Tunnel unmanned aerial vehicle positioning method and system based on laser ranging

Abstract

The application provides a tunnel unmanned aerial vehicle positioning method and system based on laser ranging, relates to the unmanned aerial vehicle positioning technical field, and comprises the following steps: S1, at time t0, establishing an unmanned aerial vehicle coordinate system with a known position of a tunnel unmanned aerial vehicle as an origin in the tunnel, and controlling the tunnel unmanned aerial vehicle to fly along the Z axis; S2, synchronously acquiring a tunnel top image captured by a camera at a first elevation angle in the YZ plane when the tunnel unmanned aerial vehicle is at the origin and a first distance measured by a laser range finder when the laser range finder irradiates a reference point in the tunnel top image; S3, at time t1, synchronously acquiring a tunnel top image captured by the camera when the camera tracks the reference point, the camera being at a second elevation angle in the YZ plane and the second elevation angle being rotated from the first elevation angle, and a second distance measured by the laser range finder when the laser range finder irradiates the reference point in the tunnel top image; and S4, performing positioning calculation according to the known position of the tunnel unmanned aerial vehicle, the first elevation angle, the first distance, the second elevation angle and the second distance.

Description

Technical Field

[0001] This invention relates to the field of UAV positioning technology, specifically to a tunnel UAV positioning method and system based on laser ranging. Background Technology

[0002] Currently, drones face severe positioning challenges in enclosed environments such as tunnels, mainly because they cannot receive GPS signals. Existing alternative technologies, such as inertial navigation, have accumulated errors that cannot be eliminated, visual positioning is limited by the single environmental features and insufficient light, and ultra-wideband technologies rely on pre-built infrastructure, which are costly and lack universality, making it difficult to achieve stable and accurate autonomous positioning. Summary of the Invention

[0003] The purpose of this invention is to provide a tunnel drone positioning method and system based on laser ranging. The method uses a camera and laser rangefinder physically attached to the drone to locate a reference point in the tunnel ceiling. Based on this, the current position of the drone is calculated according to the position of the reference point, without the need for GPS signals.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following solution:

[0005] A tunnel drone positioning method based on laser ranging, wherein the tunnel drone is equipped with a camera and a laser rangefinder physically bound together, and the tunnel drone positioning method includes:

[0006] S1. At time t0, inside the tunnel, with the known position of the tunnel UAV as the origin, establish a UAV coordinate system containing mutually perpendicular X-axis, Y-axis and Z-axis, and control the tunnel UAV to fly along the Z-axis.

[0007] S2. Simultaneously acquire the tunnel top image captured by the camera on the YZ plane at the origin of the tunnel UAV at the first elevation angle, and the first distance measured by the reference point in the tunnel top image illuminated by the laser rangefinder.

[0008] S3. At time t1, synchronously acquire the image of the tunnel top taken by the camera when it rotates from the first elevation angle to the second elevation angle in the YZ plane while tracking the reference point, and the second distance measured by the reference point in the image of the tunnel top illuminated by the laser rangefinder.

[0009] S4. Based on the known position, first elevation angle, first distance, second elevation angle and second distance of the tunnel UAV, the positioning calculation is performed to obtain the position of the tunnel UAV at time t1.

[0010] A further preferred technical solution is that the rotation axes of the camera and the laser rangefinder are synchronized, so that the light spot of the laser rangefinder is fixedly placed in the area captured by the camera with the center point of the image as the axis.

[0011] A further preferred technical solution is that S2 includes the following steps:

[0012] S21. Obtain the tunnel top image taken by the camera on the YZ plane at the origin of the tunnel UAV with the first elevation angle, and identify the reference point in the tunnel top image through the feature recognition model.

[0013] S22. Synchronously control the camera and laser rangefinder to rotate until the image captured by the camera is centered on the reference point, and the spot of the laser rangefinder falls on the area centered on the reference point.

[0014] S23. Obtain the first distance obtained by the laser rangefinder measuring the distance to the light spot.

[0015] A further preferred technical solution is that the process of acquiring the tunnel top image taken by the camera on the YZ plane at a first elevation angle when the tunnel drone is at the origin is specifically as follows:

[0016] The elevation angle range of the camera on the YZ plane is obtained in advance. At time t0, when the tunnel UAV is at the origin, the elevation angle of the current camera on the YZ plane is obtained. It is determined whether the elevation angle falls within the elevation angle range. If so, the elevation angle is taken as the first elevation angle. If not, the current camera is controlled to rotate to the minimum value in the elevation angle range, and the minimum value is taken as the first elevation angle.

[0017] A further preferred technical solution is that the process of identifying reference points in the tunnel top image through a feature recognition model is as follows:

[0018] A pre-trained feature recognition model based on tunnel features is obtained. The feature recognition model is used to perform feature recognition on the tunnel top image. When multiple features are recognized, the feature with the smallest deviation from the Z-axis and the closest distance to the camera is selected as the reference point.

[0019] A further preferred technical solution is that when the camera is tracking the reference point, it keeps the captured image centered on the reference point, and simultaneously controls the camera and laser rangefinder to rotate so that the laser rangefinder's spot falls on the area centered on the reference point; then the center of the tunnel top image captured at the second elevation angle is the reference point; and the second distance is obtained by the laser rangefinder measuring the distance of the spot falling on the area centered on the reference point.

[0020] A further preferred technical solution is that the positioning calculation process is as follows:

[0021] On the YZ plane, a trigonometric function relationship between the tunnel UAV and the reference point is constructed. Substituting the known position, first elevation angle, and first distance of the tunnel UAV, the position of the reference point is calculated. Then, substituting the position of the reference point, second elevation angle, and second distance, the position of the tunnel UAV at time t1 is calculated.

[0022] A tunnel UAV positioning system based on laser ranging, employing the aforementioned tunnel UAV positioning method based on laser ranging, includes:

[0023] Initialization module: At time t0, inside the tunnel, with the known position of the tunnel UAV as the origin, establish a UAV coordinate system containing mutually perpendicular X-axis, Y-axis and Z-axis, and control the tunnel UAV to fly along the Z-axis;

[0024] Reference point finding module: synchronously acquires the tunnel top image taken by the camera on the YZ plane at the origin of the tunnel UAV at the first elevation angle, and the first distance measured by the reference point in the tunnel top image illuminated by the laser rangefinder;

[0025] Reference point tracking module: At time t1, synchronously acquire the tunnel top image captured by the camera rotating from the first elevation angle to the second elevation angle in the YZ plane when the camera is tracking the reference point, as well as the second distance measured by the laser rangefinder illuminating the reference point in the tunnel top image;

[0026] UAV positioning module: Based on the known position, first elevation angle, first distance, second elevation angle and second distance of the tunnel UAV, the positioning calculation is performed to obtain the position of the tunnel UAV at time t1.

[0027] The beneficial effects of this invention are:

[0028] This invention provides a method and system for locating unmanned aerial vehicles (UAVs) in tunnels based on laser ranging. Applied to tunnel environments, it solves the problem of UAVs being unable to locate themselves when flying in tunnels due to the inability to receive GPS signals. It employs a camera and a laser rangefinder, which are physically bound to the UAV to form a collaborative system with visual guidance as the core for fixed-point measurement. This system tracks reference points inside the tunnel, thereby achieving the location of the UAV without the need for GPS signals. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the tunnel UAV positioning method in Embodiment 1 of the present invention;

[0030] Figure 2 This is a schematic diagram illustrating the synchronization of the camera and the laser rangefinder in Embodiment 1 of the present invention;

[0031] Figure 3This is a schematic diagram of the positioning calculation on the YZ plane in Embodiment 1 of the present invention. Detailed Implementation

[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The following description of at least one exemplary embodiment is merely illustrative and is in no way intended to limit the present invention or its application or use. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0033] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps described in these embodiments do not limit the scope of the invention.

[0034] At the same time, it should be understood that, for ease of description, the dimensions of the various parts shown in the accompanying drawings are not drawn according to actual scale.

[0035] Furthermore, for clarity and brevity, descriptions of well-known structures, functions, and configurations may have been omitted. Those skilled in the art will recognize that various changes and modifications can be made to the examples described herein without departing from the spirit and scope of this disclosure.

[0036] Techniques, methods, and equipment known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and equipment should be considered part of the specification.

[0037] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values.

[0038] The present invention will now be described in detail with reference to the accompanying drawings and embodiments:

[0039] Example 1

[0040] In this embodiment, to address the problem of drones being unable to locate themselves when flying in tunnels due to the lack of GPS signals, a tunnel drone positioning method based on laser ranging is proposed. This requires mounting a physically bound camera and laser rangefinder on the drone to form a unified sensing system. Physical binding means that the space between the camera and the laser rangefinder is relatively fixed. This relative spatial fixation ensures that the laser rangefinder's spot inevitably falls on a stable and unchanging coordinate position within the camera's field of view, providing stable and reliable spatial reference data for the drone.

[0041] like Figure 1 As shown, the tunnel UAV positioning method includes the following steps:

[0042] S1. At time t0, inside the tunnel, with the known position of the tunnel UAV as the origin, establish a UAV coordinate system containing mutually perpendicular X-axis, Y-axis and Z-axis, and control the tunnel UAV to fly along the Z-axis.

[0043] S2. Simultaneously acquire the tunnel top image captured by the camera on the YZ plane at the origin of the tunnel UAV at the first elevation angle, and the first distance measured by the reference point in the tunnel top image illuminated by the laser rangefinder.

[0044] S3. At time t1, synchronously acquire the image of the tunnel top taken by the camera when it rotates from the first elevation angle to the second elevation angle in the YZ plane while tracking the reference point, and the second distance measured by the reference point in the image of the tunnel top illuminated by the laser rangefinder.

[0045] S4. Based on the known position, first elevation angle, first distance, second elevation angle and second distance of the tunnel UAV, the positioning calculation is performed to obtain the position of the tunnel UAV at time t1.

[0046] Specifically, the subway tunnel presents a tubular space. Inside the tunnel, a drone coordinate system is established with the known location of the drone as the origin. The Z-axis of the drone coordinate system is along the central axis of the tunnel; the Y-axis of the drone coordinate system is vertically upward within the vertical screen; and the X-axis of the drone coordinate system is horizontal to the right. Since all three are perpendicular to each other, when the drone flies inside the tunnel, it flies along the Z-axis, which points in the direction of the drone's movement.

[0047] Based on this, the YZ plane refers to the plane formed by the Y-axis and Z-axis, which is a vertical plane sliced ​​along the tunnel direction. Since the top of a subway tunnel is not perfectly smooth, a laser rangefinder is used to illuminate the tunnel top, where diffuse reflection occurs, allowing the laser rangefinder to measure distances to a specific point. Therefore, this invention proposes to synchronously control the rotation of the camera and laser rangefinder. This rotation refers to rotation around the X-axis; that is, while the tunnel drone flies along the Z-axis, the camera and laser rangefinder are synchronously controlled to rotate towards the tunnel top, with the rotation angle lying on the YZ plane.

[0048] The camera and laser rangefinder rotate synchronously. When the camera is controlled to center the image around a reference point, the laser rangefinder's spotlight is fixed within the area captured by the camera, centered on the image center. However, the laser rangefinder does not measure the object itself, but rather a fixed position next to it. The system continuously locks the camera to a fixed reference point, ensuring the camera lens remains focused on one position throughout the drone's movement. Therefore, the laser rangefinder also measures the same position—the target's location. Figure 2 As shown, the target's position always falls within the camera's upward-looking viewfinder; that is, the target's position within the dotted frame is relatively fixed. Once the camera detects an object in the current frame, it can adjust its angle to keep the object centered in the image. This ensures that throughout the drone's movement, the laser rangefinder consistently measures a stable and unchanging coordinate position in space, providing the drone with stable and reliable spatial reference data. Furthermore, as the drone continues to move, the camera's rotation and detection allow for continuous target tracking. Once the camera detects a target, it essentially determines an observation direction, and the laser rangefinder's emission axis will point in that direction. Through physical bonding and joint control, the two work together to achieve continuous tracking and measurement of a specific spatial position.

[0049] A further preferred technical solution is that S2 includes the following steps:

[0050] S21. Obtain the tunnel top image taken by the camera on the YZ plane at the origin of the tunnel UAV with the first elevation angle, and identify the reference point in the tunnel top image through the feature recognition model.

[0051] S22. Synchronously control the camera and laser rangefinder to rotate until the image captured by the camera is centered on the reference point, and the spot of the laser rangefinder falls on the area centered on the reference point.

[0052] S23. Obtain the first distance obtained by the laser rangefinder measuring the distance to the light spot.

[0053] A further preferred technical solution is that the process of acquiring the tunnel top image taken by the camera on the YZ plane at a first elevation angle when the tunnel drone is at the origin is specifically as follows:

[0054] The elevation angle range of the camera on the YZ plane is obtained in advance. At time t0, when the tunnel UAV is at the origin, the elevation angle of the current camera on the YZ plane is obtained. It is determined whether the elevation angle falls within the elevation angle range. If so, the elevation angle is taken as the first elevation angle. If not, the current camera is controlled to rotate to the minimum value in the elevation angle range, and the minimum value is taken as the first elevation angle.

[0055] Based on this, the present invention sets a constraint on the rotation of the synchronously controlled camera and laser rangefinder. By constraining this rotation within a preset elevation angle range, image distortion, ranging failure, or decreased data accuracy due to excessively large or small rotation angles when tracking a reference point can be avoided, thus ensuring the reliability of the measurement results. Furthermore, it ensures that the tracked target is always located within the effective center of the sensor's field of view and measurement range.

[0056] Specifically, in this embodiment, the elevation angle range can be set to 30° to 60°, allowing for some margin in the mechanical gimbal device that drives the camera and laser rangefinder to rotate. When the drone starts, at time t0, the mechanical gimbal device is controlled to adjust the elevation angle of the camera and laser rangefinder to 30°. Then, the camera searches for feature reference points based on the pre-trained feature recognition model. In actual scenarios, reference points in the shape of holes will inevitably appear on the top of subway tunnels. This feature can be trained, and after the camera identifies a reference point, it tracks that reference point within the elevation angle range. Alternatively, if the camera does not identify a reference point, the elevation angle can be adjusted, increasing by a certain angle each time, which can be determined based on the camera's performance. Here, adjusting by 2° each time will adjust the elevation angle from 30° to 32°. The camera then identifies and selects a reference point in the tunnel top image captured at 32°.

[0057] In one embodiment, as the drone flies along the Z-axis, when the drone is at an elevation angle of 30°, it finds the first reference point using a camera and a laser rangefinder. This first reference point can be considered as the reference point furthest from the drone. At this time, the drone tracks the first reference point by synchronously rotating the camera and laser rangefinder. When the elevation angle is rotated to 60°, it can be considered that the drone has almost reached below the first reference point, indicating that the first reference point has been used up and a new reference point can be found. This allows the present invention to continuously locate the tunnel drone without the need for a GPS signal.

[0058] A further preferred technical solution is that the process of identifying reference points in the tunnel top image through a feature recognition model is as follows:

[0059] A pre-trained feature recognition model based on tunnel features is obtained. The feature recognition model is used to perform feature recognition on the tunnel top image. When multiple features are recognized, the feature with the smallest deviation from the Z-axis and the closest distance to the camera is selected as the reference point.

[0060] A further preferred technical solution is that when the camera is tracking the reference point, it keeps the captured image centered on the reference point, and simultaneously controls the camera and laser rangefinder to rotate so that the laser rangefinder's spot falls on the area centered on the reference point; then the center of the tunnel top image captured at the second elevation angle is the reference point; and the second distance is obtained by the laser rangefinder measuring the distance of the spot falling on the area centered on the reference point.

[0061] A further preferred technical solution is that the positioning calculation process is as follows:

[0062] On the YZ plane, a trigonometric function relationship between the tunnel UAV and the reference point is constructed. Substituting the known position, first elevation angle, and first distance of the tunnel UAV, the position of the reference point is calculated. Then, substituting the position of the reference point, second elevation angle, and second distance, the position of the tunnel UAV at time t1 is calculated.

[0063] Specifically, such as Figure 3 As shown, on the YZ plane, the tunnel UAV, at a known position (y0, z0) at time t0, acquires the elevation angle a0 of the camera to the reference point on the YZ plane and the distance d0 of the reference point measured by the laser rangefinder. At this point, assuming the position of the reference point is (yp, zp), a trigonometric function relationship between the tunnel UAV and the reference point is constructed on the YZ plane. The position of the reference point is calculated based on this trigonometric function relationship, using the formulas: yp = y0 + d0 * sin(a0), zp = z0 + d0 * cos(a0). Then, the tunnel UAV starts from time t... Starting from point 0, the UAV tracks the reference point along the Z-axis. At time t1, the UAV is at position (y1, z1). However, since there is no GPS signal inside the tunnel, position (y1, z1) is unknown. At this point, the elevation angle a1 of the camera to the reference point on the YZ plane and the distance d1 of the reference point measured by the laser rangefinder are obtained. Substituting these values ​​into the reference point's position (yp, zp), the position of the UAV at time t1 is obtained by inverse trigonometric functions. The inverse formulas are: y1 = yp - d1 * sin(a1), z1 = zp - d1 * cos(a1). After the UAV passes the reference point, a new reference point can be found for the next round of calculations, and so on.

[0064] In order to improve the accuracy of positioning, the present invention can also set up a detection position inside the tunnel during the positioning process of the tunnel drone using a camera and a laser rangefinder. For example, a QR code or a special mark can be set up inside the tunnel. When the drone flies to the detection position, the camera and laser rangefinder are controlled to be vertically upward, and calibration is performed through this mark.

[0065] Example 2

[0066] A tunnel UAV positioning system based on laser ranging, employing the aforementioned tunnel UAV positioning method based on laser ranging, includes:

[0067] Initialization module: At time t0, inside the tunnel, with the known position of the tunnel UAV as the origin, establish a UAV coordinate system containing mutually perpendicular X-axis, Y-axis and Z-axis, and control the tunnel UAV to fly along the Z-axis;

[0068] Reference point finding module: synchronously acquires the tunnel top image taken by the camera on the YZ plane at the origin of the tunnel UAV at the first elevation angle, and the first distance measured by the reference point in the tunnel top image illuminated by the laser rangefinder;

[0069] Reference point tracking module: At time t1, synchronously acquire the tunnel top image captured by the camera rotating from the first elevation angle to the second elevation angle in the YZ plane when the camera is tracking the reference point, as well as the second distance measured by the laser rangefinder illuminating the reference point in the tunnel top image;

[0070] UAV positioning module: Based on the known position, first elevation angle, first distance, second elevation angle and second distance of the tunnel UAV, the positioning calculation is performed to obtain the position of the tunnel UAV at time t1.

[0071] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Based on the technical essence of the present invention, any simple modifications, equivalent substitutions, and improvements made to the above embodiments within the spirit and principles of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A tunnel unmanned aerial vehicle (UAV) positioning method based on laser ranging, characterized in that, The tunnel drone is equipped with a physically attached camera and a laser rangefinder. The positioning method for the tunnel drone includes: S1. At time t0, inside the tunnel, with the known position of the tunnel UAV as the origin, establish a UAV coordinate system containing mutually perpendicular X-axis, Y-axis and Z-axis, and control the tunnel UAV to fly along the Z-axis. S2. Simultaneously acquire the tunnel top image captured by the camera on the YZ plane at the origin of the tunnel UAV at the first elevation angle, and the first distance measured by the reference point in the tunnel top image illuminated by the laser rangefinder. S2 includes the following steps: S21. Obtain the tunnel top image taken by the camera on the YZ plane at the origin of the tunnel UAV with the first elevation angle, and identify the reference point in the tunnel top image through the feature recognition model. S22. Synchronously control the camera and laser rangefinder to rotate until the image captured by the camera is centered on the reference point, and the spot of the laser rangefinder falls on the area centered on the reference point. S23. Obtain the first distance from the laser rangefinder by measuring the distance to the light spot; S3. At time t1, synchronously acquire the tunnel top image taken by the camera when it rotates from the first elevation angle to the second elevation angle in the YZ plane while tracking the reference point in the elevation angle range, and the second distance measured by the laser rangefinder illuminating the reference point in the tunnel top image. When the camera tracks the reference point within the range of elevation angles, the captured image is kept centered on the reference point. The camera and laser rangefinder are rotated synchronously so that the laser rangefinder's spot falls on the area centered on the reference point. S4. Based on the known position, first elevation angle, first distance, second elevation angle and second distance of the tunnel UAV, the positioning calculation is performed to obtain the position of the tunnel UAV at time t1.

2. The tunnel UAV positioning method based on laser ranging according to claim 1, characterized in that, The rotation axes of the camera and the laser rangefinder are synchronized, so that the laser rangefinder's spot is fixedly placed within the area captured by the camera with the center point of the image as the axis.

3. The tunnel UAV positioning method based on laser ranging according to claim 1, characterized in that, The process of acquiring the tunnel top image taken by the camera on the YZ plane at the origin by the tunnel drone at the first elevation angle is as follows: The elevation angle range of the camera on the YZ plane is obtained in advance. At time t0, when the tunnel UAV is at the origin, the elevation angle of the current camera on the YZ plane is obtained. It is determined whether the elevation angle falls within the elevation angle range. If so, the elevation angle is taken as the first elevation angle. If not, the current camera is controlled to rotate to the minimum value in the elevation angle range, and the minimum value is taken as the first elevation angle.

4. The tunnel UAV positioning method based on laser ranging according to claim 1, characterized in that, The process of identifying reference points in the tunnel top image using a feature recognition model is as follows: A pre-trained feature recognition model based on tunnel features is obtained. The feature recognition model is used to perform feature recognition on the tunnel top image. When multiple features are recognized, the feature with the smallest deviation from the Z-axis and the closest distance to the camera is selected as the reference point.

5. The tunnel UAV positioning method based on laser ranging according to claim 1, characterized in that, The center of the tunnel top image taken at the second elevation angle is the reference point; then the second distance is obtained by the laser rangefinder measuring the distance of the light spot falling within the area centered on the reference point.

6. The tunnel UAV positioning method based on laser ranging according to claim 1, characterized in that, The specific process of location calculation is as follows: On the YZ plane, a trigonometric function relationship between the tunnel UAV and the reference point is constructed. Substituting the known position, first elevation angle, and first distance of the tunnel UAV, the position of the reference point is calculated. Then, substituting the position of the reference point, second elevation angle, and second distance, the position of the tunnel UAV at time t1 is calculated.

7. A tunnel unmanned aerial vehicle (UAV) positioning system based on laser ranging, characterized in that, The method for locating unmanned aerial vehicles (UAVs) in tunnels based on laser ranging, as described in any one of claims 1-6, includes: Initialization module: At time t0, inside the tunnel, with the known position of the tunnel UAV as the origin, establish a UAV coordinate system containing mutually perpendicular X-axis, Y-axis and Z-axis, and control the tunnel UAV to fly along the Z-axis; Reference point finding module: synchronously acquires the tunnel top image taken by the camera on the YZ plane at the origin of the tunnel UAV at the first elevation angle, and the first distance measured by the reference point in the tunnel top image illuminated by the laser rangefinder; Reference point tracking module: At time t1, synchronously acquire the tunnel top image captured by the camera rotating from the first elevation angle to the second elevation angle in the YZ plane when the camera is tracking the reference point, as well as the second distance measured by the laser rangefinder illuminating the reference point in the tunnel top image; UAV positioning module: Based on the known position, first elevation angle, first distance, second elevation angle and second distance of the tunnel UAV, the positioning calculation is performed to obtain the position of the tunnel UAV at time t1.

Images