Automatic pool cleaning device, control method and computer storage medium

Abstract

A control method of an automatic pool cleaning device (20), the automatic pool cleaning device (20) comprising an image acquisition assembly (201), the control method comprising: controlling the automatic pool cleaning device (20) to move in a pool, and during the moving process, acquiring an image of an object around the automatic pool cleaning device (20) by the image acquisition assembly (201); recognizing steps and / or platforms in the pool based on the image; and controlling the automatic pool cleaning device (20) to execute a first cleaning mode on the steps and / or a second cleaning mode on the platform according to a recognition result of the image.

Description

Technical Field

[0001] This application relates to the technical field of automatic water tank cleaning devices, and more particularly to an automatic water tank cleaning device, control method, and computer storage medium. Background Technology

[0002] With the increasing popularity and diverse shapes of swimming pools, cleaning them has become a crucial aspect of their use. However, current automatic pool cleaning systems cannot effectively distinguish complex terrain such as steps or platforms within the pool. This may lead to the automatic pool cleaning system using incorrect cleaning strategies for steps or platforms, resulting in issues such as getting stuck when crossing steps or having to repeatedly clean on platforms, thus reducing the cleaning efficiency and safety of the automatic pool cleaning system. 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 component. The control method includes: controlling the automatic water tank cleaning device to move in a water tank, and during the movement, acquiring images of objects around the automatic water tank cleaning device through the image acquisition component; identifying steps and / or platforms in the water tank based on the images; and, according to the image recognition results, controlling the automatic water tank cleaning device to perform a first cleaning mode on the steps and / or a second cleaning mode on the platforms.

[0004] This application also provides a non-volatile computer storage medium storing a computer program that, when executed by a processor, implements the method described in any one of the above descriptions.

[0005] This application also provides an automatic water tank cleaning device, which is capable of performing the method described in any one of the above.

[0006] The embodiments described in this application have the following beneficial effects:

[0007] The control method of the automatic pool cleaning device provided in this application enables the automatic pool cleaning device to accurately distinguish between steps or platforms, and thus to adopt corresponding cleaning strategies for steps or platforms, thereby improving the cleaning efficiency of the automatic pool cleaning device and ensuring the safety of the automatic pool cleaning device during cleaning operations. Attached Figure Description

[0008] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings used in the description of the embodiments will be briefly introduced below. The accompanying drawings in the following description are merely exemplary embodiments of this disclosure.

[0009] Figure 1A flowchart illustrating the control method of the automatic water tank cleaning device of this application is shown; and

[0010] Figure 2 A schematic diagram of the automatic water tank cleaning device of this application located on a step is shown.

[0011] Label Explanation

[0012] 20. Automatic water tank cleaning device; 201. Image acquisition component. Detailed Implementation

[0013] The embodiments of this disclosure will now be described with reference to the accompanying drawings. The described embodiments are merely some, not all, of the embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the protection scope of this application. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0014] This application provides a control method 100 for an automatic water tank cleaning device 20. Figure 1 A flowchart illustrating the control method of the automatic water tank cleaning device of this application is shown. Figure 2 A schematic diagram of the automatic pool cleaning device of this application located on steps is shown. The automatic pool cleaning device 20 can be, for example, a cleaning device such as an automatic pool cleaning robot or an automatic pool sweeping robot. (See reference...) Figure 2 The automatic water tank cleaning device 20 includes an image acquisition component 201.

[0015] Image acquisition component 201 may include, for example, a binocular camera. A binocular camera typically consists of two monocular cameras with identical parameters, maintaining a fixed distance between the lenses. The binocular cameras simultaneously capture images of the same scene, acquiring two sets of binocular images with parallax. This allows for direct measurement of obstacle distances using the binocular images, resulting in highly accurate distance data. Image acquisition component 201 is used to acquire image data from the pool environment to achieve functions such as obstacle recognition, path planning, and cleaning target detection. The camera can be one or more of the following types depending on actual needs: visible light camera (RGB camera), used to acquire color images under sufficient lighting conditions; low light camera (such as a high-sensitivity sensor camera), suitable for image acquisition in low-light environments; infrared camera (IR camera), combined with infrared supplementary lighting, can work in turbid water or low visibility conditions; underwater-specific camera, with waterproof, pressure-resistant, and corrosion-resistant characteristics, suitable for long-term underwater operation; 3D camera or depth camera (such as structured light or ToF camera), used to acquire three-dimensional spatial information of obstacles; multispectral or polarization camera, used to enhance image recognition capabilities in specific scenarios (such as stain classification or reflection suppression); panoramic camera, providing a wider field of view to improve environmental perception capabilities. The cameras mentioned above are only a limited list; other types of cameras can be used in practice, as long as they can realize the technical concept of this application. It should be noted that the camera used should have a sealed waterproof function to ensure that it can work stably when submerged in water for a long time. The camera can be set at the front of the robot (e.g., as needed). Figure 2 (as shown), upper, side and / or rear positions, to optimize image acquisition needs in different directions.

[0016] Image acquisition component 201 may include, for example, a lidar (e.g., a laser diode scanning radar). The lidar can be used to detect or range targets around the robot. The lidar can also be mounted on a rotatable base, allowing it to scan and identify the area around the automatic pool cleaning device 20 during rotation.

[0017] Image acquisition component 201 may include, for example, an ultrasonic sensor. The ultrasonic sensor emits ultrasonic waves in the direction of travel of the automatic pool cleaning device 20 and receives the reflected echoes. Based on the data acquired by the ultrasonic sensor, the automatic pool cleaning device 20 can analyze the data using artificial intelligence software to further generate an image of the target object.

[0018] The following reference Figure 1The control method 100 of this application will be described below. The control method 100 includes: in step S101, controlling the automatic water tank cleaning device 20 to move in the water tank, and during the movement, acquiring images of objects around the automatic water tank cleaning device 20 through the image acquisition component 201; in step S102, identifying steps and / or platforms in the water tank based on the images; in step S103, controlling the automatic water tank cleaning device 20 to perform a first cleaning mode on the steps and / or a second cleaning mode on the platforms according to the recognition results of the images.

[0019] First, proceed to step S101. In step S101, the automatic water cleaning device 20 is controlled to move within the water pool, such as at the bottom or wall. During this movement, the automatic water cleaning device 20 can be powered and its direction adjusted via drive wheels, tracks, water pumps, or other devices on its body. During movement, the image acquisition component 201 acquires images of objects surrounding the automatic water cleaning device 20. The automatic water cleaning device 20 may be equipped with a controller, which may contain control circuitry such as a microprocessor, digital signal processor (DSP), or microcontroller. During the movement of the automatic water cleaning device 20, the image acquisition component 201 may acquire one or more frames of images at predetermined time intervals, and the controller will then identify the acquired images.

[0020] Next, proceed to step S102. In step S102, steps and / or platforms in the pool are identified based on the image. In the case where the image acquisition component 201 includes a binocular camera, in order to accurately identify steps and platforms in the pool, the controller may, for example, process the image acquired by the image acquisition component 201 as follows.

[0021] The controller first performs stereo correction on the images of the same scene captured by the binocular cameras. Stereo correction involves reprojecting two images captured from different perspectives (hereinafter referred to as "left image" and "right image") onto the same plane, aligning the coordinates of the two images. After correction, corresponding feature points in the left and right images will appear on the same vertical coordinate. Specifically, stereo correction calculates a set of precise perspective transformation matrices to geometrically deform the two images, eliminating radial and tangential distortions from the lens and restoring the ideal perspective projection of the images. This provides a reliable guarantee for subsequent generation of high-precision disparity maps and 3D point cloud reconstruction.

[0022] Furthermore, a disparity map is generated based on the stereo-corrected image described above. For example, a stereo matching algorithm (such as the StereoSGBM algorithm) can be used to analyze the pixel position differences of the same physical point on two imaging planes with different spatial locations to quantitatively estimate the depth information of the scene in the acquired image. Here, disparity refers to the difference in horizontal coordinates of the same point in the images acquired by the two cameras. Those skilled in the art can also apply other disparity map generation algorithms to the technical solution of this application, as long as they can achieve the technical principles of this application.

[0023] Furthermore, the initial disparity map can be depth-corrected to improve its accuracy. For example, median filtering can be used to identify pixels with low confidence or invalidity in the initial disparity map, and reasonable inferences and replacements can be made using the disparity values ​​of surrounding valid pixels. Edge-aware filtering and optimization can also be applied to the initial disparity map.

[0024] Furthermore, the disparity map that has undergone depth correction can also be used to calculate a depth map. For example, using the formula... The calculation is performed. Here, Z is the depth from a point in space to the camera plane, f is the focal length of the camera, B is the baseline distance between the optical centers of the two cameras, and d is the parallax of a point in space in the left and right images.

[0025] Finally, the disparity map or depth map is used to reconstruct a 3D point cloud. For example, OpenCV functions for quickly calculating 3D point clouds can be called to calculate the coordinates of each pixel in 3D space, thus obtaining the 3D point cloud. Alternatively, the pixels in the depth map can be projected back into 3D space, i.e., the 3D coordinates of the pixels in the coordinate system can be calculated using the intrinsic parameter matrix.

[0026] The above description of the steps for realizing 3D point cloud reconstruction is not an exhaustive list. Any solution that can realize the technical principles of this application falls within the protection scope of this application.

[0027] The controller can, for example, distinguish between steps and platforms based on the precise information from the image obtained in the above steps. For instance, semantic segmentation of each pixel based on deep learning can directly identify steps and platforms.

[0028] In step S102, identifying steps and / or platforms in the pool based on the image includes: identifying horizontal and / or vertical information of objects in the pool based on the image, wherein the horizontal information includes at least one of the following: the number of horizontal surfaces, the area of ​​the horizontal surfaces, and the height difference between adjacent horizontal surfaces; and the vertical information includes the number of vertical surfaces. Based on the horizontal and / or vertical information, steps and / or platforms in the pool are identified.

[0029] For example, the controller can also classify steps and platforms based on horizontal and / or vertical information of objects in a pool. First, for example, the RANSAC algorithm can be used to randomly sample the point cloud and fit it into a plane. The angle between the normal vector and the gravity vector of the fitted plane is calculated. If the angle is less than a predetermined angle threshold (e.g., 10 degrees), it is classified as a horizontal candidate; if the angle is less than 90 degrees minus the predetermined angle threshold, it is classified as a vertical candidate. These candidate planes are then clustered according to the distance between the normal direction and the normal vector to reduce plane redundancy. For example, the RANSAC algorithm may fit a complete plane into multiple smaller planes due to noise interference, fitting errors, etc. Normal clustering can re-identify the same physical plane that has been divided into multiple smaller planes and merge them into a complete plane, thereby simplifying the plane data and improving the controller's processing efficiency for plane data. Classifying planes according to vertical and horizontal planes makes the controller's semantic understanding of the plane clearer. Record the area, number, and average height of each plane after clustering, and sort all horizontal planes from low to high. Record the height difference between each layer. By sorting by height, the spatial hierarchy between planes can be inferred based on their positional relationship, thereby identifying steps and platforms.

[0030] At least one of the planar information obtained above can be used to identify steps and / or platforms in the pool.

[0031] For example, if the number of horizontal surfaces is greater than a predetermined threshold, the controller determines that the object in the image is a step, otherwise it is a platform.

[0032] For example, if the area of ​​the horizontal plane is greater than a predetermined area threshold, the controller determines that the object in the image is a platform, and otherwise it is a step.

[0033] For example, if the number of vertical surfaces is greater than a predetermined threshold, the controller determines that the object in the image is a step, otherwise it is a platform.

[0034] The above information can also be combined to identify steps or platforms. If the number of horizontal and vertical surfaces is equal and both exceed a predetermined threshold, the controller determines that the object in the image is a step. This further improves the accuracy of the automatic pool cleaning device 20 in identifying steps and platforms.

[0035] The above-described methods for selecting horizontal and vertical planes are merely illustrative examples. Any technical solution that can achieve the technical principles of this application falls within the protection scope of this application.

[0036] Based on the horizontal and / or vertical information, steps and / or platforms in the pool are identified, including: determining the object as a step when the number of horizontal surfaces is greater than 1, the area of ​​the horizontal surfaces is less than a preset area threshold, the height difference between adjacent horizontal surfaces is greater than a preset height difference, or the number of vertical surfaces is greater than 1; and determining the object as a platform when the number of horizontal surfaces is equal to 1, the area of ​​the horizontal surfaces is greater than a preset area threshold, or the number of vertical surfaces is equal to 1.

[0037] For example, the judgment logic for steps is as follows: The core characteristic of steps is that they consist of a continuous series of multiple treads (i.e., horizontal planes), therefore the number of horizontal planes of the object must be greater than one. Each tread of a step has a corresponding vertical plane, therefore the number of vertical planes of the object must also be greater than one. The area of ​​a step tread is usually small, for example, usually smaller than the area of ​​a platform, therefore the planar area of ​​the object is less than a preset area threshold. There is a significant height difference between each tread of a step, therefore the height difference between adjacent horizontal planes must be greater than a preset height difference (usually 15-25cm).

[0038] For example, the judgment logic for a platform is as follows. There is usually only one platform, and the area is relatively large. Therefore, the number of horizontal surfaces of the object must be equal to 1, and the corresponding number of vertical surfaces must also be equal to 1. The area of ​​the horizontal surface must be greater than a preset area threshold.

[0039] Generally, steps and platforms do not share the same or similar characteristics except for their height difference. Therefore, only one condition in the above-mentioned judgment logic, excluding the height difference, needs to be met to distinguish between steps and platforms. Those skilled in the art can also set specific judgment logic and conditions to be met according to actual circumstances; this application does not limit this. Those skilled in the art can also set judgment logic for steps and platforms separately based on the technical principles of this application.

[0040] Next, proceed to step S103. In step S103, based on the recognition result of the image, control the automatic water tank cleaning device 20 to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform.

[0041] Based on the above control steps, the automatic pool cleaning device 20 can accurately distinguish whether the object in front is a step or a platform, and implement the corresponding cleaning method. Specifically, the automatic pool cleaning device 20 performs a first cleaning mode for steps and a second cleaning mode for platforms. The first cleaning mode differs from the second cleaning mode. For example, the travel speed of the automatic pool cleaning device 20 in the second cleaning mode can be greater than the travel speed in the first cleaning mode; or, for the platform, the second cleaning mode can clean the platform surface and side walls separately, while for the step, the first cleaning mode can clean the step surface and side walls together. The specific cleaning methods will be described in detail below.

[0042] In step S103, based on the recognition result of the image, the automatic cleaning device 20 of the pool is controlled to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform, including: when the steps and / or platform in the pool are recognized based on the image, the automatic cleaning device 20 of the pool is controlled to move towards the steps or the platform, and the first cleaning mode is performed on the steps or the second cleaning mode is performed on the platform.

[0043] After the automatic water tank cleaning device 20 identifies an object as a step and / or platform, it controls the device to move towards the step or platform. This movement can be a planned path along which the device moves towards the step or platform. The path to the base station can be obtained using a path planning algorithm. For example, the path can be the shortest straight-line distance between the device and the step or platform from its current position. Alternatively, the movement can be a predetermined direction towards the step or platform. During this movement, the device continuously acquires images of the step or platform using the image acquisition component 201, thereby obtaining point cloud data. Specifically, during the movement towards the step or platform, the device acquires overall point cloud data of the step or platform using the image acquisition component 201. The automatic water tank cleaning device 20 can plan or adjust its movement parameters (such as movement path, heading angle, or movement speed) based on the overall point cloud data of the steps or platform, causing the automatic water tank cleaning device 20 to move towards the steps or platform. Once the automatic water tank cleaning device 20 has moved to the platform or steps, it begins to execute either the first cleaning mode or the second cleaning mode.

[0044] In one scenario, a platform and a step are present in the image simultaneously. The automatic pool cleaning device 20 can first move toward the platform or step closest to it and clean it, or it can move according to the control logic preset by those skilled in the art. This application does not limit this.

[0045] In step S103, based on the image recognition result, the automatic pool cleaning device 20 is controlled to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform, including: recording the position information of the steps and / or platform based on the image recognition result; and controlling the automatic pool cleaning device 20 to perform the first cleaning mode on the steps and / or the second cleaning mode on the platform based on the recorded position information.

[0046] When the automatic pool cleaning equipment is mapping or cleaning the pool bottom or walls, it can identify steps and / or platforms by acquiring images through the image acquisition component. The controller can record the position information of the steps or platforms based on the images, including information such as the shape, size, height, confidence level, and coordinates of the steps or platforms. After the automatic pool cleaning equipment finishes mapping, completes the current cleaning task, or receives a task to clean the platform or steps, it controls the automatic pool cleaning device 20 to move to the location of the steps or platforms based on the recorded position information and executes a first cleaning mode for the steps or a second cleaning mode for the platforms.

[0047] Specifically, the automatic pool cleaning device 20 can set specific parameters (such as cleaning time, distance, etc.) in a first or second cleaning mode based on the shape, size, or height information of the steps or platform, and clean the steps or platform according to the specific parameters. The automatic pool cleaning device 20 can also select different movement methods based on the level of confidence. For example, in areas with high confidence, the automatic pool cleaning device 20 can adopt a faster and more decisive cleaning method to clean the steps or platform. The automatic pool cleaning device 20 can also move towards the platform or steps based on the coordinate information of the steps or platform, as described above, which will not be elaborated further here.

[0048] After identifying a step or platform, its location is recorded so that the robot can accurately locate and clean it. Simultaneously, when the robot performs other cleaning tasks, such as cleaning pool walls or bottoms, it can avoid these steps or platforms based on the recorded location information. This ensures that the robot's current cleaning parameters are not suitable for cleaning steps or pool walls, thus maintaining cleaning efficiency. After other cleaning tasks are completed, the robot can then clean the steps or platforms again based on the recorded location information.

[0049] The first cleaning mode includes controlling the automatic cleaning device 20 of the pool to clean each step sequentially along the length of the steps.

[0050] In the first cleaning mode, the automatic pool cleaning device 20 can, for example, clean each step from bottom to top along the length of the steps.

[0051] In one scenario, the automatic pool cleaning device 20 first climbs up the vertical surface of the first step using a drive mechanism such as tracks and drive wheels. For example, as... Figure 2 As shown, the automatic cleaning device 20 of the pool is tilted at a certain angle relative to the vertical direction by the difference in wheel speed between the tracks or drive wheels on both sides, and moves along the length of the step. Figure 2 The robot is controlled to move to the right, thus completing the cleaning of the vertical surface. Afterwards, the automatic water tank cleaning device 20 climbs onto the horizontal surface of the step, continuing to clean this surface from left to right or right to left along the length of the step. After cleaning this horizontal surface, the robot climbs up to the vertical surface of the next step and repeats the cleaning steps for both the vertical and horizontal surfaces described above. Generally, the width of the horizontal surface of a step is relatively small. When cleaning the steps, the robot's head faces upward and its tail faces downward, as shown in the image. Figure 2 As shown, the machine lies on the steps and moves left and right to clean them. Of course, if the width of the steps is relatively large, the entire machine can be moved to the platform and the head can be facing the length of the steps, moving left and right to clean the platform.

[0052] The tilt angle of the automatic pool cleaning device 20 can be set according to the size of the machine or the steps (for example, the lower the height of the steps, the larger the tilt angle of the automatic pool cleaning device 20 can be). The cleaning method of the automatic pool cleaning device 20 on the steps is not limited to the above description. Any cleaning method that can realize the technical principle of this application is within the protection scope of this application.

[0053] When the automatic water tank cleaning device 20 cleans each step, at least a portion of the cleaning component of the automatic water tank cleaning device 20 covers the area where the horizontal and vertical surfaces of the step meet.

[0054] The automatic pool cleaning device 20 can, for example, clean dirt in the pool using cleaning components (e.g., roller brushes) located at the bottom or side. When the automatic pool cleaning device 20 cleans steps, the cleaning components also need to clean the area where the horizontal and vertical surfaces meet (e.g., the angle between the vertical surface of the step and the horizontal surface of the step, the edges of the step). For example, when the automatic pool cleaning device 20 moves along the length of the step to clean the vertical or horizontal surface of the step, the machine head faces upwards, the machine body lies on the step, and a portion of the cleaning components on the machine contacts and cleans the area where the horizontal and vertical surfaces meet. Also, for example, when the automatic pool cleaning device 20 moves from a vertical surface to a horizontal surface, or from a horizontal surface to a vertical surface, all or at least a portion of the cleaning components contacts and cleans the area where the horizontal and vertical surfaces meet. The above control method can effectively clean areas such as the edges and corners of the steps that are difficult to clean in conventional ways, thus improving the cleaning effect of the automatic cleaning device for the pool on the steps.

[0055] When the automatic pool cleaning device 20 is cleaning the steps, the method further includes: based on the distance value detected by the downward-looking sensor on the automatic pool cleaning device 20, when the automatic pool cleaning device 20 moves from one step to another, controlling the automatic pool cleaning device 20 to reduce its moving speed.

[0056] The automatic pool cleaning device 20 may include a downward-facing sensor, which may be located at the bottom of the device. For example, the downward-facing sensor may contain an infrared emitting diode that emits an infrared beam, and an infrared receiving diode that receives reflected infrared light. The controller can calculate the distance based on the time difference between the emission and reception of the infrared light.

[0057] In one scenario, the downward-facing sensor is positioned at the bottom of the automatic pool cleaning device 20 and near its head. When the device moves from one step to another (e.g., from the horizontal plane of one step to the horizontal plane of another), the controller calculates the changes in the downward-facing sensor reading. If the reading shows a pattern of "first increasing, then decreasing," it indicates that the device has crossed to another step. The controller then reduces the device's speed to prevent it from colliding with the side wall (vertical plane) of the next step due to excessive speed, thus avoiding damage.

[0058] This control method improves the safety of the automatic water tank cleaning device 20. The description of the corresponding downward-looking sensors is not an exhaustive list. Any solution that can achieve the technical principle of this application falls within the protection scope of this application.

[0059] The second cleaning mode includes controlling the automatic cleaning device 20 of the pool to clean the horizontal and vertical surfaces of the platform along a predetermined path.

[0060] The automatic pool cleaning device 20 first moves vertically towards the platform. After reaching the platform, it can clean the platform according to a predetermined path. For example, it can first clean the platform's side walls (vertical surfaces), and then clean the horizontal surfaces after cleaning the side walls. Alternatively, it can clean the platform's side walls and pool walls together, and then clean the platform surface separately. Or, it can clean the vertical and horizontal surfaces of the platform alternately, depending on the set cleaning path. The predetermined path may include a pre-designed cleaning path or cleaning strategy, which is not specifically limited in this embodiment.

[0061] It should be noted that in practice, terms such as "predetermined path" and "bow-shaped path" do not necessarily require the robot to pre-plan a movement trajectory and store the corresponding information in the robot's memory. In this field, a "path" typically refers to a planned movement rule, which can be a trajectory planned according to a global map. This movement rule can also be a movement rule that controls the robot's movement, such as a bow-shaped path movement rule. For example, controlling the robot to move continuously in a predetermined direction (i.e., along the long side of the bow), turning 90 degrees to the right upon encountering an obstacle, then moving forward a preset distance or for a preset time (i.e., along the short side of the bow), turning 90 degrees to the right again, and so on.

[0062] The above description of the preset path is merely an example. Those skilled in the art can set the preset path according to the actual situation, and this application does not limit it.

[0063] This application also includes a non-volatile computer storage medium storing a computer program that, when executed by a processor, implements the method described in any one of the above descriptions.

[0064] It should be understood that the non-volatile computer storage medium may be located on at least one of the multiple network servers in a computer network. Optionally, in this application, 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.

[0065] This application also includes an automatic pool cleaning device 20, which is capable of performing any of the methods described above.

[0066] The automatic pool cleaning device 20 can be, for example, a pool cleaning robot or a pool sweeping robot. The automatic pool cleaning device 20 can execute the control program described above, which has been explained in detail above and will not be repeated here.

[0067] The control method of the automatic pool cleaning device 20 provided in this application enables the automatic pool cleaning device 20 to accurately distinguish between steps or platforms, and thus adopt corresponding cleaning strategies for steps or platforms, thereby improving the cleaning efficiency of the automatic pool cleaning device 20 and ensuring the safety of the automatic pool cleaning device 20 during cleaning operations.

[0068] 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.

[0069] 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 expressly and specifically defined.

[0070] 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.

[0071] 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 (20), the automatic water tank cleaning device (20) including an image acquisition component (201), the control method comprising: The automatic water tank cleaning device (20) is controlled to move in the water tank, and during the movement, images of objects around the automatic water tank cleaning device (20) are acquired by the image acquisition component (201). Based on the image, steps and / or platforms in the pool were identified; Based on the recognition results of the image, the automatic cleaning device (20) of the pool is controlled to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform.

2. The control method according to claim 1, wherein, Identifying steps and / or platforms in the pool based on the image includes: Based on the image, the horizontal and / or vertical information of objects in the pool is identified, wherein the horizontal information includes at least one of the following: the number of horizontal surfaces, the area of ​​the horizontal surfaces, and the height difference between adjacent horizontal surfaces; and the vertical information includes the number of vertical surfaces. Based on the horizontal and / or vertical information, steps and / or platforms in the pool are identified.

3. The control method according to claim 2, wherein, Based on the horizontal and / or vertical information, steps and / or platforms in the pool are identified, including: If the number of horizontal surfaces of an object is greater than 1, the area of ​​the horizontal surfaces is less than a preset area threshold, the height difference between adjacent horizontal surfaces is greater than a preset height difference, or the number of vertical surfaces is greater than 1, the object is determined to be a step. If the number of horizontal surfaces of an object is equal to 1, the area of ​​the horizontal surfaces is greater than a preset area threshold, or the number of vertical surfaces is equal to 1, then the object is determined to be a platform.

4. The control method according to claim 1, wherein, Based on the image recognition results, the automatic pool cleaning device (20) is controlled to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform, including: When the steps and / or platforms in the pool are identified based on the image, the pool automatic cleaning device (20) is controlled to move toward the steps or the platforms and perform a first cleaning mode on the steps or a second cleaning mode on the platforms.

5. The control method according to claim 1, wherein, Based on the image recognition results, the automatic pool cleaning device (20) is controlled to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform, including: Record the position information of the steps and / or platform based on the recognition results of the image; The automatic cleaning device (20) of the pool is controlled to perform a first cleaning mode on the steps and / or a second cleaning mode on the platform according to the recorded location information.

6. The control method according to claim 1, wherein, The first cleaning mode includes: The automatic cleaning device (20) of the pool is controlled to clean each step sequentially along the length of the steps.

7. The control method according to claim 6, wherein, When the automatic water tank cleaning device (20) cleans each step, at least a portion of the cleaning component of the automatic water tank cleaning device (20) covers the area where the horizontal and vertical surfaces of the step meet.

8. The control method according to claim 6, wherein, When the automatic pool cleaning device (20) cleans the steps, the method further includes: Based on the distance value detected by the downward-facing sensor on the automatic pool cleaning device (20), when the automatic pool cleaning device (20) is detected moving from one step to another, the automatic pool cleaning device (20) is controlled to reduce its moving speed.

9. The control method according to claim 1, wherein, The second cleaning mode includes: The automatic cleaning device (20) of the pool is controlled to clean the horizontal and vertical surfaces of the platform along a predetermined path.

10. A non-volatile computer storage medium storing a computer program that, when executed by a processor, implements the method of any one of claims 1-9.

11. An automatic water tank cleaning device (20) capable of performing the method of any one of claims 1-9.

Images