Optical positioning-based underwater cleaning robot coverage visualization method

Abstract

The present application relates to underwater cleaning robot technical field, specifically for underwater cleaning robot coverage visualization method based on optical positioning, method includes the following steps: deploying multiple optical sensors above the working water area, based on optical sensor captures the image sequence of optical mark point installed on underwater cleaning robot; based on the image sequence determines the current position coordinates of underwater cleaning robot in the three-dimensional space of working water area; according to the current position coordinates, in the preset working water area model, the motion trajectory of underwater cleaning robot is updated and rendered;Preset coverage determination rule, based on the motion trajectory according to the coverage determination rule determines the cleaned area of working water area;The cleaned area and the non-cleaned area are differentially visually presented, and the overall cleaning coverage is calculated and displayed in real time, which can improve the accuracy of underwater robot operation, and also enhance the user's control ability to the cleaning process through intuitive visualization interface.

Description

Technical Field

[0001] This invention relates to the field of underwater cleaning robot technology, specifically a method for visualizing the coverage of underwater cleaning robots based on optical positioning. Background Technology

[0002] In the field of underwater cleaning, underwater cleaning robots are widely used for cleaning closed bodies of water with regular geometric shapes (including but not limited to rectangular and circular pools), such as swimming pools, industrial aquaculture ponds, and large aquariums. However, the current level of intelligence in these robots is generally low, with the following significant drawbacks: Once a traditional underwater cleaning robot submerges in water, its precise location, movement path, and real-time attitude cannot be directly observed. Operators and managers cannot know whether the robot is working according to the predetermined plan, or whether it is spinning in place for a long time, getting stuck, or missing large areas. This invisibility of the process leads to management relying on experience and speculation, making effective real-time monitoring and intervention impossible.

[0003] To address these issues, we propose an optical positioning-based method for visualizing the coverage of underwater cleaning robots. Summary of the Invention

[0004] The purpose of this invention is to provide a method for visualizing the coverage of underwater cleaning robots based on optical positioning, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for visualizing the coverage of an underwater cleaning robot based on optical positioning, the method comprising the following steps:

[0006] Multiple optical sensors are deployed above the operating water area to capture image sequences of optical markers mounted on the underwater cleaning robot.

[0007] The current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area are determined based on the image sequence;

[0008] Based on the current location coordinates, update and render the underwater cleaning robot's trajectory in the preset operating water area model;

[0009] Pre-defined coverage determination rules are used to determine the cleaned areas of the work area based on the movement trajectory and coverage determination rules;

[0010] The system presents cleaned and uncleaned areas with distinct visual representations and calculates and displays the overall cleaning coverage rate in real time.

[0011] Preferably, the step of deploying multiple optical sensors above the operating water area and capturing image sequences of optical markers mounted on the underwater cleaning robot based on the optical sensors includes:

[0012] Multiple optical sensors are deployed above the operating area, and the optical sensors are configured to cover the entire operating area from different perspectives;

[0013] Optical markers are placed on the top of the underwater cleaning robot. These optical markers are multiple active light-emitting or passive reflective elements arranged in a specific spatial pattern to form a unique code that can be used for identification.

[0014] Multiple optical sensors are triggered synchronously to capture a synchronized sequence of images containing the optical markers.

[0015] Preferably, the step of determining the current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area based on the image sequence includes:

[0016] Acquire a sequence of images simultaneously captured by multiple optical sensors deployed above the operating water area, wherein the image sequence contains images of optical markers on the underwater cleaning robot;

[0017] Optical markers in each frame of the image are identified and extracted to obtain their two-dimensional pixel coordinates in the pixel coordinate system of each optical sensor.

[0018] Based on the pre-calibrated internal and external parameters of the multiple optical sensors, the coordinates of multiple two-dimensional pixels are fused through a multi-view visual triangulation algorithm to calculate the current position coordinates of the optical marker in the three-dimensional space of the working water area.

[0019] The current position coordinates are output as the real-time pose information of the underwater cleaning robot.

[0020] Preferably, the step of updating and rendering the underwater cleaning robot's trajectory in a preset operating water area model based on the current location coordinates includes:

[0021] Obtain the real-time serialized current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area;

[0022] The real-time serialized current position coordinates are spliced ​​with the historical position coordinates to form continuous motion trajectory data;

[0023] In the preset 3D or 2D unfolded model of the working water area, a trajectory line representing the movement path of the underwater cleaning robot is dynamically rendered and updated based on the motion trajectory data.

[0024] Preferably, the steps of the preset coverage determination rule include:

[0025] Obtain three-dimensional model data of the work area to be cleaned; based on the three-dimensional model data, discretize the cleanable surface of the work area to obtain multiple coverage determination units, wherein the discretization process is to divide the cleanable surface into planar grid units of equal area, or to establish a three-dimensional voxel model on the cleanable surface of the three-dimensional model.

[0026] A preset coverage determination rule is established, wherein the coverage determination rule is: a preset spatial neighborhood is established with the center of each coverage determination unit as the reference; when the positioning coordinates of the underwater cleaning robot fall into the preset spatial neighborhood of any coverage determination unit, the coverage determination unit is determined to have been cleaned.

[0027] During the operation of the underwater cleaning robot, the cleaned area of ​​the working water area is determined based on its real-time status and coverage determination rules.

[0028] Preferably, the step of determining the cleaned area of ​​the working water area based on its real-time status and coverage determination rules during the operation of the underwater cleaning robot includes:

[0029] Set selection points for each coverage determination unit;

[0030] The current position coordinates of the underwater cleaning robot are obtained in real time, and the coverage determination unit corresponding to the current position coordinates in the 3D model data is used as the current coverage determination unit.

[0031] The selected point corresponding to the current coverage determination unit is copied to obtain multiple sub-selected points, and communication connections are established between the selected point and the sub-selected points.

[0032] Obtain other coverage determination units adjacent to the current coverage determination unit, distribute multiple sub-selection points to the other adjacent coverage determination units respectively, and mark the coverage determination unit where the sub-selection point is located as a candidate clean state. Based on multiple sub-selection points, obtain multiple candidate clean states.

[0033] The system collects the next coverage determination unit corresponding to the trajectory line of the underwater cleaning robot in real time, obtains the candidate clean state corresponding to the next coverage determination unit based on multiple candidate clean states as the target clean state, and re-marks the other coverage determination units corresponding to the remaining candidate clean states as unclean states.

[0034] The selected points on other coverage determination units corresponding to the target's clean status are copied and distributed, and the area composed of all coverage determination units on the trajectory line is taken as the cleaned area of ​​the operating water area.

[0035] Preferably, the step of visually differentiating cleaned and uncleaned areas and calculating and displaying the overall cleaning coverage in real time includes:

[0036] Acquire real-time positioning data and coverage determination results of the underwater cleaning robot in the operating area;

[0037] Based on the coverage determination results, the cleaned area is rendered in the first visual style on the visualization interface, and the uncleaned area is rendered in the second visual style that can be distinguished from the first visual style, forming a differentiated visual presentation.

[0038] Based on the coverage determination result, the proportion of the currently cleaned area to the total cleanable area is calculated in real time to obtain the overall cleaning coverage rate; the overall cleaning coverage rate value is updated and displayed in real time along with the differentiated visual presentation interface.

[0039] Preferably, the step of calculating the proportion of the currently cleaned area to the total cleanable area in real time based on the coverage determination result to obtain the overall cleaning coverage rate includes:

[0040] Obtain the total cleanable surface area of ​​the operating water area. Based on the preset coverage determination rules, the total cleanable surface area is discretized into multiple grid cells of equal area;

[0041] Count the total number of grid cells marked as "clean" in real time. Through formula Calculate the cumulative area of ​​the areas that have been marked as "cleaned". ,in, Represents the area of ​​a single grid cell;

[0042] Calculate and update overall cleaning coverage in real time. The corresponding formula is ;

[0043] The overall cleaning coverage is displayed in real time on the visualization interface using at least one of the following methods: numerical percentage and graphical progress bar.

[0044] Compared with the prior art, the beneficial effects of the present invention are:

[0045] 1. By deploying a relatively low-cost optical sensor network above the pool, combined with coded optical markers on the robot, real-time 3D positioning with sub-centimeter accuracy was achieved. Administrators can view a dynamically extending, precise movement trajectory in real-time on a 2D or 3D model of the pool, improving the accuracy of underwater robot operations. Furthermore, the intuitive visualization interface enhances user control over the cleaning process, enabling effective real-time monitoring and intervention of the underwater cleaning robot.

[0046] 2. By pre-copying and distributing "sub-selection points" to adjacent potential next coverage units within the current cycle, the marking and pre-rendering of coverage areas are brought forward. Logically, this implements a parallel processing mode of "prediction-confirmation." When the real location data arrives at the next moment, only a quick "confirmation" or "cancel" operation is needed, instead of calculating and rendering from scratch. This significantly shortens the latency from data reception to visualization updates, marking potentially rendering coverage units as "pending" and pre-rendering them, essentially preparing the data and graphics instructions needed for the next frame in advance. When updates are needed, the cached pre-rendered results can be directly called, greatly reducing the computation and communication required for rendering each frame, thereby improving overall graphics rendering efficiency and significantly enhancing the efficiency and real-time performance of coverage area visualization. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of the method flow of the present invention;

[0049] Figure 2 This is a schematic diagram of the trajectory line of the present invention. Detailed Implementation

[0050] 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. 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.

[0051] Example

[0052] Please see Figures 1 to 2 This invention provides a technical solution for a method to visualize the coverage of an underwater cleaning robot based on optical positioning: The method for visualizing the coverage of an underwater cleaning robot based on optical positioning includes the following steps:

[0053] S1: Deploy multiple optical sensors above the operating water area to capture image sequences of optical markers mounted on the underwater cleaning robot based on the optical sensors;

[0054] The steps of deploying multiple optical sensors above the operating area and capturing image sequences of optical markers mounted on an underwater cleaning robot based on the optical sensors include: deploying multiple optical sensors above the operating area, the optical sensors being configured to cover the entire operating area from different perspectives; setting optical markers on the top of the underwater cleaning robot, wherein the optical markers are multiple actively emitting or passively reflecting elements arranged in a specific spatial pattern to form a unique code that can be used for identification; and synchronously triggering multiple optical sensors to capture a synchronized image sequence containing the optical markers.

[0055] Specifically, the optical sensor is a global shutter industrial camera, and the optical markers are active light-emitting elements, preferably LEDs; alternatively, the optical markers are passive reflective elements, preferably reflective spheres or reflective marks; the operating water area is a closed body of water with a regular geometric shape, including swimming pools, aquaculture ponds, or water tanks; the number of multiple optical sensors is 4 to 6, evenly deployed above and around the operating water area; image sequences of optical markers on the underwater cleaning robot are simultaneously captured by multiple optical sensors; these are used for subsequent real-time calculation of the current position coordinates of the underwater cleaning robot in the three-dimensional space of the operating water area based on multi-view vision; the synchronous triggering of multiple optical sensors can reduce the depth uncertainty of monocular vision and the calculation error caused by the asynchrony of images from different cameras due to robot movement, thereby improving the accuracy of the calculation; by setting coded markers, the identity of the current underwater cleaning robot can be accurately identified, allowing the system to perform fast and unambiguous identity recognition and matching when there are multiple robots or interfering light sources in the field of view, avoiding mistracking.

[0056] S2: Determine the current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area based on the image sequence;

[0057] The steps for determining the current position coordinates of an underwater cleaning robot in the three-dimensional space of the operating area based on an image sequence include: acquiring an image sequence synchronously collected by multiple optical sensors deployed above the operating area, wherein the image sequence contains images of optical markers on the underwater cleaning robot; identifying and extracting the optical markers in each frame of the image to obtain their two-dimensional pixel coordinates in the pixel coordinate system of each optical sensor; fusing the multiple two-dimensional pixel coordinates using a multi-view visual triangulation algorithm based on the pre-calibrated internal and external parameters of the multiple optical sensors to calculate the current position coordinates of the optical markers in the three-dimensional space of the operating area; and outputting the current position coordinates as the real-time pose information of the underwater cleaning robot.

[0058] Specifically, a sub-pixel localization algorithm is used to extract the two-dimensional pixel coordinates of the optical markers. This algorithm includes the gray-level centroid method or Gaussian fitting. The multi-view visual triangulation algorithm is specifically a minimization of reprojection error algorithm, which is achieved by solving the following objective function: ,in, This represents the optimal three-dimensional position coordinates that minimize the objective function. Let the three-dimensional position of the robot be the solution. For the first The projection function of an optical sensor. In the first The pixel coordinates of optical markers observed in each optical sensor. Indicates the robot's position based on the current assumption. Through the first The projection model of a camera is used to calculate the theoretical pixel coordinates at which the robot marker should appear. Indicates all Sum the squares of the reprojection errors of each camera; Represents the total error function The parameter that gets the minimum value By utilizing information from multiple perspectives for mutual verification and constraint, the target's position in three-dimensional space can be deduced very accurately. Based on pre-calibrated internal and external parameters of multiple optical sensors, the coordinates of multiple two-dimensional pixels are fused using a multi-view vision triangulation algorithm to calculate the current position coordinates of the optical marker in the three-dimensional space of the working water area. Simultaneously, bundle adjustment optimization is periodically performed to minimize the cumulative reprojection error of the entire system in the time dimension by jointly optimizing the three-dimensional positions of all optical markers and the poses of multiple optical sensors in multiple frames of images. The current position coordinates calculated based on multi-view vision are fused with data from the inertial measurement unit built into the underwater cleaning robot. Through a Kalman filter or extended Kalman filter, smoothed and predicted higher-frequency robot pose information is obtained and used as the final real-time pose information. The sensor fusion step can continuously output the robot's pose information based on inertial data when the optical marker is temporarily occluded, causing pure vision positioning to fail.

[0059] S3: Based on the current location coordinates, update and render the movement trajectory of the underwater cleaning robot in the preset working water area model;

[0060] The steps for updating and rendering the motion trajectory of an underwater cleaning robot in a preset working water area model based on its current position coordinates include: obtaining the real-time serialized current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area; splicing the real-time serialized current position coordinates with the historical position coordinates to form continuous motion trajectory data; and dynamically rendering and updating a trajectory line representing the movement path of the underwater cleaning robot in a preset three-dimensional model or two-dimensional unfolded model of the working water area based on the motion trajectory data.

[0061] When rendering trajectory lines within a 3D model, the trajectory lines are configured to always conform to the bottom or vertical surface of the working water area to realistically reflect the robot's actual contact path. This surface-conforming requirement addresses the issue of pure spatial coordinates potentially causing the trajectory to "float" in the water, failing to accurately reflect the robot's movement along the pool wall / bottom, thus improving the realism and accuracy of the visualization. The visual attributes of the trajectory lines dynamically change according to preset rules. These attributes include at least one of color, thickness, transparency, or style. The preset rule for dynamic changes is: assigning different colors or transparency to the trajectory lines based on the freshness of the robot's movement trajectory, with the most recent trajectory having higher visual salience than historical trajectories; dynamic rendering and... The trajectory line is updated using an incremental drawing method, which means that only the latest position coordinates are connected to the trajectory endpoint of the previous moment and added to the visualization interface. This avoids the computational overhead of redrawing the entire huge trajectory chain with each update, ensuring the smoothness of the system under high-frequency positioning data. In the visualization interface, an independent, time-updating graphical icon represents the current instantaneous position and attitude of the underwater cleaning robot in real time. The shape or color of the graphical icon is used to indicate the current working status or operating mode of the underwater cleaning robot. Binding the icon to the working status (such as "cleaning", "paused", "return to charging") makes the visualization interface more information-rich, improves user experience and monitoring efficiency.

[0062] S4: Preset coverage determination rules, determine the cleaned area of ​​the working water area based on the movement trajectory and the coverage determination rules;

[0063] The steps for setting up coverage determination rules include: acquiring 3D model data of the work area to be cleaned; discretizing the cleanable surface of the work area based on the 3D model data to obtain multiple coverage determination units, wherein the discretization process involves dividing the cleanable surface into planar grid units of equal area, or establishing a 3D voxel model on the cleanable surface of the 3D model; setting up coverage determination rules, wherein the coverage determination rules are: establishing a preset spatial neighborhood based on the center of each coverage determination unit; when the positioning coordinates of the underwater cleaning robot fall within the preset spatial neighborhood of any coverage determination unit, the coverage determination unit is determined to have been cleaned; during the operation of the underwater cleaning robot, the cleaned area of ​​the work area is determined according to its real-time status and the coverage determination rules;

[0064] Specifically, the preset spatial neighborhood is a circular area centered on the center of the grid cell with a radius within a certain threshold, such as a radius between 5cm and 15cm, or a cubic area centered on the center of the voxel with a side length between 10cm and 20cm; the preset coverage determination rules may also include: determining whether the cleaning mechanism of the underwater cleaning robot is in an active working state; when the cleaning mechanism is in an active working state, and the robot's positioning coordinates satisfy the spatial association condition with a certain coverage determination cell, then the coverage determination cell is determined to have been cleaned; acquiring the real-time attitude data of the underwater cleaning robot; based on the attitude data The spatial relationship between the underwater cleaning robot and the coverage determination unit is used to determine whether the cleaning mechanism is effectively acting on the unit. When it is determined to be effective, the coverage determination unit is determined to be clean. The number of times the underwater cleaning robot covers the same coverage determination unit is recorded. Only when the number of coverages reaches or exceeds a preset threshold is the coverage determination unit finally determined to be clean. By setting a threshold for the coverage frequency, the underwater cleaning robot can handle stubborn stains. Only after cleaning the same area multiple times is it considered clean. This makes the visualization results more reflective of the real cleaning effect and improves the authenticity and accuracy of the coverage statistics.

[0065] During the operation of an underwater cleaning robot, the steps for determining the cleaned area of ​​the working water area based on its real-time status and coverage determination rules include: setting selection points for each coverage determination unit; acquiring the current position coordinates of the underwater cleaning robot in real time, and using the coverage determination unit corresponding to the current position coordinates in the 3D model data as the current coverage determination unit; copying the selection point corresponding to the current coverage determination unit to obtain multiple sub-selection points, and establishing communication connections between the selection points and sub-selection points; acquiring other coverage determination units adjacent to the current coverage determination unit, distributing the multiple sub-selection points to the other adjacent coverage determination units, and simultaneously... The coverage determination unit where the sub-selection point is located is marked as a candidate clean state, and multiple candidate clean states are obtained based on multiple sub-selection points; the next coverage determination unit corresponding to the trajectory line of the underwater cleaning robot is collected in real time, and the candidate clean state corresponding to the next coverage determination unit is obtained as the target clean state based on multiple candidate clean states, and the other coverage determination units corresponding to the remaining candidate clean states are remarked as unclean states; the selection points on the other coverage determination units corresponding to the target clean state are copied and distributed, and the area composed of all coverage determination units corresponding to the trajectory line is taken as the clean area of ​​the working water area.

[0066] Specifically, selection points are set in each coverage determination unit. Before the next underwater cleaning robot's position information is transmitted to the 3D model data, the selection points are copied and differentiated into sub-selection points. Each selection point and sub-selection point belongs to a virtual control port. Multiple virtual control ports are interconnected for data transmission and control between multiple coverage determination units and the control terminal. They can store and forward messages, send control commands to other endpoints, and receive and send data and control instructions. Sub-selection points are distributed to other coverage determination units adjacent to the coverage determination unit corresponding to the current underwater cleaning robot's position information, serving as pre-selected landing points for the underwater cleaning robot's next move. Distributing sub-selection points to the corresponding coverage determination unit indicates that the coverage determination unit has been selected as a potential cleaned area, suggesting it may be a cleaned area. At this point, the cleaned area is rendered. The number of differentiated sub-selection points is the same as the number of pre-selected landing points for the next move. After preparation, the next coverage determination unit corresponding to the underwater cleaning robot's next position information is obtained, and the corresponding next landing point is extracted from multiple other coverage determination units. The other coverage determination units of the next coverage determination unit have already been marked as cleaned areas. Therefore, they can be directly visualized in subsequent rendering. At the same time, based on the communication connection between the sub-selection points, since the easily cleanable area corresponding to the next location information has been determined, it means that the other coverage determination units corresponding to the other sub-selection points do not correspond to the next coverage determination unit. Therefore, the markings in the other coverage determination units where the other sub-selection points are located are canceled, and they are restored to the uncleaned areas. Since the other coverage determination units have been confirmed before the location information of the next underwater cleaning robot is obtained, when the location information of the next underwater cleaning robot is obtained, the corresponding coverage determination units can be directly regarded as cleaned areas, and the others can be canceled. Since the process of marking as cleaned areas has been completed in advance, there is no need to mark the cleaned areas when obtaining the location information of the next underwater cleaning robot, which improves the efficiency of determining the cleaned areas of the working water area, thereby improving the efficiency of subsequent visualization rendering of the cleaned areas.

[0067] S5: Visually differentiates cleaned and uncleaned areas and calculates and displays the overall cleaning coverage in real time.

[0068] The steps for visually differentiating cleaned and uncleaned areas and calculating and displaying the overall cleaning coverage rate in real time include: acquiring real-time positioning data and coverage determination results of the underwater cleaning robot in the operating water area; based on the coverage determination results, rendering the cleaned area in a first visual style and rendering the uncleaned area in a second visual style that can be distinguished from the first visual style on the visualization interface to form a differentiated visual presentation; calculating the proportion of the currently cleaned area to the total cleanable area in real time based on the coverage determination results to obtain the overall cleaning coverage rate; and updating and displaying the overall cleaning coverage rate value together with the differentiated visual presentation interface in real time.

[0069] The system receives user instructions to select a historical time period and responds to the instructions by replaying the robot trajectory and cleaning area changes for the corresponding time period on the visualization interface, while dynamically displaying the coverage change curve within that historical time period; in response to user data export instructions, it exports data records including timestamps, robot positions, postures, and corresponding cleaning coverage rates into a structured data file.

[0070] Specifically, the visualization interface is a two-dimensional unfolded view; this view is formed by unfolding the bottom surface and multiple elevations of the working water area onto the same plane; the first and second visual styles use different color fills, pattern fills, or transparency; the visualization interface is a three-dimensional perspective view; in the three-dimensional water model, cleaned areas are highlighted by assigning a first color or first transparency to their corresponding three-dimensional voxels, while uncleaned areas retain the model's original color or a different second transparency. Between the cleaned voxels and uncleaned areas in the three-dimensional perspective view, there is also a gradient visual transition area, which can more smoothly display the cleaning boundary, improving visual effects and user experience; differentiated visual presentation uses a heatmap format; where color or color depth variations are used to represent coverage density, which is positively correlated with the frequency or dwell time of the underwater cleaning robot in the corresponding area; the overall cleaning coverage rate is displayed simultaneously in a prominent position on the visualization interface as a percentage progress bar and a numerical percentage, supporting free switching between 2D and 3D views to meet the observation needs of different scenarios and improve the adaptability of coverage visualization;

[0071] Based on the coverage determination results, the steps to calculate the proportion of the currently cleaned area to the total cleanable area in real time and obtain the overall cleaning coverage rate include: obtaining the total cleanable surface area of ​​the operating water area. Based on preset coverage determination rules, the total cleanable surface area is discretized into multiple grid cells of equal area, and the total number of grid cells marked as "clean" in real time is counted. Through formula Calculate the cumulative area of ​​the areas that have been marked as "cleaned". ,in, Represents the area of ​​a single grid cell; calculates and updates the overall cleaning coverage in real time. The corresponding formula is The overall cleaning coverage is displayed in real time on the visual interface using at least one of the following methods: numerical percentage and graphical progress bar.

[0072] By deploying a relatively low-cost optical sensor network above the pool and combining it with coded optical markers on the robot, real-time 3D positioning with sub-centimeter accuracy was achieved. Administrators can view a dynamically extending, precise movement trajectory in real time on a 2D or 3D model of the pool, improving the accuracy of underwater robot operations. Furthermore, the intuitive visualization interface enhances users' control over the cleaning process, enabling effective real-time monitoring and intervention of the underwater cleaning robot.

[0073] By pre-copying and distributing "sub-selection points" to adjacent potential next coverage units within the current cycle, the marking and pre-rendering of coverage areas are brought forward. Logically, this implements a parallel processing mode of "prediction-confirmation." When the real-time location data arrives, only a quick "confirmation" or "cancellation" operation is needed, instead of calculating and rendering from scratch. This significantly shortens the latency from data reception to visualization updates, marking potentially rendering coverage units as "pending" and pre-rendering them, essentially preparing the data and graphics instructions needed for the next frame in advance. When updates are needed, the cached pre-rendered results can be directly called, greatly reducing the computational and communication load required for each frame's rendering, thereby improving overall graphics rendering efficiency and significantly enhancing the efficiency and real-time performance of coverage area visualization.

[0074] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0075] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for visualizing the coverage of an underwater cleaning robot based on optical positioning, characterized in that, Includes the following steps: Multiple optical sensors are deployed above the operating water area to capture image sequences of optical markers mounted on the underwater cleaning robot. The current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area are determined based on the image sequence; Based on the current location coordinates, update and render the underwater cleaning robot's trajectory in the preset operating water area model; Pre-defined coverage determination rules are used to determine the cleaned areas of the work area based on the movement trajectory and coverage determination rules; The steps of the preset coverage determination rule include: Obtain three-dimensional model data of the work area to be cleaned; based on the three-dimensional model data, discretize the cleanable surface of the work area to obtain multiple coverage determination units, wherein the discretization process is to divide the cleanable surface into planar grid units of equal area, or to establish a three-dimensional voxel model on the cleanable surface of the three-dimensional model. A preset coverage determination rule is established, wherein the coverage determination rule is: a preset spatial neighborhood is established with the center of each coverage determination unit as the reference; when the positioning coordinates of the underwater cleaning robot fall into the preset spatial neighborhood of any coverage determination unit, the coverage determination unit is determined to have been cleaned. During the operation of the underwater cleaning robot, the cleaned area of ​​the working water area is determined based on its real-time status and coverage determination rules; The steps for determining the cleaned area of ​​the working water area based on its real-time status and coverage determination rules during the operation of the underwater cleaning robot include: Set selection points for each coverage determination unit; The current position coordinates of the underwater cleaning robot are obtained in real time, and the coverage determination unit corresponding to the current position coordinates in the 3D model data is used as the current coverage determination unit. The selected point corresponding to the current coverage determination unit is copied to obtain multiple sub-selected points, and communication connections are established between the selected point and the sub-selected points. Obtain other coverage determination units adjacent to the current coverage determination unit, distribute multiple sub-selection points to the other adjacent coverage determination units respectively, and mark the coverage determination unit where the sub-selection point is located as a candidate clean state. Based on multiple sub-selection points, obtain multiple candidate clean states. The system collects the next coverage determination unit corresponding to the trajectory line of the underwater cleaning robot in real time, obtains the candidate clean state corresponding to the next coverage determination unit based on multiple candidate clean states as the target clean state, and re-marks the other coverage determination units corresponding to the remaining candidate clean states as unclean states. The selected points on other coverage determination units corresponding to the target's clean status are copied and distributed, and the area composed of all coverage determination units on the trajectory line is taken as the cleaned area of ​​the operating water area. The system presents cleaned and uncleaned areas with distinct visual representations and calculates and displays the overall cleaning coverage rate in real time.

2. The method for visualizing the coverage of an underwater cleaning robot based on optical positioning according to claim 1, characterized in that: The step of deploying multiple optical sensors above the operating water area and capturing image sequences of optical markers mounted on the underwater cleaning robot based on the optical sensors includes: Multiple optical sensors are deployed above the operating area, and the optical sensors are configured to cover the entire operating area from different perspectives; Optical markers are placed on the top of the underwater cleaning robot. These optical markers are multiple active light-emitting or passive reflective elements arranged in a specific spatial pattern to form a unique code that can be used for identification. Multiple optical sensors are triggered synchronously to capture a synchronized sequence of images containing the optical markers.

3. The method for visualizing the coverage of an underwater cleaning robot based on optical positioning according to claim 1, characterized in that: The step of determining the current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area based on the image sequence includes: Acquire a sequence of images simultaneously captured by multiple optical sensors deployed above the operating water area, wherein the image sequence contains images of optical markers on the underwater cleaning robot; Optical markers in each frame of the image are identified and extracted to obtain their two-dimensional pixel coordinates in the pixel coordinate system of each optical sensor. Based on the pre-calibrated internal and external parameters of the multiple optical sensors, the coordinates of multiple two-dimensional pixels are fused using a multi-view visual triangulation algorithm to calculate the current position coordinates of the optical marker in the three-dimensional space of the working water area. The current position coordinates are output as the real-time pose information of the underwater cleaning robot.

4. The method for visualizing the coverage of an underwater cleaning robot based on optical positioning according to claim 1, characterized in that: The step of updating and rendering the underwater cleaning robot's trajectory in a preset operating water area model based on the current location coordinates includes: Obtain the real-time serialized current position coordinates of the underwater cleaning robot in the three-dimensional space of the working water area; The real-time serialized current position coordinates are spliced ​​with the historical position coordinates to form continuous motion trajectory data; In the preset 3D or 2D unfolded model of the working water area, a trajectory line representing the movement path of the underwater cleaning robot is dynamically rendered and updated based on the motion trajectory data.

5. The method for visualizing the coverage of an underwater cleaning robot based on optical positioning according to claim 1, characterized in that: The steps of visually differentiating cleaned and uncleaned areas and calculating and displaying the overall cleaning coverage in real time include: Acquire real-time positioning data and coverage determination results of the underwater cleaning robot in the operating area; Based on the coverage determination results, the cleaned area is rendered in the first visual style on the visualization interface, and the uncleaned area is rendered in the second visual style that can be distinguished from the first visual style, forming a differentiated visual presentation. Based on the coverage determination result, the proportion of the currently cleaned area to the total cleanable area is calculated in real time to obtain the overall cleaning coverage rate; the overall cleaning coverage rate value is updated and displayed in real time along with the differentiated visual presentation interface.

6. The method for visualizing the coverage of an underwater cleaning robot based on optical positioning according to claim 5, characterized in that: The step of calculating the proportion of the currently cleaned area to the total cleanable area in real time based on the coverage determination result to obtain the overall cleaning coverage rate includes: Obtain the total cleanable surface area of ​​the operating water area. Based on the preset coverage determination rules, the total cleanable surface area is discretized into multiple grid cells of equal area; Count the total number of grid cells marked as "clean" in real time. Through formula Calculate the cumulative area of ​​the areas that have been marked as "cleaned". ,in, Represents the area of ​​a single grid cell; Calculate and update overall cleaning coverage in real time. The corresponding formula is ; The overall cleaning coverage is displayed in real time on the visualization interface using at least one of the following methods: numerical percentage and graphical progress bar.

Images