A visual-based underwater tethered robot homing method

Abstract

The application relates to a visual-based underwater tethered robot docking guidance method, which comprises the following steps: S1: collecting images by using a calibrated camera arranged on an underwater robot; S2: processing the collected images to identify whether a preset light source arranged on an underwater docking station exists in the images; and S3: when the preset light source is not identified, performing a heading angle guidance stage: processing the collected images containing a pipe cable, calculating an extension direction angle of the pipe cable, and providing a docking heading guidance for the underwater robot according to the extension direction angle. The application realizes the full-course autonomous docking guidance of the underwater tethered robot from a long distance to a short distance through pure visual information, reduces the system complexity and cost, and improves the concealment.

Description

Technical Field

[0001] This invention belongs to the field of underwater robot technology, specifically relating to a vision-based method for guiding a tethered underwater robot back to dock. Background Technology

[0002] Underwater robots can replace humans in carrying out observation, operation, and maintenance tasks in harsh underwater environments, playing an increasingly important role in marine resource development, environmental monitoring, and national defense. By establishing a cable connection between the underwater robot and the underwater docking station, a continuous energy supply, high-speed data transmission, and material exchange channel can be provided for the robot, significantly improving its operating time and capabilities.

[0003] Currently, the docking guidance method for underwater robots mainly adopts a phased acoustic and optical combined guidance scheme: acoustic equipment (such as ultra-short baseline) is used for coarse positioning at long distances, and optical equipment (such as underwater lights and cameras) is used for precise positioning at close distances. This method has the following shortcomings: First, both the underwater robot and the underwater docking station need to be equipped with both acoustic and optical guidance equipment, resulting in a complex system structure and high hardware costs; second, the propagation of acoustic signals is easily detected by passive sonar, leading to reduced stealth in underwater operations; third, existing docking guidance technology is mainly designed for untethered autonomous underwater vehicles (AUVs), and there is a lack of dedicated research and technical solutions for docking guidance of tethered underwater robots that are physically connected to the docking station.

[0004] Therefore, how to provide a simple, low-cost, highly concealed, and targeted docking guidance method for underwater tethered robots is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a vision-based underwater tethered robot docking guidance method. This method realizes the autonomous docking guidance of the underwater tethered robot from long distance to short distance through pure visual information, reducing system complexity and cost, and improving stealth.

[0006] The technical solution of this invention is to provide a vision-based underwater tethered robot docking guidance method, comprising the following steps:

[0007] S1: Acquire images using a calibrated camera mounted on the underwater robot;

[0008] S2: Process the acquired image to identify whether a preset light source set on the underwater dock station exists in the image;

[0009] S3: When the preset light source is not detected, the heading angle guidance stage is executed: the acquired image containing the cable is processed, the extension direction angle of the cable is calculated, and the underwater robot is provided with docking guidance based on the extension direction angle.

[0010] S4: When the preset light source is detected, the pose guidance stage is executed: the acquired image containing the preset light source is processed, the relative pose between the underwater dock and the underwater robot is calculated, and the underwater robot is provided with docking pose guidance based on the relative pose.

[0011] S5: Determine if a docking completion signal has been received. If not, repeat steps S1 to S4. If received, end docking guidance.

[0012] This invention provides guidance for docking based on visual images of cables and light sources. It eliminates the need for acoustic guidance, requiring only optical guidance. The equipment is simple, inexpensive, widely applicable, and highly concealed, making it of significant research value and practical application for docking of underwater tethered robots.

[0013] Preferably, step S2, which processes the image to identify the preset light source, specifically includes:

[0014] S21: Use threshold segmentation to extract suspected light source regions in the image;

[0015] S22: Extract the outline of the suspected light source area;

[0016] S23: The extracted contours are filtered according to the preset contour area threshold and the major-minor axis ratio threshold of the fitted ellipse; if a contour passes the filter, it is determined that the preset light source exists; if no contour passes the filter, it is determined that the preset light source does not exist.

[0017] Preferably, the calculation of the cable extension direction angle in step S3 specifically includes:

[0018] S31: Use camera calibration parameters to correct distortion in the acquired cable images;

[0019] S32: Enhance the image after distortion correction;

[0020] S33: Perform edge detection on the enhanced image and use morphological operations to remove noise and extract suspected cable edges;

[0021] S34: Based on the preset edge area threshold, edge roundness, and region growing method based on a specific region, the extracted suspected cable edges are screened to obtain the cable edges;

[0022] S35: Calculate the equation of the straight line at the edge of the cable using the Hough transform, and calculate the relative angle between the straight line and the underwater robot;

[0023] S36: Obtain the current heading angle of the underwater robot, and calculate the extension direction angle of the cable by combining the relative angle.

[0024] Preferably, step S4, calculating the relative pose between the underwater dock and the underwater robot, specifically includes:

[0025] S41: Perform distortion correction on the acquired light source image using camera calibration parameters;

[0026] S42: Use threshold segmentation to extract suspected light source regions in the distortion-corrected image;

[0027] S43: Extract the contour of the suspected light source area;

[0028] S44: The extracted contours are filtered according to the preset contour area threshold and the major-minor axis ratio threshold of the fitted ellipse. When there are multiple contours that pass the filter, the contour with the largest area is selected as the light source contour.

[0029] S45: Using the center point of the ellipse that fits the contour of the light source as the center point of the light source, calculate the relative position of the preset light source with respect to the underwater camera based on the camera imaging model and calibration parameters.

[0030] S46: Obtain the attitude angle data of the underwater robot and the preset attitude angle data of the underwater dock station, and calculate the relative pose between the underwater dock station and the underwater robot by combining the relative positions.

[0031] Preferably, the calibrated camera includes a first camera and a second camera. The first camera is used to acquire images containing the cable, and the second camera is used to acquire images containing a preset light source. The preset light source is set on the underwater docking station, and when the underwater robot returns to the dock, the optical axis of the second camera coincides with the center normal of the preset light source on the same straight line.

[0032] The first camera is an underwater camera that acquires images of the cable. Its installation requirements are as follows: the underwater camera is installed on the underwater robot, and the center point of the optical axis of the underwater camera and the geometric center point of the cable interface of the underwater robot are located in the same vertical plane; the roll angle of the underwater camera is 0 degrees, and the horizontal axis of its imaging sensor is parallel to the ground plane.

[0033] The second camera is an underwater camera that collects images of the underwater docking station light source. Its installation requirements are as follows: the underwater camera is installed on the underwater robot, and the underwater light source is installed on the underwater docking station. After returning to the dock, the optical axis of the underwater camera and the center normal of the underwater light source are aligned on the same straight line, and the coordinates of the center point of the light source in the resulting light source image coincide with the principal point of the underwater camera.

[0034] Preferably, in the heading angle guidance stage described in step S3, the underwater robot travels along the extension direction of the cable to gradually approach the underwater dock station.

[0035] Preferably, in the pose guidance stage described in step S4, the underwater robot adjusts its position and attitude according to the calculated relative pose in order to complete the precise docking with the underwater dock station.

[0036] Compared with the prior art, the present invention has the following advantages:

[0037] This invention addresses the characteristics of underwater tethered robots by providing a method for docking guidance based on visual images from cables and light sources. It eliminates the need for acoustic guidance, requiring only optical guidance. The equipment is simple, inexpensive, widely applicable, and highly concealed. Attached Figure Description

[0038] Figure 1 This is a flowchart of the method of the present invention.

[0039] Figure 2 This is a flowchart of S2 in this invention.

[0040] Figure 3 This is a flowchart of S3 in this invention.

[0041] Figure 4 This is a flowchart of S4 in this invention.

[0042] Figure 5 This is the information flow diagram for this embodiment.

[0043] Figure 6 This is a schematic diagram of S3 in this embodiment.

[0044] Figure 7 This is a schematic diagram of S4 in this embodiment. Detailed Implementation

[0045] The present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0046] like Figure 5As shown, the underwater robot is equipped with underwater camera #1, underwater camera #2, underwater light source #1, attitude sensor #1, and control board #1. The underwater dock is equipped with underwater light source #2, attitude sensor #2, and control board #2. A cable connects the underwater robot and the underwater dock, enabling RS485 serial communication between the two ends. Underwater camera #1 is used to acquire images from the cable; underwater light source #1 provides illumination for underwater camera #1 imaging; underwater camera #2 is used to acquire images from underwater light source #2 at the underwater dock; attitude sensors #1 and #2 are used to measure the attitude angles of the underwater robot and the underwater dock; control board #1 receives image information from underwater cameras #1 and #2, brightness information from underwater light source #1, attitude information from attitude sensor #1, and attitude information from the underwater dock sent by control board #2, and runs guidance software; control board #2 receives brightness information from underwater light source #2 and attitude information from attitude sensor #2. The optical axis center point of underwater camera #1 and the geometric center point of the underwater robot cable interface are located in the same vertical plane; the roll angle of underwater camera #1 is 0 degrees, and the horizontal axis of its imaging sensor is parallel to the ground plane. After returning to dock, the optical axis of underwater camera #2 and the center normal of underwater light source #2 coincide on the same straight line, and the coordinates of the center point of the light source in the resulting light source image coincide with the principal point of the underwater camera.

[0047] like Figure 1 As shown, a vision-based underwater tethered robot docking guidance method includes the following steps:

[0048] S1: Use calibrated underwater cameras #1 and #2 to acquire image #1 and image #2 respectively;

[0049] S2: Perform image processing on image #2 acquired by underwater camera #2 to identify whether underwater light source #2 on the underwater dock station exists in image #2;

[0050] S3: If underwater light source #2 on the underwater dock is not identified in image #2, then heading angle guidance is performed. Image processing is performed on image #1 acquired by underwater camera #1 to calculate the cable extension direction angle, providing heading angle guidance for the underwater robot's return to dock. Figure 6 As shown;

[0051] S4: If underwater light source #2 on the underwater dock is identified in the image, pose guidance is performed. Image processing is performed on image #1 acquired by underwater camera #1 to calculate the relative pose between the underwater dock and the underwater robot, providing pose guidance for the underwater robot's return to dock, such as... Figure 7 As shown;

[0052] S5: If no docking completion signal is received, repeat steps S1 to S4; if a docking completion signal is received, end the process.

[0053] like Figure 2As shown, the image processing described in S2 includes the following steps:

[0054] S21: Image segmentation, using a threshold segmentation method to extract suspected light source regions;

[0055] S22: Contour extraction, extracting the contour of the suspected light source area;

[0056] S23: Contour filtering. Light source regions are filtered by contour area and the ratio of the major and minor axes of the fitted ellipse. If a contour passes the filter, the image is considered to have a light source; if no contour passes the filter, the image is considered to have no light source.

[0057] like Figure 3 As shown, the image processing described in S3 includes the following steps:

[0058] S31: Distortion correction, using camera calibration parameters to correct image distortion;

[0059] S32: Image enhancement, using filtering and histogram equalization methods to enhance the image;

[0060] S33: Edge detection, using the Canny method to detect edge features in the image, removing noise through morphological operations, and extracting suspected cable edges;

[0061] S34: Edge filtering, filtering cable edges by edge area, edge roundness, and region growth based on a specific area;

[0062] S35: Hough Transform, using the Hough Transform to calculate the equation of the straight line at the edge of the cable and to calculate the relative angle between the straight line at the edge of the cable and the underwater robot;

[0063] S36: Angle calculation: Based on the underwater robot's heading angle and the relative angle between the straight line at the edge of the cable and the underwater robot, calculate the cable extension direction angle.

[0064] like Figure 4 As shown, the image processing described in S4 includes the following steps:

[0065] S41: Distortion correction, using camera calibration parameters to correct image distortion;

[0066] S42: Image segmentation, using a threshold segmentation method to extract suspected light source regions;

[0067] S43: Contour extraction, extracting the contour of the suspected light source area;

[0068] S44: Contour filtering. The light source area is filtered by contour area and the ratio of the major and minor axes of the fitted ellipse. If multiple contours pass the filtering, the contour with the largest area is selected as the light source contour.

[0069] S45: Position calculation. The center point of the fitted ellipse is taken as the center point of the light source. Based on the camera imaging model and calibration parameters, and assuming that the vertical distance between the underwater light source and the underwater camera is 1 meter, the relative position of the underwater light source with respect to the underwater camera is calculated.

[0070] S46: Pose calculation. The underwater robot control board #1 receives the attitude angle data fed back by the attitude sensor #1 and the underwater dock station attitude angle data transmitted by the underwater dock station attitude sensor #2 through the control board #2. Combined with the relative position of the underwater light source #2 relative to the underwater camera #2, the relative pose of the underwater dock station and the underwater robot is calculated.

[0071] This invention provides guidance for docking based on visual images of the cable and light source, eliminating the need for acoustic guidance and requiring only optical guidance. The equipment is simple, low-cost, widely applicable, and highly concealed, making it significant for the research and application of underwater tethered robots returning to dock. When the underwater robot is far from the dock, the light source image is too small or invisible in the field of view. In this case, the method automatically operates in the heading angle guidance stage, guiding the robot along the cable direction. As the robot approaches, the light source gradually appears in the image and meets the recognition threshold, at which point the method automatically switches to the pose guidance stage, using the light source for precise positioning. The entire switching process requires no manual intervention, achieving fully autonomous docking.

[0072] The above description only illustrates preferred embodiments of the present invention and should not be construed as limiting the scope of the claims. Any equivalent procedural modifications made using this specification are included within the patent protection scope of this invention.

Claims

1. A vision-based method for guiding an underwater tethered robot back to dock, characterized in that: Includes the following steps, S1: Acquire images using a calibrated camera mounted on the underwater robot; S2: Process the acquired image to identify whether a preset light source set on the underwater dock station exists in the image; S3: When the preset light source is not detected, the heading angle guidance stage is executed: the acquired image containing the cable is processed, the extension direction angle of the cable is calculated, and the underwater robot is provided with docking guidance based on the extension direction angle. S4: When the preset light source is detected, the pose guidance stage is executed: the acquired image containing the preset light source is processed, the relative pose between the underwater dock and the underwater robot is calculated, and the underwater robot is provided with docking pose guidance based on the relative pose. S5: Determine if a docking completion signal has been received. If not, repeat steps S1 to S4. If received, end docking guidance.

2. The vision-based underwater tethered robot docking guidance method according to claim 1, characterized in that: Step S2, which involves processing the image to identify the preset light source, specifically includes: S21: Use threshold segmentation to extract suspected light source regions in the image; S22: Extract the outline of the suspected light source area; S23: The extracted contours are filtered according to the preset contour area threshold and the major-minor axis ratio threshold of the fitted ellipse; if a contour passes the filter, it is determined that the preset light source exists; if no contour passes the filter, it is determined that the preset light source does not exist.

3. The vision-based underwater tethered robot docking guidance method according to claim 1, characterized in that: Step S3, calculating the cable extension direction angle, specifically includes: S31: Use camera calibration parameters to correct distortion in the acquired cable images; S32: Enhance the image after distortion correction; S33: Perform edge detection on the enhanced image and use morphological operations to remove noise and extract suspected cable edges; S34: Based on the preset edge area threshold, edge roundness, and region growing method based on a specific region, the extracted suspected cable edges are screened to obtain the cable edges; S35: Calculate the equation of the straight line at the edge of the cable using the Hough transform, and calculate the relative angle between the straight line and the underwater robot; S36: Obtain the current heading angle of the underwater robot, and calculate the extension direction angle of the cable by combining the relative angle.

4. The vision-based underwater tethered robot docking guidance method according to claim 1, characterized in that: Step S4, calculating the relative pose between the underwater dock and the underwater robot, specifically includes: S41: Perform distortion correction on the acquired light source image using camera calibration parameters; S42: Use threshold segmentation to extract suspected light source regions in the distortion-corrected image; S43: Extract the contour of the suspected light source area; S44: The extracted contours are filtered according to the preset contour area threshold and the major-minor axis ratio threshold of the fitted ellipse. When there are multiple contours that pass the filter, the contour with the largest area is selected as the light source contour. S45: Using the center point of the ellipse that fits the contour of the light source as the center point of the light source, calculate the relative position of the preset light source with respect to the underwater camera based on the camera imaging model and calibration parameters. S46: Obtain the attitude angle data of the underwater robot and the preset attitude angle data of the underwater dock station, and calculate the relative pose between the underwater dock station and the underwater robot by combining the relative positions.

5. The vision-based underwater tethered robot docking guidance method according to claim 1, characterized in that: The calibrated camera includes a first camera and a second camera. The first camera is used to acquire images containing the cable, and the second camera is used to acquire images containing a preset light source. The preset light source is set on the underwater docking station, and when the underwater robot returns to the dock, the optical axis of the second camera coincides with the center normal of the preset light source on the same straight line.

6. The vision-based underwater tethered robot docking guidance method according to claim 5, characterized in that: The installation requirements for the first camera are as follows: its optical axis center point and the geometric center point of the underwater robot's cable interface are located in the same vertical plane, its installation roll angle is 0 degrees, and the horizontal axis of the imaging sensor is parallel to the ground plane.

7. The vision-based underwater tethered robot docking guidance method according to claim 1, characterized in that: In the heading angle guidance stage described in step S3, the underwater robot travels along the extension direction of the cable to gradually approach the underwater dock station.

8. The vision-based underwater tethered robot docking guidance method according to claim 1, characterized in that: In the pose guidance stage described in step S4, the underwater robot adjusts its position and attitude according to the calculated relative pose in order to complete the precise docking with the underwater dock station.

Images