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 a monocular image acquisition device, and the monocular image acquisition device is used for image acquisition. The method comprises the following steps: controlling the pool automatic cleaning device to move on the water surface; acquiring water surface image information through the monocular image acquisition device during the movement; performing target object identification on the water surface image information; and in the case that a target object floating on the water surface within a first predetermined distance from the pool automatic cleaning device is identified, controlling the pool automatic cleaning device to clean the target object. The method controls the pool automatic cleaning device to actively clean the garbage floating on the water surface within the first predetermined distance from the pool automatic cleaning device, thereby improving the cleaning efficiency of the pool.

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 development of computer technology, robotics technology has also developed rapidly. Currently, underwater robots are being used more and more widely in various fields, assisting people in underwater operations, including underwater cleaning, underwater exploration, and underwater tourism.

[0003] Robots used for cleaning pools, such as automated pool cleaning devices, typically move randomly across the water surface when cleaning. During this random movement, the automated pool cleaning device does not actively control its yaw angle or steering; it usually maintains a single direction of movement until it hits the pool wall, at which point it adjusts its direction. This method means that the automated pool cleaning device does not actively clean floating debris around its predetermined path, resulting in poor cleaning efficiency in this random movement mode. Summary of the Invention

[0004] According to one aspect of this disclosure, a control method for an automatic water tank cleaning device is provided. The automatic water tank cleaning device includes a monocular image acquisition device for image acquisition. The method includes: controlling the automatic water tank cleaning device to move on the water surface; acquiring water surface image information through the monocular image acquisition device during the movement; identifying target objects in the water surface image information; and, if a target object floating on the water surface is identified within a first predetermined distance from the automatic water tank cleaning device, controlling the automatic water tank cleaning device to clean the target object.

[0005] According to the control method of the automatic pool cleaning device provided in this disclosure, when there is only one target object, controlling the automatic pool cleaning device to clean the target object includes: determining the location information of the target object; and controlling the automatic pool cleaning device to clean the target object based on the location information.

[0006] According to the control method of the automatic cleaning device for a water tank provided in this disclosure, the position information of the target object is determined based on the calibration parameters of the monocular image acquisition device.

[0007] According to the control method of the automatic pool cleaning device provided in this disclosure, when there are multiple target objects, controlling the automatic pool cleaning device to clean the target objects includes: determining the depth information corresponding to each of the target objects; and controlling the automatic pool cleaning device to clean the target objects based on the depth information corresponding to each of the target objects.

[0008] According to the control method of the automatic cleaning device for a water tank provided in this disclosure, determining the depth information corresponding to each of the target objects includes: inputting the water surface image information into a pre-trained monocular depth estimation model to obtain the depth information corresponding to each of the target objects output by the monocular depth estimation model.

[0009] According to the control method of the automatic pool cleaning device provided in this disclosure, the step of controlling the automatic pool cleaning device to clean the target objects based on the depth information corresponding to each of the target objects includes: determining the target object closest to the automatic pool cleaning device and the distance relationship between the target objects based on the depth information corresponding to each of the target objects; taking the target object closest to the automatic pool cleaning device as the cleaning starting point and determining the target cleaning route based on the distance relationship between the target objects; and controlling the automatic pool cleaning device to clean each of the target objects based on the target cleaning route.

[0010] According to the control method of the automatic pool cleaning device provided in this disclosure, the step of controlling the automatic pool cleaning device to clean the target objects based on the depth information corresponding to each of the target objects includes: determining the target object closest to the automatic pool cleaning device based on the depth information corresponding to each of the target objects, and cleaning the closest target object.

[0011] According to the control method of the automatic cleaning device for a pool provided in this disclosure, during the cleaning of the target object, the monocular image acquisition device performs real-time image acquisition; after determining that the target object gradually disappears from the field of view of the monocular image acquisition device based on the acquired image, the automatic cleaning device for the pool is controlled to continue cleaning along the current cleaning direction for a predetermined distance or a predetermined duration.

[0012] According to the control method of the automatic water tank cleaning device provided in this disclosure, the monocular image acquisition device is disposed at the head of the automatic water tank cleaning device near the top.

[0013] According to a second aspect of this disclosure, an automatic water tank cleaning device is provided, wherein the automatic water tank cleaning device is capable of performing the control method described in any of the above descriptions.

[0014] According to a third aspect of this disclosure, a computer storage medium is provided, wherein a computer program is stored therein, which, when executed by a processor, implements any of the methods described above.

[0015] The embodiments described in this application have the following beneficial effects: The control method of the automatic water tank cleaning device provided in this application can collect image information based on the monocular image acquisition device of the automatic water tank cleaning device during the process of controlling the automatic water tank cleaning device to move on the water surface for random cleaning, and can identify floating garbage around the automatic water tank cleaning device by identifying target objects in the image information, thereby controlling the automatic water tank cleaning device to actively clean the garbage floating on the water surface within a first predetermined distance from the automatic water tank cleaning device, thereby improving the cleaning efficiency of the water tank. Attached Figure Description

[0016] 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.

[0017] Figure 1 A schematic flowchart of the control method for the automatic water tank cleaning device provided in this application is shown;

[0018] Figure 2 A schematic diagram of the field of view of the monocular image acquisition device provided in this application is shown; and

[0019] Figure 3 A schematic diagram of the defined target cleaning route provided in this application is shown. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] This application provides a control method for an automatic pool cleaning device. The automatic pool cleaning device of this application can clean 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. It is understood that the automatic pool cleaning device includes a monocular image acquisition device, which can be used to acquire water surface image information of the pool. 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.

[0022] Unless otherwise specified, the following description will use a robot as an example of the automatic pool cleaning device and a swimming pool as an example of a pool or pool-shaped structure.

[0023] 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.

[0024] Figure 1 A flowchart illustrating the control method of the automatic water tank cleaning device provided in this application is shown. Figure 1 As shown, the control method 100 includes steps 101 to 104. Steps 101 to 104 will be described in detail below.

[0025] In step 101, the automatic cleaning device for the water tank is controlled to move on the water surface.

[0026] For example, when cleaning a pool, the robot can include a variety of cleaning operation modes, such as pool bottom cleaning mode, pool wall cleaning mode, water surface edge cleaning mode, and water surface random mode. The embodiments of this application mainly focus on the robot cleaning in the water surface random mode.

[0027] In the random surface mode, the robot moves across the surface of the pool to clean it. Also known as the boat mode, this mode involves the robot moving along random or relatively random paths to remove floating debris and surface contaminants. In this mode, the robot's vertical movement is primarily driven by gravity and buoyancy. The robot adjusts its buoyancy to control its vertical movement, essentially adjusting its height relative to the water's surface. The robot's direction of movement on the water can be influenced by various external environmental factors. For example, the speed and direction of water flow can exert forces and torques on the robot, causing it to deviate from its intended path. Furthermore, collisions with pool walls or obstacles can generate reaction forces that alter the robot's direction of motion. Friction also plays a significant role in the robot's interaction with the water surface. Although the coefficient of friction is low, it still creates resistance to the robot's movement, especially at low speeds. Friction between the robot and the pool wall also affects its direction of movement on the water surface. Furthermore, wind direction also influences the robot's movement, especially on open water, where wind can exert a horizontal thrust, causing it to deviate from its intended course. Therefore, the robot's direction of movement on the water surface may be influenced by a combination of external factors, which collectively determine its trajectory.

[0028] During the movement, step 102 is performed, in which water surface image information is acquired through the monocular image acquisition device.

[0029] Specifically, during the process of controlling the robot to move on the water surface, image information of the pool's surface can be acquired using a monocular image acquisition device. Such a device could be a monocular camera or a monocular video camera. The water surface image information acquired by the monocular image acquisition device can be multi-dimensional, including static two-dimensional image data, as well as various attributes such as color, texture, contour, and spatial relationships.

[0030] Figure 2 A schematic diagram of the field of view of the monocular image acquisition device provided in this application is shown. Figure 2 As shown, if a monocular image acquisition device is installed on the robot and near the water surface, for example... Figure 2At point A, during image acquisition, only a small area of ​​the water surface can be captured. Furthermore, the closer the monocular image acquisition device is to the water surface, the smaller the corresponding water surface area it can capture. Additionally, it's difficult to capture objects on the water surface clearly and completely from its acquisition angle. For example, when the monocular image acquisition device is set at a horizontal plane, part of its field of view captures useless underwater information, while another part is captured at an angle parallel to the horizontal plane, where objects overlap and are difficult to clearly display. Therefore, when setting up a monocular image acquisition device, it should be placed at a relatively high position on the robot to ensure that it can capture sufficient areas of features on the water surface. In one embodiment, the monocular image acquisition device is positioned near the top of the head of the automatic water cleaning device, such as... Figure 2 Point B in the diagram. Of course, the monocular image acquisition device can also be positioned on top of the robot, for example... Figure 2 At position C in the diagram.

[0031] After acquiring the water surface image information, proceed to step 103. In step 103, target object identification is performed on the water surface image information.

[0032] Understandably, before identifying targets in water surface images, preprocessing is necessary. Image preprocessing is a crucial step before image analysis (feature extraction, segmentation, matching, and recognition), aiming to eliminate irrelevant information, restore useful real-world information, enhance the detectability of relevant information, and simplify data to the maximum extent possible, thereby improving the reliability of feature extraction, image segmentation, matching, and recognition. Image preprocessing can include, but is not limited to, various methods such as image denoising, image geometric transformation, image filtering, image enhancement, image data normalization, and image restoration.

[0033] The target objects in water surface images can be floating debris that needs cleaning, such as leaves, plastic bags, and other impurities that affect the cleanliness of the water. When identifying target objects in water surface images, deep learning models can be used. These models include, but are not limited to, R-CNN, Faster R-CNN, SSD, and YOLO series models. It's understood that the deep learning model for water surface image recognition is a pre-trained model. During training, a large amount of water surface image data can be collected beforehand and manually labeled with target object tags. A training set is then constructed using this large amount of water surface image data and the manually labeled tags. The deep learning model is then trained based on this training set to obtain a pre-trained deep learning model that meets the recognition requirements.

[0034] Next, proceed to step 104. In step 104, if a target object is detected floating on the water surface within a first predetermined distance from the automatic water cleaning device, control the automatic water cleaning device to clean the target object.

[0035] It is understandable that the robot's movement direction on water may be influenced by a variety of external factors, which collectively determine the robot's trajectory. Therefore, it is difficult to precisely control the robot's movement path on water. The farther the target object is from the robot, the more difficult it is to accurately control the robot to move to the target object; the closer the target object is to the robot, the easier it is to accurately control the robot to move to the target object. Based on this, after performing water surface image recognition and identifying a target object floating on the water surface, the robot can be controlled to actively clean only targets within a first predetermined distance from the robot. For targets beyond the first predetermined distance, on the one hand, the target object is too far from the robot, making it difficult to accurately control the robot to move to the target object for cleaning; on the other hand, the target object is too far from the robot, requiring the robot to travel a long distance to clean it, resulting in high energy consumption and low cleaning efficiency. Therefore, targets beyond the first predetermined distance can be ignored. In one embodiment, the first predetermined distance can be, for example, 80 cm to 120 cm.

[0036] The robot is equipped with a water pump and a trash can inside, and has a suction port and a drain port on its body. When cleaning a target object within a predetermined distance, the robot first moves to the location of the target object or its vicinity. Then, the water pump draws water from the target object and its surroundings into the robot through the suction port. Inside the robot, the trash can filters out the target object, leaving it inside, while the clean water is drained out through the drain port, thus achieving the cleaning of the target object.

[0037] The embodiments described in this application have the following beneficial effects:

[0038] The control method for the automatic water tank cleaning device provided in this application, during the process of controlling the automatic water tank cleaning device to move on the water surface for random cleaning, can acquire image information based on the monocular image acquisition device of the automatic water tank cleaning device, and identify floating garbage around the automatic water tank cleaning device by identifying target objects in the image information, thereby controlling the automatic water tank cleaning device to actively clean the garbage floating on the water surface within a first predetermined distance from the automatic water tank cleaning device, thus improving the cleaning efficiency of the water tank.

[0039] In one embodiment, when there is only one target object, controlling the automatic pool cleaning device to clean the target object includes: determining the location information of the target object; and controlling the automatic pool cleaning device to clean the target object based on the location information.

[0040] The target object can be a single object (such as a leaf or a twig) or a single object formed by multiple objects stacked or clustered together (for example, multiple leaves stacked or clustered together are considered as one target object; multiple twigs stacked or clustered together are considered as one target object). In the case of multiple objects stacked or clustered together, the camera often cannot distinguish the multiple independent objects, so they can be treated as a single target object.

[0041] The case of a single target object includes both the case of having only one target object and the case of multiple targets overlapping each other being considered as one target object. For example, a single leaf floating on the water and two overlapping leaves floating on the water are each considered as one target object. The following will explain in detail using a leaf as an example.

[0042] In the case of a single leaf (including both single leaves and multiple leaves overlapping), the first step is to determine the leaf's position information. This can be achieved in various ways, such as using a depth sensor mounted on the robot (e.g., ultrasonic sensors, infrared sensors, vision sensors, lidar, and / or distance sensors), or using a depth estimation model. In one embodiment, the leaf's position information can also be determined based on the calibration parameters of the monocular image acquisition device. The monocular image acquisition device can be pre-calibrated before leaving the factory. These calibration parameters can include intrinsic and extrinsic parameters. Intrinsic parameters may include focal length, principal point (optical center) coordinates, distortion coefficients, etc., while extrinsic parameters may include rotation matrices and translation vectors. The leaf's position relative to the monocular image acquisition device can be determined using its calibration parameters. Furthermore, the position of the monocular image acquisition device on the robot is known in advance. After determining the leaf's position relative to the monocular image acquisition device, coordinate system transformation can be used to determine the leaf's position relative to the robot. This position information may include, for example, distance and direction. Once the position of the leaf relative to the robot is determined, the robot can be controlled to adjust its trajectory and posture according to the position information, and move towards the leaf to clean it.

[0043] It is understood that the above description of determining the position information of the leaf is merely exemplary, and the method of determining the position information of the leaf protected by this application is not limited to the content listed above. Those skilled in the art can make adjustments according to the actual situation, as long as the technical principles of this application can be achieved.

[0044] In one embodiment, when there are multiple target objects, controlling the automatic pool cleaning device to clean the target objects includes: determining the depth information corresponding to each of the target objects; and controlling the automatic pool cleaning device to clean the target objects based on the depth information corresponding to each of the target objects.

[0045] Specifically, when there are multiple targets, before cleaning multiple targets, it is necessary to first determine the depth information corresponding to each target. Then, based on the depth information corresponding to each target, the robot can be controlled to move and clean multiple targets step by step.

[0046] In one embodiment, determining the depth information corresponding to each of the target objects may include: inputting the water surface image information into a pre-trained monocular depth estimation model to obtain the depth information corresponding to each of the target objects output by the monocular depth estimation model. For example, water surface image information acquired by a monocular image acquisition device can be input into a pre-trained monocular depth estimation model. During training, the monocular depth estimation model learns from a large amount of image data and can predict the depth information of each target object in a scene from a single image. This depth information can characterize the distance between the target object and the robot. The monocular depth estimation model may be, for example, RCNN, Faster R-CNN, SSD, or YOLO series models. The monocular depth estimation model, by learning from a large amount of image data, can predict the depth information of each object in a 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.

[0047] In one embodiment, controlling the automatic pool cleaning device to clean the target objects based on the depth information corresponding to each of the target objects includes: determining the target object closest to the automatic pool cleaning device and the distance relationship between the target objects based on the depth information corresponding to each of the target objects; taking the target object closest to the automatic pool cleaning device as the cleaning starting point and determining the target cleaning route based on the distance relationship between the target objects; and controlling the automatic pool cleaning device to clean each of the target objects based on the target cleaning route.

[0048] Specifically, based on the depth information corresponding to each target object, the target object closest to the robot and the distance relationships between the targets can be determined. Furthermore, the target object closest to the robot can be used as the cleaning starting point, and a target cleaning route can be determined according to the distance relationships between the targets, so as to control the robot to perform the subsequent cleaning work according to the determined target cleaning route.

[0049] For example, when determining the target cleaning route, after taking the nearest target object as the starting point for cleaning, the target object closest to the previous target object can be taken as the next cleaning point, and so on, until the target cleaning route is determined based on multiple target objects. Figure 3 A schematic diagram of the defined target cleaning route provided in this application is shown, such as... Figure 3 As shown, for example, if there are 5 targets, and target 1 is closest to the robot, then target 1 is the cleaning starting point. Target 3 is closest to target 1, so target 3 is the next cleaning point. Target 2 is closest to target 3, so target 2 is the next cleaning point after target 3. Target 5 is closest to target 2, so target 5 is the next cleaning point after target 2. Target 4 is closest to target 5, so target 4 is the next cleaning point after target 5. The resulting target cleaning route is as follows: Figure 2 The numbers 1-3-2-5-4.

[0050] Based on the cleaning route determined by the above method, the next cleaning target is always the object closest to the current robot. This makes it easier to control the robot to move towards the target object more accurately. At the same time, it can also reduce the moving distance of the robot during the cleaning process, save energy and improve the cleaning efficiency of the robot.

[0051] In one embodiment, controlling the automatic pool cleaning device to clean the target objects based on the depth information corresponding to each of the target objects includes: determining the target object closest to the automatic pool cleaning device based on the depth information corresponding to each of the target objects, and cleaning the closest target object.

[0052] Specifically, when there are multiple targets within a first predetermined distance, the target closest to the robot can be determined based on the depth information corresponding to each target. Then, the robot is controlled to clean the closest target. Furthermore, after the robot cleans the closest target, the target closest to the robot can be re-determined based on the robot's current position and cleaned again. This process is repeated until there are no more targets within the first predetermined distance from the robot.

[0053] Understandably, if there is no target object within the first predetermined distance from the robot, the robot will continue to execute the random water surface mode to clean the water surface.

[0054] In one embodiment, during the cleaning of the target object, images are acquired in real time by the monocular image acquisition device; after determining that the target object gradually disappears from the field of view of the monocular image acquisition device based on the acquired images, the automatic pool cleaning device is controlled to continue cleaning along the current cleaning direction for a predetermined distance or a predetermined duration.

[0055] Specifically, such as Figure 2 As shown, when the monocular image acquisition device is positioned at location B, there is a certain blind zone between the lower edge of the monocular image acquisition device's field of view and the robot (such as the area between the dotted line below the box at location B and the robot). Although this blind zone is not visible in the monocular image acquisition device's field of view, it is still an area untouched by the robot. When a target object is located in this blind zone, the monocular image acquisition device cannot see the target object, but this does not mean that the target object has been cleared by the robot. Furthermore, there is a certain distance between the robot's bottom suction port and its front edge. During the robot's movement towards the target object, if the target object is located between the bottom suction port and the robot's front edge, the monocular image acquisition device cannot acquire an image of the target object, but the target object has not yet been sucked into the bottom suction port.

[0056] Therefore, during the process of controlling the robot to clean the target object, images can still be acquired in real time using a monocular image acquisition device. After the target object gradually disappears from the field of view of the monocular image acquisition device based on the acquired images, the robot still needs to continue moving along the current cleaning direction and cleaning for a predetermined distance or duration to ensure that the target object in front of the robot and located in the blind spot of the monocular image acquisition device is cleaned. The predetermined distance can be determined comprehensively based on the installation position, field of view size, and shooting direction of the monocular image acquisition device, and the predetermined duration can be determined comprehensively based on the installation position, field of view size, shooting direction of the monocular image acquisition device, and the robot's moving speed, etc. These details are not elaborated upon in this embodiment.

[0057] Based on the above method, it can be ensured that even after the target object disappears from the field of view of the monocular image acquisition device, the target object in the blind spot of the field of view can still be cleaned reasonably, thus ensuring the cleaning effect of the robot.

[0058] According to a second aspect of this application, an automatic water tank cleaning device is also provided. The automatic water tank cleaning device is capable of performing the control methods described in the various embodiments above.

[0059] According to a third aspect of this application, a non-transitory computer-readable storage medium is also provided, on which a computer program is stored. When executed by a processor, the computer program implements a control method for the automatic water tank cleaning device provided in the above embodiments. The automatic water tank cleaning device includes a monocular image acquisition device for image acquisition. The method includes: controlling the automatic water tank cleaning device to move on the water surface; acquiring water surface image information through the monocular image acquisition device during the movement; identifying a target object in the water surface image information; and, if a target object floating on the water surface is identified within a first predetermined distance from the automatic water tank cleaning device, controlling the automatic water tank cleaning device to clean the target object. The principle and scheme of the control method are described above in conjunction with the various embodiments and accompanying drawings, and will not be repeated here.

[0060] Fourthly, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the control method for the automatic water tank cleaning device provided by the above methods. The automatic water tank cleaning device includes a monocular image acquisition device for image acquisition. The method includes: controlling the automatic water tank cleaning device to move on the water surface; acquiring water surface image information through the monocular image acquisition device during the movement; identifying target objects in the water surface image information; and controlling the automatic water tank cleaning device to clean the target object when a target object floating on the water surface is identified within a first predetermined distance from the automatic water tank cleaning device. The principle and scheme of the control method are described above in conjunction with various embodiments and accompanying drawings, and will not be repeated here.

[0061] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0062] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0063] 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.

[0064] 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.

[0065] 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.

[0066] 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 comprising a monocular image acquisition device, the monocular image acquisition device being used for image acquisition, the method comprising: Control the movement of the automatic cleaning device on the water surface; During the movement, water surface image information is acquired through the monocular image acquisition device; Target object identification is performed on the water surface image information; If a target object is detected floating on the water surface within a first predetermined distance from the automatic water cleaning device, the automatic water cleaning device is controlled to clean the target object.

2. The control method according to claim 1, wherein when there is only one target object, controlling the automatic cleaning device of the pool to clean the target object includes: Determine the location information of the target object; Based on the location information, the automatic cleaning device of the pool is controlled to clean the target object.

3. The control method according to claim 2, wherein, The location information of the target object is determined based on the calibration parameters of the monocular image acquisition device.

4. The control method according to claim 1, when there are multiple target objects, the step of controlling the automatic water tank cleaning device to clean the target objects includes: Determine the depth information corresponding to each of the aforementioned targets; Based on the depth information corresponding to each of the target objects, the automatic cleaning device of the pool is controlled to clean the target objects.

5. The control method according to claim 4, wherein determining the depth information corresponding to each of the target objects includes: The water surface image information is input into a pre-trained monocular depth estimation model to obtain the depth information corresponding to each of the target objects output by the monocular depth estimation model.

6. The control method according to claim 4, wherein controlling the automatic water tank cleaning device to clean the target objects based on the depth information corresponding to each of the target objects includes: Based on the depth information corresponding to each of the target objects, the target object closest to the automatic cleaning device of the pool and the distance relationship between the target objects are determined. The target object closest to the automatic cleaning device of the pool is taken as the cleaning starting point, and the target cleaning route is determined based on the distance relationship between the target objects; The automatic cleaning device for the water tank is controlled to clean each of the target objects according to the target cleaning route.

7. The control method according to claim 4, wherein controlling the automatic water tank cleaning device to clean the target objects based on the depth information corresponding to each of the target objects includes: Based on the depth information corresponding to each of the target objects, the target object closest to the automatic cleaning device of the pool is determined, and the nearest target object is cleaned.

8. The control method according to any one of claims 1-7, wherein, During the cleaning process of the target object, images are acquired in real time using the monocular image acquisition device; After determining that the target object gradually disappears from the field of view of the monocular image acquisition device based on the acquired image, the automatic water tank cleaning device is controlled to continue cleaning along the current cleaning direction for a predetermined distance or a predetermined duration.

9. The control method according to any one of claims 1-7, wherein, The monocular image acquisition device is positioned near the top of the head of the automatic water tank cleaning device.

10. 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 1-9.

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

Images