view of the section intercepted by line AA in the diagram;

[0016] Figure 4 A 3D scanner according to an embodiment of the present disclosure is shown along... Figure 2 A schematic cross-sectional view of the section intercepted by the BB line; and

[0017] Figure 5 A schematic diagram of the camera field of view and radar field of view of a 3D scanner according to an embodiment of the present disclosure is shown. Detailed Implementation

[0018] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0019] In the description of embodiments of this disclosure, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0020] A 3D scanner, also known as a 3D camera or virtual reality camera, is used to detect and analyze the shape (geometric structure) and appearance data (such as color, surface albedo, etc.) of objects or environments in the real world. A 3D scanner used for scanning room structures typically consists of two parts: a first part, also called the drive unit fixing part, which is used to fix drive components such as stepper motors; and a second part, also called the sensing part, which typically includes an image acquisition device such as a camera and a lidar sensor. The sensing part can rotate relative to the first part about a rotation axis under the drive of the drive components. When scanning the interior structure of a room, the 3D scanner is usually fixed at a predetermined position in the room to be scanned by a support device such as a tripod. The drive components drive the sensing part to rotate by a predetermined angle to complete the scanning of the entire interior structure of the room, thereby obtaining image data and point cloud data about the interior structure of the room.

[0021] A 3D scanner's image acquisition device is used to acquire image data of objects in the environment. An important parameter of this device is the field of view (FOV). In optical engineering, the field of view, also known as the field of view area, determines the range of the optical instrument's field of view. A camera's field of view is typically a cone-shaped region centered on the normal and with the camera's origin as its vertex; the apex angle of this cone is the field of view. The field of view of each camera may differ when measured in the lateral and longitudinal directions. For example, some cameras may have a longitudinal field of view of approximately 100°, while the lateral field of view may be around 80°.

[0022] LiDAR, also known as Laser Radar, is an abbreviation for Light Detection and Ranging. It's a radar system that uses sensing technology to emit low-power, eye-safe laser pulses for measurement, measuring the time it takes for the laser to travel round-trip between the sensor and the target. The resulting aggregated data is used to generate 3D point cloud images, providing spatial location and depth information for identifying, classifying, and tracking moving objects. Similar to cameras, LiDAR also has a field of view, which is the angle covered by the sensor. The field of view of LiDAR is centered on the radar normal.

[0023] Traditional 3D scanners typically employ a configuration where the camera normal of the image acquisition device and the radar normal of the LiDAR are arranged coaxially or parallel to each other. This facilitates the calibration and integration of data collected by the two sensors. However, this arrangement limits the scanning range of both the camera and LiDAR. Furthermore, current traditional 3D scanners have relatively low scanning efficiency and pixel density, failing to meet the requirements for efficient full-space scanning.

[0024] The present disclosure provides a 3D scanner to solve, or at least partially solve, the aforementioned or other potential problems of conventional 3D scanners. The image acquisition device of the 3D scanner according to the present disclosure has a more rational layout, enabling the image acquisition device to obtain a full-space panoramic image by rotating only half a revolution around the longitudinal axis (i.e., the rotation axis mentioned below) when acquiring image data. Furthermore, the LiDAR of the 3D scanner according to the present disclosure also improves the scanning range and scanning method. Through the arrangement of the image acquisition device and LiDAR, as well as various improvements to parameters such as the field of view, scanning range, and scanning method, the 3D scanner according to the present disclosure can reduce depth information errors, better optimize image stitching effects based on depth information, and significantly improve scanning efficiency.

[0025] The specific structure and improvements of the 3D scanner according to embodiments of the present disclosure will now be described with reference to the accompanying drawings. Figure 1 A front view of a 3D scanner according to an embodiment of the present disclosure is shown. Figure 2 A left-side view of a 3D scanner according to an embodiment of the present disclosure is shown. Figure 3 A 3D scanner according to an embodiment of the present disclosure is shown along... Figure 1 A schematic cross-sectional view of the section intercepted by line AA in the diagram. Figure 4 A 3D scanner according to an embodiment of the present disclosure is shown along... Figure 2 A schematic cross-sectional view taken by the BB line in the diagram. Figure 5 A schematic diagram of the camera field of view and radar field of view of a 3D scanner according to an embodiment of the present disclosure is shown. Figures 1 to 5 As shown, the 3D scanner according to embodiments of the present disclosure generally includes a mounting bracket 10, a plurality of image acquisition devices 20, a lidar 30, and a rotating gimbal 40.

[0026] Mounting bracket 10 is used to mount multiple image acquisition devices 20, LiDAR 30, and rotating gimbal 40. For example... Figures 1 to 5 As shown, the mounting bracket 10 includes a base 11 and a camera mounting bracket 12 and a radar mounting bracket 13 disposed on the base 11. The base 11 supports the camera mounting bracket 12 and the radar mounting bracket 13 and is rotatably connected to a gimbal 40. The camera mounting bracket 12 has a mounting section 121 on the side facing the radar mounting bracket 13, which is used to mount multiple image acquisition devices 20. The radar mounting bracket 13 is used to mount a lidar 30. The gimbal 40 can drive the base 11 to rotate around a rotation axis 101, thereby causing the multiple image acquisition devices 20 and the lidar 30 to rotate around the rotation axis 101.

[0027] like Figure 1 and Figure 3As shown, multiple image acquisition devices 20 are arranged circumferentially on the mounting portion 121. The outer surface of the mounting portion 121 may be generally columnar. Multiple camera windows arranged circumferentially spaced apart may be provided on the mounting portion 121. The multiple camera windows correspond to the positions of the multiple image acquisition devices 20 arranged in the mounting portion 121. The multiple image acquisition devices 20 are arranged circumferentially on the mounting portion 121, and each of the multiple image acquisition devices 20 can acquire image data of objects within the camera's field of view 21 through its corresponding camera window. The multiple image acquisition devices 20 can be fixed in the mounting portion 121 by any suitable means, such as bolts, clips, or other types of fasteners; the embodiments of this disclosure are not limited in this regard.

[0028] In some embodiments, such as Figures 1 to 5 As shown, the camera mounting bracket 12 also includes a main body portion 122 connected to the base 11. The mounting portion 121 protrudes from the main body portion 122 toward the radar mounting bracket 13 and is recessed relative to the main body portion 122 by a certain distance in the radial direction. In this way, the lens modules of the multiple image acquisition devices 20 can be better protected from impacts or damage. In some embodiments, a transparent cover made of a material such as glass or plastic can be provided at the camera window on the mounting portion 121 through which the camera field of view 21 of the image acquisition device 20 passes, thereby further protecting the lens modules of the image acquisition device 20.

[0029] exist Figure 3 In the diagram, each camera field of view 21 is indicated by a solid-lined triangle, and the camera normal 22 of each image acquisition device 20 is indicated by a dashed line. Each camera field of view 21 is distributed in a roughly conical shape around its corresponding camera normal 22. The camera normals 22 of the multiple image acquisition devices 20 are arranged in the same plane, which will be referred to as the normal plane below. The normal plane extends substantially in the longitudinal direction. Of course, it should be understood that the multiple camera normals 22 may not be located in the same plane, but rather in a conical surface. That is, in some embodiments, the normal plane containing the multiple camera normals 22 may also be a conical surface.

[0030] from Figure 3As can be seen, the camera fields of view 21 of the multiple image acquisition devices 20 overlap with each other. In some embodiments, the overlap of the camera fields of view 21 of two adjacent image acquisition devices 20 is greater than 10°. The overlapping camera fields of view 21 of the multiple image acquisition devices 20 facilitates the calibration of the parameters of each image acquisition device 20, thereby improving the calibration accuracy. In this way, the control unit of the 3D scanner can reliably calibrate the parameters of the image acquisition devices 20 using the overlapping fields of view, thereby improving the processing effect of subsequent data processing. On the other hand, the overlapping fields of view of the image acquisition devices 20 can reduce or eliminate image distortion that may exist at the edges of the image acquisition devices 20, thereby significantly improving the imaging effect of the final panoramic image.

[0031] Figure 3 The image also shows that the camera normals 22 of multiple image acquisition devices 20 intersect at a single point, namely, the normal intersection point 23. Figure 3 In the embodiment shown with six image acquisition devices 20, the six camera normals 22 of the six image acquisition devices 20 intersect at the normal intersection point 23. This arrangement can eliminate parallax that may exist between multiple image acquisition devices 20, thereby reducing the difficulty of stitching panoramic images and thus improving the imaging effect of panoramic images.

[0032] It should be noted that the figures, values, numbers, etc. mentioned above and elsewhere in this disclosure are exemplary and are not intended to limit the scope of this disclosure in any way. Any other suitable figures, values, numbers, etc. are possible.

[0033] like Figure 3 As shown, the rotation axis 101 is also in the normal plane. For example, in some embodiments, the rotation axis 101 may pass through the intersection of the normals 23. In this way, parallax that may occur when the image acquisition device 20 rotates around the rotation axis 101 can be effectively reduced or eliminated, thereby reducing the workload of subsequent data processing and improving the model creation effect.

[0034] The lidar 30 is used to measure depth information of the environment. For example... Figure 1 , Figure 4 and Figure 5 As shown, the lidar 30 is arranged on the side of the radar mounting bracket 13 facing the camera mounting bracket 12 to collect point cloud data of objects within the radar field of view 31. Figure 5 The radar field of view 31 of the lidar 30 is shown to be a roughly conical region centered on the radar normal. The so-called conical shape refers to the shape formed by subtracting a coaxial cone with a smaller apex angle from a cone with a larger apex angle (e.g., close to or greater than 180°). Figure 5The radar field of view 31 of the LiDAR 30 shown is merely a cross-sectional shape. It should be understood that the actual shape of the radar field of view 31 is a toroidal cone shape obtained by rotating the illustrated cross-sectional shape around the radar normal 32. The LiDAR 30 can acquire point cloud data of all objects within the radar field of view 31. In this way, the efficiency of full-space scanning and the reliability of point cloud data can be further improved. Furthermore, by arranging the LiDAR 30 on the side of the radar mounting bracket 13 facing the camera mounting bracket 12, the optical center of the LiDAR 30 can be brought as close as possible to the intersection point 23 of the normals, thereby reducing the calibration error of depth information and better optimizing the image stitching effect based on the depth information.

[0035] It should be understood that Figure 5 The radar field of view 31 of the lidar 30 shown is merely illustrative and is not intended to limit the scope of this disclosure. The 3D scanner according to embodiments of this disclosure may employ lidar 30 with any other field of view and scanning method. For example, in some embodiments, lidar 30 may also employ line-scan radar or any other suitable radar.

[0036] According to an embodiment of this disclosure, the radar normal 32 of the lidar 30 of the 3D scanner and the normal plane containing the camera normal 22 of the image acquisition device 20 form a non-zero angle. For example, in... Figure 4 and Figure 5 In the illustrated embodiment, when the normal plane is planar, the radar normal 32 is perpendicular to the normal plane. This arrangement reduces the influence of the image acquisition device 20 and the lidar 30 on each other's field of view, thereby facilitating a compact arrangement of the image acquisition device 20 and the lidar 30, and promoting the miniaturization of the 3D scanner.

[0037] exist Figure 5 In the illustrated embodiment, the radar field of view 31 of the LiDAR 30 and the camera field of view 21 of the image acquisition device 20 partially overlap. This arrangement allows the image acquisition device 20 and the LiDAR 30 to be calibrated using image data and point cloud data acquired at a fixed location during 3D scanner calibration. This allows for the calculation of calibration parameters, thereby improving parameter accuracy and ultimately enhancing the fusion effect of image data and point cloud data. Furthermore, this arrangement enables the LiDAR 30 and the image acquisition device 20 of the 3D scanner to simultaneously scan the space, acquiring point cloud data and image data of at least a certain area in the space at the same time. The point cloud data and image data can be processed immediately and simultaneously for fusion to generate a virtual reality image with depth information about that area. In this way, compared to the conventional solution where the radar needs to acquire point cloud data of a certain area while scanning, the 3D scanner according to the present disclosure can further improve data acquisition efficiency and data fusion quality.

[0038] In some embodiments, when the normal plane is planar, the radar normal 32 is not only perpendicular to the normal plane but also passes through the intersection point 23 of the normals, making the radar normal 32 perpendicular to the rotation axis 101. This arrangement increases the overlap area between the camera field of view 21 and the radar field of view 31, effectively reducing or eliminating the parallax between the image acquisition device 20 and the lidar 30. This allows the gimbal 40 to acquire all images and depth information in less than one rotation, thereby improving the data processing effect, reducing the workload of image data and point cloud data processing, and thus obtaining a better spatial model.

[0039] As described above, the normal plane containing the multiple camera normals 22 can also be a conical surface. In this case, the radar normals 32 can be arranged collinear with the center line of the conical surface. This arrangement also facilitates the compact arrangement of the image acquisition device 20 and the lidar 30, further promoting the overlap of the fields of view of the image acquisition device 20 and the lidar 30, as well as subsequent parameter calibration, and thus improving the quality of data fusion.

[0040] In some embodiments, the field of view of the image acquisition device 20 can also be adjusted or selected according to the number of image acquisition devices 20. For example, if six image acquisition devices 20 are used, the field of view of each image acquisition device 20 can be set to 60°. If four image acquisition devices 20 are used, the field of view of each image acquisition device 20 can be set to 99.43°. If three image acquisition devices 20 are used, the field of view of each image acquisition device 20 can be set to greater than or equal to 120°. If five image acquisition devices 20 are used, the field of view of each image acquisition device 20 can be set to 77° to 80°.

[0041] The field of view of the image acquisition device 20 can be set to the required angle at the factory. In some embodiments, the field of view of the image acquisition device 20 can also be adjusted according to an algorithm. When adjusting the field of view using an algorithm, areas with severe edge distortion can be cropped out, thereby further improving image quality.

[0042] In some embodiments, the camera normals 22 of the multiple image acquisition devices 20 can be distributed within a predetermined angular range in the circumferential direction. For example, the predetermined angular range can be from 180° to 240°. For example, in some embodiments, the camera normals 22 of the multiple image acquisition devices 20 can be distributed within an angle of approximately 220° in the circumferential direction. The normals of the multiple image acquisition devices 20 may or may not equally divide this angle, but the fields of view of the multiple image acquisition devices 20 centered on the camera normals 22 can at least partially overlap. In this way, it is beneficial for the stitching and calibration of image data acquired by the multiple image acquisition devices 20.

[0043] With six image acquisition devices 20 having a field of view of approximately 60°, excluding a partial blind zone at the bottom, the six image acquisition devices 20 can acquire image data within a circumferential range of approximately 320°. That is, when the 3D scanner is in a fixed position, the six image acquisition devices 20 can acquire image data from approximately one full circle from front to back. In this case, to acquire image data across the entire spatial range, it is only necessary to rotate the six image acquisition devices 20 half a circle (180°) around the rotation axis 101. Compared to conventional 3D scanners that can only acquire images in one direction, the 3D scanner according to the embodiments of this disclosure can significantly reduce the number of acquisition angles required, thereby reducing errors caused by multiple image stitching and improving scanning efficiency.

[0044] In some embodiments, such as Figure 3 and Figure 4 As shown, a gimbal 40 is positioned at the bottom of the base 11 to drive the mounting bracket 10 and its components to rotate around the rotation axis 101. Driven by the gimbal 40, the mounting bracket 10 can rotate the image acquisition device 20 and the LiDAR 30 mounted thereon around the rotation axis 101, thereby achieving circumferential full-space scanning. The gimbal 40 can be coupled to a support device for supporting the 3D scanner and, driven by a power component such as a motor, can rotate the mounting bracket 10 around the rotation axis 101, thereby driving the image acquisition device 20 and the LiDAR 30 to rotate around the rotation axis 101.

[0045] The various embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A three-dimensional scanner, comprising: a mounting bracket (10) including a base (11) and a camera mounting frame (12) and a radar mounting frame (13) disposed on the base (11), a side of the camera mounting frame (12) facing the radar mounting frame (13) being provided with a mounting portion (121); a rotating holder (40) rotatably connected to the base (11) and adapted to drive the base (11) to rotate about a rotation axis (101); a plurality of image acquisition devices (20) arranged along a circumference of the mounting portion (121) and adapted to acquire image data of an object within a corresponding camera field of view (21), each of the camera field of view (21) being centered on a corresponding camera normal line (22), the camera normal lines (22) of the plurality of image acquisition devices (20) being in a same normal plane and intersecting at a normal intersection point (23), the rotation axis (101) passing through the normal intersection point (23); and a laser radar (30) arranged on a side of the radar mounting frame (13) facing the camera mounting frame (12) and adapted to acquire point cloud data of an object within a radar field of view (31), the radar field of view (31) being centered on a radar normal line (32).

2. The three-dimensional scanner of claim 1, wherein the normal plane in which the camera normal lines (22) of the plurality of image acquisition devices (20) lie is a plane, and the radar normal line (32) is perpendicular to the normal plane.

3. The three-dimensional scanner of claim 2, wherein the radar normal line (32) passes through the normal intersection point (23).

4. The three-dimensional scanner of claim 1, wherein the normal plane in which the camera normal lines (22) of the plurality of image acquisition devices (20) lie is a conic plane, and the radar normal line (32) is collinear with a center line of the conic plane.

5. The three-dimensional scanner of claim 1, wherein the radar field of view (31) partially overlaps the camera field of view (21).

6. The three-dimensional scanner of claim 1, wherein the plurality of image acquisition devices (20) are arranged circumferentially in a range of more than 180° in circumference of the mounting portion (121).

7. The three-dimensional scanner of claim 1, wherein the camera mounting frame (12) further includes a main body portion (122) connected with the base (11), the mounting portion (121) protruding from the main body portion (122) toward the radar mounting frame (13) and being recessed relative to the main body portion (122) in a radial direction of the mounting portion (121). ​

Images