Automatic pool cleaning device, control method, and computer storage medium

Abstract

The application provides a pool automatic cleaning device, a control method and a computer storage medium. The pool automatic cleaning device comprises an image acquisition device configured to acquire images. The method comprises: acquiring image information in front of the pool automatic cleaning device by the image acquisition device; identifying obstacles in the image information; and in the case that an obstacle is identified in front of the pool automatic cleaning device, controlling the pool automatic cleaning device to avoid the obstacle based on an obstacle avoidance strategy corresponding to a current operation mode of the pool automatic cleaning device.

Description

Technical Field

[0001] This application relates to the field of cleaning device technology, and in particular to an automatic pool cleaning device, control method, and computer storage medium. Background Technology

[0002] With the increasing popularity of swimming pools and the significant advancements in robotics technology, more and more consumers are opting for automated pool cleaning robots to perform pool cleaning tasks. However, the pool environment is complex and varied, containing various fixed and dynamic obstacles, such as leaves, ladders, wall lights, ground lights, and drains. How to promptly identify obstacles and accurately avoid them using different cleaning modes has become a pressing technical challenge for pool cleaning robots. Summary of the Invention

[0003] This application addresses the shortcomings of the prior art by providing a control method for an automatic water tank cleaning device. The automatic water tank cleaning device includes an image acquisition device for image acquisition. The method includes: acquiring image information in front of the automatic water tank cleaning device using the image acquisition device; identifying obstacles in the image information; and, when an obstacle is identified in front of the automatic water tank cleaning device, controlling the automatic water tank cleaning device to avoid obstacles based on an obstacle avoidance strategy corresponding to the current operating mode of the automatic water tank cleaning device.

[0004] Furthermore, when the current operating mode is a random water surface mode, controlling the automatic water cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic water cleaning device includes: determining the type of the obstacle, wherein, when the obstacle is a floating object to be cleaned, controlling the automatic water cleaning device to clean the floating object to be cleaned; and when the obstacle is a fixed obstacle, controlling the automatic water cleaning device to avoid the fixed obstacle.

[0005] Furthermore, controlling the automatic cleaning device of the pool to clean the floating objects to be cleaned includes: when there is only one floating object to be cleaned, determining the location information of the floating object to be cleaned, and cleaning the floating object to be cleaned based on the location information of the floating object to be cleaned; when there are multiple floating objects to be cleaned, determining the location information of each floating object to be cleaned, determining a target cleaning route based on the location information of each floating object to be cleaned, and cleaning the floating objects to be cleaned according to the target cleaning route.

[0006] Furthermore, when the current operating mode is the water surface edge mode, controlling the automatic cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic cleaning device includes: determining the type of obstacle, wherein, when the obstacle is a floating object to be cleaned, controlling the automatic cleaning device to clean the floating object to be cleaned, or controlling the automatic cleaning device to ignore the floating object to be cleaned and continue moving towards the front edge; when the obstacle is a fixed obstacle, controlling the automatic cleaning device to avoid the fixed obstacle.

[0007] Furthermore, the control method further includes: during the process of controlling the automatic pool cleaning device to avoid the fixed obstacle, determining whether the automatic pool cleaning device has crossed the obstacle, wherein if the automatic pool cleaning device has crossed the obstacle, controlling the automatic pool cleaning device to return to a position close to the pool wall and continue to operate in the water surface edge mode.

[0008] Furthermore, controlling the automatic water tank cleaning device to avoid the fixed obstacle includes: determining the distance between the automatic water tank cleaning device and the fixed obstacle; and controlling the automatic water tank cleaning device to avoid the fixed obstacle when the distance between the automatic water tank cleaning device and the fixed obstacle is a predetermined distance.

[0009] Furthermore, the automatic pool cleaning device includes a first sensor for detecting the distance between the automatic pool cleaning device and the fixed obstacle.

[0010] Furthermore, when the image acquisition device is a monocular image acquisition device, determining the distance between the automatic water tank cleaning device and the fixed obstacle includes: inputting the image information into a pre-trained monocular depth estimation model to obtain the depth information corresponding to the fixed obstacle output by the monocular depth estimation model; and determining the distance between the automatic water tank cleaning device and the fixed obstacle based on the depth information corresponding to the fixed obstacle.

[0011] This application also discloses an automatic water tank cleaning device, which can implement the method described in any embodiment of this application when executed.

[0012] This application also discloses a computer storage medium storing a computer program, which, when executed by a processor, implements the methods described in any embodiment of this application.

[0013] The embodiments described in this application have the following beneficial effects:

[0014] The control method for the automatic pool cleaning device provided in this application can acquire image information in front of the automatic pool cleaning device through its image acquisition device, and identify obstacles based on the image information. When an obstacle is identified in front of the automatic pool cleaning device, the method controls the automatic pool cleaning device to avoid the obstacle based on the obstacle avoidance strategy corresponding to the current operation mode of the automatic pool cleaning device. This ensures the normal progress of the cleaning operation and avoids collisions or friction between the automatic pool cleaning device and obstacles, thereby improving the cleaning efficiency of the pool and meeting the needs of users. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The drawings in the following description are merely exemplary embodiments of this application.

[0016] Figure 1 This is a flowchart illustrating a control method for an automatic water tank cleaning device according to an embodiment of this application. Detailed Implementation

[0017] The technical solutions in this application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0018] This application provides a control method for an automatic pool cleaning device, an automatic pool cleaning device using this control method, and a computer storage medium. The automatic pool cleaning device of this application is capable of cleaning a pool. The pool is, for example, a pool-shaped structure. The pool-shaped structure can be a swimming pool, a water storage tank, a spa pool, a water tank, a water storage trough, etc. The automatic pool cleaning device can be a device such as an automatic cleaning device or a pool cleaning robot, capable of cleaning the pool-shaped structure. This application does not limit the specific presentation of the automatic pool cleaning device or the pool-shaped structure, as long as the principle of this application is achieved. In the following description, unless otherwise specified, a robot will be used as an example of the automatic pool cleaning device, and a swimming pool will be used as an example of a pool or pool-shaped structure. In the following description, unless otherwise specified, the terms "pool bottom," "pool bottom surface," and "pool base" all refer to the bottom surface of a swimming pool.

[0019] The control method 100 of the automatic water tank cleaning device of this application will be described in detail below with reference to the accompanying drawings. Figure 1A flowchart illustrating a control method for an automatic water tank cleaning device according to an embodiment of this application is shown. The control method 100 includes steps S101 to S103. Steps S101 to S103 will be described below.

[0020] In step S101, image information in front of the automatic cleaning device of the water tank is acquired by the image acquisition device.

[0021] In one embodiment, during the cleaning operation, an image acquisition device (e.g., an image acquisition device positioned in front of the robot) can acquire image information in front of the robot, the image information including one or more reference objects.

[0022] For example, the robot can use a monocular and / or binocular and / or depth camera mounted on its top to capture image information of the area in front of it while performing a cleaning task. To improve environmental adaptability and target recognition accuracy, other sensors, including but not limited to high-definition visible light cameras, infrared thermal imaging cameras, and LiDAR, can be combined with the image acquisition equipment to capture image information of the area in front of the robot while performing a cleaning task. The acquired image information includes, but is not limited to, information such as the color, texture, size, outline, quantity, and relative position of the reference object.

[0023] It should be noted that after acquiring image information, image preprocessing operations can be performed to improve image quality, making subsequent detection and analysis more accurate and reliable. Image preprocessing operations include, but are not limited to: filtering image information, sharpening image information, removing noise from image information, enhancing image contrast, and normalizing data in image information.

[0024] It should be noted that the above description of the image information acquisition method, the content of the acquired image information, and the preprocessing operation of the image information is merely exemplary. The image information acquisition method, the content of the acquired image information, and the preprocessing operation of the image information protected by this application are not limited to the content listed above. Those skilled in the art can adjust the image information acquisition method and the content of the acquired image information according to the actual situation, and can also adopt one or more preprocessing operation methods to process the image information according to the actual situation, as long as the technical principle of this application can be realized.

[0025] Next, proceed to step S102. In step S102, obstacle identification is performed on the image information.

[0026] For safety reasons, as well as for cleaning, maintenance, recreation, training, competition, and natural factors, swimming pools often contain obstacles such as filters, pipes, ladders, ground lights, drains, wall lights, water pipes, and skimmers.

[0027] In one embodiment, obstacle recognition of the image information can be performed using deep learning models, including but not limited to: the YOLO series (e.g., YOLOv5, YOLOv7, YOLOv8, etc.), SSD, Faster R-CNN, etc. The obstacle recognition process will be described in detail below using the YOLO series as an example.

[0028] For example, YOLO divides the input image (e.g., the image information) into a fixed-size grid, with each grid responsible for detecting targets within its region (e.g., YOLOv1 divides the image into a 7×7 grid). Each grid predicts multiple bounding boxes, each containing an object. The bounding box is represented by its center coordinates (x, y), width (w), and height (h). For each bounding box, YOLO uses a convolutional neural network to extract features and performs classification through fully connected layers to predict the object's category (i.e., the obstacle category). Each bounding box also predicts a confidence score, representing the probability of an object being present within the bounding box and the accuracy of the bounding box. Simultaneously, YOLO uses Non-Maximum Suppression (NMS) to remove duplicate detections. NMS selects the bounding box with the highest confidence score as the final detection result and discards other bounding boxes with high overlap. Through these steps, deep learning models can be effectively used for obstacle recognition of image information, improving detection accuracy and real-time performance.

[0029] In another embodiment, obstacle recognition can also be performed on the image information using edge detection algorithms, including but not limited to Sobel and Canny algorithms. The following will use the Sobel algorithm as an example to explain the obstacle recognition process in detail.

[0030] For example, the Sobel operator uses two 3x3 convolution kernels to detect edges in the horizontal and vertical directions of an image, respectively. The horizontal convolution kernel is as follows:

[0031]

[0032] The vertical convolution kernel is:

[0033]

[0034] These two convolution kernels are applied to each pixel in the image information to calculate its horizontal direction (G). x ) and in the vertical direction (G y The gradient of a pixel is calculated. Specifically, for each pixel, the convolution kernel is convolved with the gray values ​​of that pixel and its eight surrounding pixels to obtain the gradient values ​​in the horizontal and vertical directions. Edges are determined by calculating the magnitude and direction of the horizontal and vertical gradients. The gradient magnitude G can be calculated using the following formula:

[0035]

[0036] The gradient direction θ can be calculated using the following formula:

[0037]

[0038] The gradient direction θ can help identify the regions with the most significant gray-level changes in an image, i.e., edges. Thresholding is performed based on the gradient magnitude, marking pixels with values ​​greater than a certain threshold as edge points. The edge image obtained through the Sobel operator can highlight the contours of objects in the image, facilitating the location and identification of obstacle contours.

[0039] It should be noted that after edge detection, other image processing techniques (such as morphological operations and region growing) can be combined to further extract and identify obstacles. For example, morphological operations can remove noise and small edge fragments, preserving larger, continuous edge regions, thereby more accurately identifying obstacles.

[0040] In another embodiment, obstacle recognition can also be performed on the image information using a threshold segmentation method.

[0041] For example, the acquired image information is converted into a grayscale image. Grayscale conversion simplifies image data, reduces computation, and preserves the main structural information of the image. Filters (e.g., Gaussian filtering, median filtering, etc.) are used to denoise the image to reduce the impact of noise on threshold segmentation. A global threshold is selected to classify pixels in the image into two classes. Common global thresholding methods include the Otsu method, which automatically selects the optimal threshold by maximizing the inter-class variance. Different thresholds are used in different regions of the image, suitable for situations where the grayscale difference between the target and background is uneven. The image is binarized according to the selected threshold, marking pixels with grayscale values ​​greater than the threshold as target regions (usually white) and pixels with grayscale values ​​less than the threshold as background regions (usually black). Obstacle regions are extracted from the binarized image; these regions are usually connected pixel blocks. Feature analysis is performed on the extracted regions, such as area, shape, and boundary features, to further confirm and identify obstacles.

[0042] It should be noted that the above description of the types of obstacles and the methods of obstacle identification is merely exemplary. The types of obstacles and the methods of obstacle identification protected by this application are not limited to those listed above. Those skilled in the art can adjust the types of obstacles and the methods of obstacle identification according to the actual situation, and can also use one or more methods of obstacle identification to process the image information, as long as the technical principles of this application can be realized.

[0043] Finally, proceed to step S103. In step S103, if an obstacle is detected in front of the automatic water tank cleaning device, the automatic water tank cleaning device is controlled to avoid the obstacle based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic water tank cleaning device.

[0044] The automatic pool cleaning device can have a variety of cleaning operation modes, including but not limited to the following: pool bottom cleaning mode, which focuses on thoroughly cleaning and vacuuming the bottom of the pool; water surface random mode, which allows the automatic pool cleaning device to move on the water surface in a random or relatively random path to remove floating objects and surface impurities; water surface along the edge mode, which allows the automatic pool cleaning device to move along the edge of the pool to ensure the cleanliness of the edge area; and pool wall cleaning mode, which is used to remove dirt and sediment attached to the side walls of the pool.

[0045] It's important to note that in the water surface random mode of the cleaning operation, the robot's movement direction on the water surface may be affected by various external environmental factors. Specifically, the speed and direction of the water flow exert forces and torques on the robot, causing it to deviate from its intended path. Furthermore, when the robot collides with pool walls or obstacles, the resulting reaction force also alters its direction of motion. Simultaneously, friction plays a significant role when the robot is in contact with the water surface. Although the coefficient of friction on the water surface is low, it still creates some resistance to the robot's movement, especially at low speeds, where the effect of friction is more pronounced. Friction between the robot and the pool walls also affects its movement direction on the water surface. In addition, wind direction also influences the robot's movement direction on the water surface, especially on open water surfaces, where wind can exert a horizontal thrust, causing it to deviate from its intended course. Therefore, the robot's movement direction on the water surface may be influenced by the combined effects of multiple external factors, which collectively determine its trajectory. In the water surface random mode, the robot does not actively control its yaw angle or steering.

[0046] It should be noted that in the water surface edge-following mode, the robot maintains a certain distance from the pool wall, and its movement direction is parallel, roughly parallel, or at a predetermined angle to the pool wall. This mode is typically used for cleaning the waterline area of ​​the pool. Furthermore, this mode is also widely used in the pool mapping process. By moving along the pool wall, the robot can effectively scan the edge contour of the pool, thereby generating an accurate map of the pool. This mapping process is crucial for subsequent cleaning path planning and navigation, helping the robot identify the pool boundaries and the location of obstacles.

[0047] In one embodiment, when the current operating mode is a random water surface mode, controlling the automatic water cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic water cleaning device includes: determining the type of the obstacle, wherein, when the obstacle is a floating object to be cleaned, controlling the automatic water cleaning device to clean the floating object to be cleaned; and when the obstacle is a fixed obstacle, controlling the automatic water cleaning device to avoid the fixed obstacle.

[0048] In the random mode on the water surface, when an obstacle is identified in front of the robot through step S102, different obstacle avoidance strategies are adopted for different types of obstacles (e.g., floating objects to be cleaned and fixed obstacles) to control the robot to avoid them.

[0049] In the random water surface mode, if the obstacle is a floating object to be cleaned (e.g., branches, leaves, plastic particles, paper scraps, etc.), controlling the automatic water cleaning device to clean the floating object includes: when there is only one floating object, determining the location information of the floating object and cleaning it based on the location information; when there are multiple floating objects, determining the location information of each floating object, determining a target cleaning route based on the location information of each floating object, and cleaning the floating objects according to the target cleaning route.

[0050] For example, in a water surface cleaning task, if the obstacle is identified as a single floating object to be cleaned in step S102, the robot's depth sensor system (e.g., ultrasonic sensor, infrared sensor, vision sensor, lidar, distance sensor, etc.) can be used to locate the single floating object on the water surface. Alternatively, a monocular depth estimation model can be used to locate the single floating object based on the image information in step S101, so as to accurately obtain the location information of the floating object to be cleaned. After obtaining the location information of the floating object to be cleaned, the robot adjusts its own trajectory and posture according to the location information to move towards the location of the floating object. Upon approaching the floating object, the robot activates its cleaning device (e.g., suction port, belt conveyor, rotating brush, etc.) to suck in and filter the floating object and place it in the waste collection container. The following will use a leaf as an example to illustrate this in detail.

[0051] The robot uses devices such as LiDAR (Light Detection and Ranging), depth sensors, Time-of-Flight (ToF) sensors, or vision sensors to acquire distance information between itself and the leaf. After determining the leaf's location, the robot plans its movement path, generating an optimal path from its current position to the leaf's location. Following the planned path, the robot adjusts its direction and speed to move towards the leaf. During this movement, the robot monitors its distance and relative position to the leaf in real time to ensure accurate arrival at the target location. Upon reaching the leaf, the robot uses cleaning devices (e.g., suction port, belt conveyor, rotating brush, etc.) to transport the leaf from the water surface to a waste collection container.

[0052] It's important to note that area array ToF sensors typically use infrared light or lasers as their light source, emitting modulated light signals. These signals are reflected back when they encounter the surface of objects such as leaves. The receiver in the sensor captures the reflected light and measures the time difference between the emission and reception of the light signal, i.e., the "time of flight." Based on the speed of light and the time of flight, the sensor calculates the distance to the object. Since the speed of light is constant, high-precision distance information can be obtained by accurately measuring the time difference.

[0053] For example, in a water surface cleaning task, if multiple floating objects to be cleaned are identified as obstacles in step S102, the robot will use a depth sensor system (e.g., ultrasonic sensors, infrared sensors, vision sensors, lidar, distance sensors, etc.) or a monocular depth estimation model to locate each floating object on the water surface based on the image information in step S101. This allows the robot to obtain the position coordinates, size, and distribution of each floating object, as well as the distance information between each object and the robot. Based on the obtained distance information, all floating objects to be cleaned are sorted according to their distance from the robot. Objects closer to the robot are placed first, and those farther away are placed later, thus forming a distance-based priority sequence. Based on this priority sequence, the robot's movement path (e.g., the target cleaning route) is planned sequentially. The robot adjusts its direction and speed according to the planned path. It first moves to the location of the nearest floating object to be cleaned and sucks it up and filters it through cleaning devices (such as suction port, belt conveyor, rotating brush, etc.). After completing the cleaning task of the nearest floating object, the robot moves to the location of other floating objects to be cleaned in turn according to the planned path and repeats the cleaning action.

[0054] It should be noted that throughout the entire cleaning process of the floating objects to be cleaned, the robot will monitor the status of the floating objects and changes in the surrounding environment in real time. If new floating objects to be cleaned are detected or the position of the floating objects to be cleaned changes, the robot will dynamically adjust the cleaning priority and path to ensure that all floating objects to be cleaned are effectively cleaned.

[0055] In one embodiment, devices such as LiDAR (LDS), depth sensors, Time-of-Flight (ToF) sensors, or vision sensors are used to scan and detect multiple leaves on the water surface, acquiring distance information between each leaf and the robot. The distance data acquired by the sensors is processed to calculate the actual distance of each leaf relative to the robot. Simultaneously, the position coordinates of each leaf are recorded to provide a basis for subsequent cleaning path planning. Based on the acquired distance information, all leaves are sorted according to their distance from the robot. Leaves that are closer are placed first, and leaves that are farther away are placed last, forming a distance-based priority sequence. In some cases, the priority may need to be fine-tuned. For example, if a distant leaf is located on the robot's path or its cleaning difficulty is low, its priority can be appropriately increased to optimize the cleaning path and efficiency. Based on the priority sequence and factors such as the robot's movement range, speed, and cleaning efficiency, one or more optimal paths are generated. The robot first moves to the location of the nearest leaf and uses cleaning devices (e.g., suction port, belt conveyor, rotating brush, etc.) to suck in and filter the leaf. During the cleaning process, the robot needs to precisely control the movements of the cleaning device to ensure that the leaves are effectively removed. After cleaning the leaves at close range, the robot moves to the locations of other leaves in sequence according to a planned path and repeats the cleaning action. For leaves that are farther away, the robot may need to adjust its movement speed and cleaning strategy to adapt to different distances and environmental conditions.

[0056] It's important to note that throughout the cleaning process, the robot monitors the condition of the leaves and changes in the surrounding environment in real time. If it detects new leaves or a leaf has moved, the robot dynamically adjusts its cleaning priority and path to ensure that all leaves are effectively cleaned.

[0057] In the random water surface mode, if the obstacle is a fixed obstacle, controlling the automatic water cleaning device to avoid the fixed obstacle includes: determining the distance between the automatic water cleaning device and the fixed obstacle; and controlling the automatic water cleaning device to avoid the fixed obstacle when the distance between the automatic water cleaning device and the fixed obstacle is a predetermined distance.

[0058] In one embodiment, the automatic pool cleaning device includes a first sensor for detecting the distance between the automatic pool cleaning device and the fixed obstacle.

[0059] For example, a robot may be equipped with a depth sensor (e.g., a first sensor), such as an ultrasonic sensor, infrared sensor, vision sensor, lidar, distance sensor, etc. The depth sensor continuously emits signals and receives reflected signals, and determines the distance to obstacles by calculating the round-trip time of the signals.

[0060] In another embodiment, when the image acquisition device is a monocular image acquisition device, determining the distance between the automatic water tank cleaning device and the fixed obstacle includes: inputting the image information into a pre-trained monocular depth estimation model to obtain the depth information corresponding to the fixed obstacle output by the monocular depth estimation model; and determining the distance between the automatic water tank cleaning device and the fixed obstacle based on the depth information corresponding to the fixed obstacle.

[0061] For example, if the image acquisition device is a monocular image acquisition device (e.g., a monocular camera), the image information in front of the robot acquired by the monocular image acquisition device in step S101 is input into a pre-trained monocular depth estimation model. The monocular depth estimation model, by learning from a large amount of image data, can predict the depth information of various objects in the scene from a single image. After processing the image information in front of the robot, the model outputs the depth information corresponding to the fixed obstacles. This depth information is typically presented in the form of pixel-level depth maps, with each pixel corresponding to a depth value. Based on the depth information output by the monocular depth estimation model, the actual distance between the robot and the fixed obstacles can be calculated.

[0062] It should be noted that, in order to improve the accuracy of distance measurement, the results of monocular depth estimation can be fused and calibrated with data from other sensors (e.g., ultrasonic sensors, infrared sensors, vision sensors, lidar, distance sensors, etc.).

[0063] It should be noted that the above description of determining the distance between the automatic pool cleaning device and the fixed obstacle is merely exemplary. The method of determining the distance between the automatic pool cleaning device and the fixed obstacle protected by this application is not limited to the content listed above. Those skilled in the art can adjust the method of determining the distance between the automatic pool cleaning device and the fixed obstacle according to the actual situation, as long as the technical principle of this application can be achieved.

[0064] After determining the distance between the robot and the fixed obstacle, it's possible to compare this distance with a predetermined distance threshold set in the control system. If the distance between the robot and the fixed obstacle is detected to be less than the predetermined threshold, it indicates that the robot is too close to the obstacle, requiring the triggering of an avoidance strategy to avoid the obstacle and prevent damage to the robot. Common strategies include, but are not limited to: stopping or slowing down, i.e., temporarily halting the robot's movement or reducing its speed to avoid a collision; and steering to avoid the obstacle, i.e., adjusting the robot's direction of travel by changing the speed or direction of the drive wheels to bypass the obstacle and continue moving forward.

[0065] It should be noted that throughout the entire avoidance process, the distance between the robot and the fixed obstacle needs to be continuously monitored to ensure the accuracy and safety of the avoidance maneuver.

[0066] In another embodiment, when the current operating mode is the water surface edge mode, controlling the automatic water cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic water cleaning device includes: determining the type of the obstacle, wherein, when the obstacle is a floating object to be cleaned, controlling the automatic water cleaning device to clean the floating object to be cleaned, or controlling the automatic water cleaning device to ignore the floating object to be cleaned and continue moving towards the front edge; when the obstacle is a fixed obstacle, controlling the automatic water cleaning device to avoid the fixed obstacle.

[0067] In the water surface edge mode, when an obstacle is identified in front of the robot in step S102, different obstacle avoidance strategies are adopted for different types of obstacles (e.g., floating objects to be cleaned and fixed obstacles) to control the robot to avoid them.

[0068] In the edge-cleaning mode, if the obstacle is a floating object to be cleaned (e.g., twigs, leaves, plastic particles, paper scraps, etc.), and the object is within the robot's current cleaning range, the object is transported from the water surface to the waste collection container via a cleaning device (e.g., suction port, belt conveyor, rotating brush, etc.). If the object is outside the robot's current cleaning range, or if cleaning it would affect the robot's edge-cleaning efficiency, the robot can choose to ignore the object and continue moving towards the edge. For example, if the object is a large twig or piece of plastic and is outside the robot's current cleaning range, or if cleaning it would require a significant adjustment to the robot's trajectory, the robot can ignore the object and continue moving towards the edge; that is, the robot maintains the edge-cleaning mode and continues moving along the edge of the pool.

[0069] It should be noted that for areas containing uncollected floating debris that are overlooked, the robot can mark them as uncleaned areas and address them in subsequent cleaning tasks. Alternatively, the robot can report this information to the monitoring system or operators so that other measures can be taken.

[0070] In the water surface edge mode, if the obstacle is a fixed obstacle, controlling the automatic pool cleaning device to avoid the fixed obstacle includes: determining the distance between the automatic pool cleaning device and the fixed obstacle; and controlling the automatic pool cleaning device to avoid the fixed obstacle when the distance between the automatic pool cleaning device and the fixed obstacle is a predetermined distance.

[0071] In one embodiment, the automatic pool cleaning device includes a first sensor for detecting the distance between the automatic pool cleaning device and the fixed obstacle.

[0072] For example, the robot can be equipped with a depth sensor (e.g., a first sensor), which can be an ultrasonic sensor, infrared sensor, vision sensor, lidar, distance sensor, etc. The depth sensor continuously emits signals and receives reflected signals, determining the distance to the obstacle by calculating the round-trip time of the signal. The automatic pool cleaning device can also determine the distance between itself and the fixed obstacle using a monocular image acquisition device combined with a pre-trained monocular depth estimation model. The depth sensor, the monocular image acquisition device, and the pre-trained monocular depth estimation model are described in the above embodiments and will not be repeated here.

[0073] After determining the distance between the robot and the fixed obstacle, it can be compared to a predetermined distance threshold set in the control system. If the distance between the robot and the fixed obstacle is detected to be less than the predetermined distance threshold, it indicates that the robot is too close to the obstacle and an avoidance strategy needs to be triggered to avoid the obstacle and prevent damage to the robot. Common strategies include, but are not limited to: stopping or decelerating, i.e., temporarily stopping the robot's forward movement or reducing its speed to avoid a collision with the obstacle; and steering to avoid the obstacle, i.e., adjusting the robot's direction of travel by changing the speed or direction of the drive wheels to bypass the obstacle and continue moving forward.

[0074] It should be noted that throughout the entire avoidance process, the distance between the robot and the fixed obstacle needs to be continuously monitored to ensure the accuracy and safety of the avoidance maneuver.

[0075] In another embodiment, the control method further includes: during the process of controlling the automatic pool cleaning device to avoid the fixed obstacle, determining whether the automatic pool cleaning device has crossed the obstacle, wherein if the automatic pool cleaning device has crossed the obstacle, the automatic pool cleaning device is controlled to return to a position close to the pool wall and continue to operate in the water surface edge mode.

[0076] For example, when controlling a robot to avoid fixed obstacles, sensors on the robot (e.g., inertial measurement unit, ultrasonic sensors, infrared sensors, vision sensors, lidar, distance sensors, etc.) are needed to continuously monitor its relative position to the obstacle. Real-time data acquired by the sensors is used to calculate the distance changes and orientation relationship between the robot and the obstacle during the avoidance process. Combining the robot's trajectory and avoidance strategy, it is analyzed whether the robot has successfully bypassed the obstacle. For example, if the robot uses a turning avoidance maneuver, the turning angle and the distance traveled after the turn are detected to determine if the obstacle avoidance action has been completed. Alternatively, sensor data can be fused with the device's built-in map or environmental model to comprehensively determine whether the robot has crossed the obstacle. For example, given a known pool layout and obstacle location, the robot's movement trajectory on the map can more accurately determine whether it has bypassed the obstacle. Once it is determined that the robot has crossed the obstacle, a return control procedure is initiated. Based on the distance and orientation between the robot's current position and the pool wall, a return path is planned, allowing the robot to smoothly move to a position close to the pool wall. After the robot returns to the vicinity of the pool wall, the water surface edge mode is restarted. Adjust the robot's direction and speed to continue cleaning along the edge of the pool. Simultaneously, ensure the robot maintains an appropriate distance from the pool wall to avoid collisions and improve cleaning efficiency. Continuously monitor the robot's status and environmental changes throughout the return and edge-cleaning process. If new obstacles or other abnormal situations are encountered, adjust the robot's movement strategy promptly to ensure it completes the cleaning task safely and efficiently.

[0077] The control method 100 for the automatic pool cleaning device provided in this application can acquire image information in front of the automatic pool cleaning device through its image acquisition device, and identify obstacles based on the image information. When an obstacle is identified in front of the automatic pool cleaning device, the method controls the automatic pool cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current working mode of the automatic pool cleaning device. This ensures the normal progress of the cleaning operation and avoids collisions or friction between the automatic pool cleaning device and obstacles, enabling the automatic pool cleaning device to complete the cleaning task efficiently, improving the cleaning efficiency of the pool, and meeting the needs of users.

[0078] This application also provides an automatic water tank cleaning device. The automatic water tank cleaning device is capable of performing the control methods described in the various embodiments above.

[0079] This embodiment discloses a computer storage medium storing a computer program, which, when executed by a processor, implements the control method described above.

[0080] It should be understood that, in this embodiment, the aforementioned computer storage medium may be located at at least one of the multiple network servers in a computer network. Optionally, in this embodiment, the aforementioned storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0081] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments.

[0082] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," 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 this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0083] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0084] In this application, unless otherwise stated, directional terms such as "up" and "down" are generally used in relation to the direction shown in the accompanying drawings, or in relation to the vertical, perpendicular, or gravitational direction; similarly, for ease of understanding and description, "left" and "right" are generally used in relation to the left and right shown in the accompanying drawings; "inner" and "outer" refer to the inner and outer contours of each component itself, but the above directional terms are not intended to limit this application.

[0085] The above description is merely an exemplary embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope described in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A control method for an automatic water tank cleaning device, the automatic water tank cleaning device including an image acquisition device, the image acquisition device being used for image acquisition, the method comprising: The image acquisition device acquires image information of the area in front of the automatic cleaning device for the water tank. Obstacle identification is performed on the image information; When an obstacle is detected in front of the automatic pool cleaning device, the device is controlled to avoid the obstacle based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic pool cleaning device.

2. The control method according to claim 1, wherein when the current operating mode is a random water surface mode, controlling the automatic water tank cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic water tank cleaning device includes: Determine the type of the obstacle, wherein, When the obstacle is a floating object to be cleaned, the automatic cleaning device of the pool is controlled to clean the floating object. When the obstacle is a fixed obstacle, the automatic cleaning device for the pool is controlled to avoid the fixed obstacle.

3. The control method according to claim 2, wherein controlling the automatic cleaning device of the pool to clean the floating objects to be cleaned includes: When there is only one floating object to be cleaned, determine the location information of the floating object to be cleaned, and clean the floating object based on the location information of the floating object to be cleaned; When there are multiple floating objects to be cleaned, the location information of each floating object to be cleaned is determined, a target cleaning route is determined based on the location information of each floating object to be cleaned, and the floating objects to be cleaned are cleaned according to the target cleaning route.

4. The control method according to claim 1, wherein when the current operating mode is the water surface edge mode, controlling the automatic water tank cleaning device to avoid obstacles based on the obstacle avoidance strategy corresponding to the current operating mode of the automatic water tank cleaning device includes: Determine the type of the obstacle, wherein, When the obstacle is a floating object to be cleaned, the automatic cleaning device of the pool is controlled to clean the floating object, or the automatic cleaning device of the pool is controlled to ignore the floating object and continue to move towards the front edge. When the obstacle is a fixed obstacle, the automatic cleaning device for the pool is controlled to avoid the fixed obstacle.

5. The control method according to claim 4 further includes: During the process of controlling the automatic water tank cleaning device to avoid the fixed obstacle, it is determined whether the automatic water tank cleaning device has passed the obstacle. If the automatic pool cleaning device passes the obstacle, control the automatic pool cleaning device to return to a position close to the pool wall and continue operating in the water surface edge mode.

6. The control method according to claim 2 or 4, wherein controlling the automatic cleaning device of the pool to avoid the fixed obstacle includes: Determine the distance between the automatic water tank cleaning device and the fixed obstacle; When the distance between the automatic water tank cleaning device and the fixed obstacle is a predetermined distance, the automatic water tank cleaning device is controlled to avoid the fixed obstacle.

7. The control method according to claim 6, wherein the automatic water tank cleaning device includes a first sensor, the first sensor being used to detect the distance between the automatic water tank cleaning device and the fixed obstacle.

8. The control method according to claim 6, wherein when the image acquisition device is a monocular image acquisition device, determining the distance between the automatic water tank cleaning device and the fixed obstacle includes: The image information is input into a pre-trained monocular depth estimation model to obtain the depth information corresponding to the fixed obstacle output by the monocular depth estimation model. The distance between the automatic water tank cleaning device and the fixed obstacle is determined based on the depth information corresponding to the fixed obstacle.

9. An automatic water tank cleaning device, wherein, The automatic water tank cleaning device is capable of performing the control method according to any one of claims 1-8.

10. A computer storage medium storing a computer program that, when executed by a processor, implements the method of any one of claims 1-8.

Images