Control method of swimming pool robot
By equipping the pool robot with an image acquisition device and performing dual image acquisitions in moderately turbid water to confirm the object's position, the problem of insufficient cleaning and obstacle avoidance accuracy of the pool robot in turbid water was solved, achieving higher cleaning and obstacle avoidance accuracy.
Patent Information
- Application Number
- CN202511671933.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-14
- Publication Date
- 2026-02-10
AI Technical Summary
Existing pool robots struggle to accurately control cleaning and obstacle avoidance in murky water, resulting in insufficient cleaning and obstacle avoidance precision.
The pool robot is equipped with image acquisition equipment. It identifies the turbidity of the water through images and performs two image acquisitions when the water is moderately turbid to confirm the location of objects. The robot's movement is controlled by combining the image recognition results to achieve precise cleaning and obstacle avoidance.
This improves the accuracy of the pool robot in cleaning and obstacle avoidance in turbid water, reduces the probability of misidentification, and ensures the accuracy of cleaning and obstacle avoidance operations.
Smart Images
Figure CN121500969A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of pool robot technology, and more particularly to a control method for a pool robot. Background Technology
[0002] With the rapid advancement of robotics technology, pool robots are being used more and more widely in the field of pool cleaning. When working in water, pool robots need to perform tasks such as cleaning and obstacle avoidance.
[0003] However, when the pool becomes murky, how to control the pool robot to achieve better cleaning and obstacle avoidance is an urgent problem to be solved. Summary of the Invention
[0004] In view of this, embodiments of this application provide a control method for a pool robot, enabling the pool robot to perform better cleaning and obstacle avoidance, and improving the accuracy of cleaning and obstacle avoidance.
[0005] This application provides a control method for a swimming pool robot. The swimming pool robot moves on a cleaning surface to perform cleaning operations on the swimming pool. The cleaning surface includes one or more of the pool wall, pool bottom, and water surface. The swimming pool robot is equipped with at least one image acquisition device, which is used to acquire images of the pool surface to identify objects within the pool. The method includes: After determining that the first object has been identified based on at least one frame of the first image, if the turbidity of the pool water is moderate, the pool robot is controlled to move towards the first position of the first object, so that the image acquisition device can acquire at least one frame of the second image containing the first position during the movement; wherein, the turbidity of the water corresponding to moderate is greater than or equal to a first threshold and less than or equal to a second threshold; the first position refers to the position of the first object when it is identified based on the first image. The movement of the pool robot is controlled based on the object recognition results of at least one frame of the second image.
[0006] In some embodiments, the turbidity of the pool water is identified by images acquired through an image acquisition device.
[0007] In some embodiments, the images acquired by the image acquisition device are also used to identify object types, including trash that needs to be cleaned and obstacles that need to be avoided.
[0008] In some embodiments, the method further includes: controlling a pool robot to move along a bow-shaped path in the pool to identify trash in the pool; If, based on at least one frame of a first image, a first object is identified and the first object is trash, the pool robot is controlled to move towards a first position closer to the first object, so that the image acquisition device acquires at least one frame of a second image containing the first position during the movement, including: The pool robot is controlled to move toward a first position, such that during the movement of the pool robot toward the first position or when it moves a preset distance, the image acquisition device acquires at least one frame of a second image containing the first position.
[0009] In some embodiments, controlling the movement of a pool robot based on object recognition results from at least one frame of a second image includes: If a second object is identified based on at least one frame of the second image and the second object is trash, then the pool robot is controlled to clean the trash based on the second position; the second position refers to the location of the second object when it is identified based on the second image. Alternatively, if the second object is not recognized based on at least one frame of the second image, the pool robot is controlled to return to the bow-shaped path; Alternatively, if a second object is identified based on at least one frame of the second image and the second object is an obstacle, then the pool robot is controlled to return to the bow-shaped path.
[0010] In some embodiments, if it is determined based on at least one frame of a first image that a first object has been identified and that the first object is an obstacle, controlling the pool robot to move toward a first position closer to the first object, so that the image acquisition device acquires at least one frame of a second image containing the first position during the movement, includes: The pool robot is controlled to continue moving along the direction of travel when the first object is identified based on the first image, so that the image acquisition device can acquire at least one frame of the second image containing the first position during the movement or when the robot has moved a preset distance.
[0011] In some embodiments, controlling the movement of a pool robot based on object recognition results from at least one frame of a second image includes: If the second object is not identified based on at least one frame of the second image, the pool robot is controlled to continue moving in the direction of travel when the first object was identified based on the first image. Alternatively, if a second object is identified based on at least one frame of the second image and the second object does not affect the movement of the pool robot, then the pool robot is controlled to continue moving in the direction of travel when the first object was identified based on the first image. Alternatively, if a second object is identified based on at least one frame of the second image and the second object is an obstacle affecting the movement of the pool robot, then the pool robot is controlled to initiate obstacle avoidance operation based on the second position, where the second position refers to the location of the second object when it is identified based on the second image.
[0012] In some embodiments, controlling the pool robot to continue moving along the direction of travel when the first object is recognized based on the first image includes: The pool robot is controlled to continue moving at a second speed along the direction of travel when the first object is recognized based on the first image. The second speed is less than the first speed, and the first speed is the moving speed of the pool robot when the first object is recognized based on the first image.
[0013] In some embodiments, the water turbidity also includes low degree, which corresponds to water turbidity being less than a first threshold; when the water turbidity of the pool is low, after determining that the first object is identified based on at least one frame of the first image, the pool robot is controlled to execute a movement strategy for the first object. And / or, the water turbidity also includes severe turbidity, which corresponds to water turbidity greater than a second threshold; if the water turbidity also includes severe turbidity, after determining that the first object has been identified based on at least one frame of the first image, the pool robot is controlled to continue moving along the direction of travel when the first object was identified based on the first image.
[0014] In some embodiments, the method further includes: Based on the received user command or based on the turbidity of the water, the water purification module is activated to perform water purification actions.
[0015] In some embodiments, it also includes: The water turbidity is sent to the client corresponding to the pool robot so that the water turbidity can be displayed on the client periodically or in real time.
[0016] In some embodiments, it also includes: When the turbidity of the water is low, the water depth data collected by the water depth sensor installed on the pool robot and the image collected by the image acquisition device are used to identify the location of the pool robot, which is either below or above the water surface. When the turbidity of the pool water is detected to be moderate or severe, the location of the pool robot is identified by the water depth data collected by the water depth sensor installed on the pool robot.
[0017] The beneficial effects of this application embodiment compared with the prior art are as follows: The pool robot is equipped with at least one image acquisition device, which is used to acquire images of the pool to identify objects in the pool through the images; after determining that the first object has been identified based on at least one first image, if the turbidity of the pool water is moderate, the pool robot is controlled to move towards a first position close to the first object, so that the image acquisition device acquires at least one second image containing the first position during the movement. Based on the object recognition result of at least one second image, the pool robot is controlled to move; the turbidity of the water corresponding to moderate is greater than or equal to a first threshold and less than or equal to a second threshold. The first position refers to the position of the first object when it is identified based on the first image. This realizes that the pool robot can be controlled to move through at least two image recognitions when the water is turbid, ensuring the accuracy of the pool robot moving towards the object, enabling the pool robot to perform better cleaning and obstacle avoidance, and improving the accuracy of cleaning and obstacle avoidance. Attached Figure Description
[0018] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0019] Figure 1 This is a schematic diagram of the swimming pool provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the pool robot control method provided in the embodiments of this application. Detailed Implementation
[0020] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.
[0021] The following describes in detail, with reference to the accompanying drawings, a control method for a swimming pool robot according to an embodiment of this application.
[0022] Figure 1 This is a diagram of the swimming pool, such as... Figure 1As shown, the cleaning surface of the swimming pool includes one or more of the pool wall surface 11, pool bottom surface 12, and water surface 13. The pool robot 2 moves on the cleaning surface to perform cleaning operations on the pool. The pool robot is equipped with at least one image acquisition device, which is used to acquire images of the pool to identify objects in the pool through the images.
[0023] In some examples, the pool may be rectangular or irregular in shape; this is not a limitation.
[0024] The pool robot may include a main body, a filter box, an opening, and a bottom cover. The filter box is at least partially located inside the main body, and the opening is at least partially located at the bottom of the filter box. The bottom cover can be movably opened or closed. The filter box includes a frame and at least one filter screen. At least part of the opening is located on the bottom of the frame, the filter screen is located on the frame, and the bottom cover is movably located on the frame to open or close the opening. Through the filter box, the pool robot can clean up debris in the pool. The pool robot can move on the pool walls, bottom, and surface according to instructions.
[0025] In addition, the pool robot is equipped with at least one image acquisition device, which can be a camera. This device can capture images of the pool surface, allowing it to identify objects within the pool. These objects can include debris such as leaves and sediment that need to be cleaned, as well as obstacles that need to be avoided, such as steps, ladders, underwater lights, and pool wall lights.
[0026] In the embodiments described in this specification, the pool robot can detect water turbidity or receive water turbidity transmitted by other devices.
[0027] When detecting water turbidity, the pool robot can use a water quality analyzer installed to detect turbidity; or it can acquire images of the pool using image acquisition equipment and analyze the turbidity data and images (such as the distribution of suspended matter) using a deep learning model to obtain the water turbidity; or it can use a built-in turbidimeter (usually based on the principle of scattering) to emit light of a specific wavelength (such as 860nm infrared light) to irradiate the water and detect the intensity of the scattered light in a 90° direction. The more suspended particles there are, the stronger the scattered light, thereby calculating the turbidity value.
[0028] In addition, the turbidity of the water can be detected by a separately set turbidity detector. The pool robot can establish a communication connection with the turbidity detector. After the turbidity detector detects the turbidity of the water, it transmits the data to the pool robot, which then receives the turbidity value.
[0029] In some embodiments, the turbidity of the pool water can be identified by images acquired by the image acquisition device.
[0030] When identifying the turbidity of swimming pool water through image recognition, optionally, a ResNet model can be used to train water quality image classification (e.g., clear / low turbidity / moderate turbidity / heavy turbidity, etc.). Then, based on the real-time image recognition, the real-time image classification of the water body can be obtained. Here, clear and low turbidity can be considered as turbidity less than a first threshold. Alternatively, the proportion of suspended matter pixels in the image can be statistically analyzed. If the proportion is less than a first preset proportion (e.g., 15%), the water turbidity is considered less than the first threshold, and thus low turbidity. If the proportion of suspended matter pixels is greater than or equal to the first preset proportion but less than a second preset proportion (e.g., 20%), the water turbidity is considered greater than or equal to the first threshold and less than or equal to the second threshold, thus moderate turbidity.
[0031] In addition, the same image can be used to identify the turbidity of the water in the pool, or it can be used to identify objects, or it can identify both the turbidity of the water and the objects at the same time. There is no specific limitation that the same image can only identify one item.
[0032] By using image recognition to determine the turbidity of the pool water, the real-time detection of water turbidity is ensured.
[0033] In some embodiments, when the turbidity of the pool water is identified using images acquired by the image acquisition device, the identification operation can be triggered by the user through the client corresponding to the pool robot, or it can be automatically triggered by the pool robot (e.g., triggered at the start of the cleaning process).
[0034] In some implementations, after triggering the identification of water turbidity, if the image acquisition device is close to the pool wall, the pool robot can be controlled to move away from the pool wall until the image acquisition device maintains a preset distance from the pool wall, so as to avoid misidentifying the pool wall as dirt and affecting the judgment of water turbidity identification.
[0035] In some implementations, after triggering the identification of water turbidity, if the environment in which the pool robot is currently located is too bright or too dark, the robot can be controlled to pause turbidity identification and resume identification only after the ambient brightness meets the preset requirements. For example, if the pool robot is equipped with a brightness adjustment device such as a supplementary light, the parameters of the supplementary light can be adjusted until the ambient brightness meets the preset requirements before resuming identification, thus avoiding the influence of ambient brightness on the judgment of water turbidity.
[0036] In some implementations, if the pool robot is in an unstable state, such as surfacing or diving, climbing the pool wall, or getting stuck, when water turbidity recognition is triggered, the recognition can be paused to avoid result deviation. When turbidity recognition is paused, a prompt can be issued to the user, such as asking them to wait for the pool robot to stabilize before re-triggering turbidity recognition, or prompting the user to wait for the pool robot to stabilize before providing the final recognition result. Prompt methods include, but are not limited to, one or more of the following: images, text, voice, and lighting effects.
[0037] In some implementations, after triggering water turbidity identification, the pool robot can move a preset distance or for a preset time to collect multiple turbidity identification results during this process. These results are then comprehensively analyzed to arrive at a final identification result, thereby improving the accuracy of turbidity identification. The comprehensive analysis can be any method and is not limited here.
[0038] In some implementations, the pool robot may acquire a turbidity identification result only once during a single cleaning process (without repeating the identification later), or it may continuously perform turbidity identification. When the turbidity identification result changes, the display of the turbidity identification result can be updated synchronously, and / or the corresponding operating mode of the pool robot can be adjusted.
[0039] The turbidity of swimming pool water can be classified into any two categories: low, medium, and high. This classification can be achieved by setting a preset threshold. For example, the preset threshold could be 1 NTU, 1.5 NTU, etc., without any specific limitation here.
[0040] Water turbidity can be classified as low, moderate, and severe. Low turbidity corresponds to a turbidity level below a first threshold, moderate turbidity corresponds to a turbidity level greater than or equal to the first threshold and less than or equal to a second threshold, and severe turbidity corresponds to a turbidity level greater than the second threshold. For example, the first threshold could be 1.5 NTU, and the second threshold could be 2 NTU.
[0041] When the turbidity of the water is less than or equal to a second threshold, objects can be identified using images. In some embodiments, object type and / or object location can also be identified based on images. In some embodiments, object types include trash that needs to be cleaned and obstacles that need to be avoided. The swimming pool robot's speed and / or attitude (e.g., the robot's roll angle, pitch angle, and yaw angle can be used to characterize its attitude) can be adjusted to clean trash and avoid obstacles. For example, if the object type is trash, the swimming pool robot is controlled to move towards the object's location to clean it, ensuring cleaning efficiency and accuracy. If the object type is an obstacle, the swimming pool robot is controlled to initiate obstacle avoidance to ensure accuracy. For trash, subtypes such as leaves, fruit, and sand can be identified based on images. For obstacles, subtypes such as roller shutters, floor drains, and bar pillars can be identified based on images.
[0042] It should be noted that, in the following embodiments, for the sake of distinction, the image corresponding to the first time an object is detected can be referred to as the first image, the detected object as the first object, and the location of the first object as the first location. The image captured during the movement towards the first object or the movement of a preset distance can be referred to as the second image, the detected object as the second object, and the location of the second object as the second location. The first object and the second object can be the same object or different objects. The first location and the second location can be the same or different.
[0043] For example, when the pool robot is far from an object, if the object appears within the image acquisition range of the image acquisition device, the object can be identified based on the image. The initial object identification result can be output as the first identification result. Alternatively, to further ensure accuracy, a baseline distance for object identification can be preset. When an object is identified based on the image, if the distance between the pool robot and the object is greater than the baseline distance, the identification result is ignored, and the pool robot continues along a planned or random path. When an object is identified based on the image, if the distance between the pool robot and the object is less than or equal to the baseline distance, the object identification result is output as the first identification result.
[0044] Accordingly, the image corresponding to the first recognition result can be used as the first image, and the object recognized by the first recognition result can be used as the first object. The first recognition result may only include the recognized first object. The first recognition result may also include the object type of the first object and / or the first position of the first object. Of course, the first position of the first object can also be identified in conjunction with other sensors.
[0045] In some embodiments, the water turbidity is low. When the turbidity of the pool water is low, the pool robot is controlled to execute a movement strategy for the first object after determining that the first object has been identified based on at least one frame of the first image.
[0046] If the water turbidity is low, this initial identification result can be used as the final object identification result, allowing the pool robot to perform cleaning or obstacle avoidance operations based on this result. When the water turbidity is less than a first threshold, it indicates high water clarity, meaning the acquired images can accurately identify objects with a low probability of misidentification. Furthermore, when identifying object type or location based on the image, the robot can accurately determine the object type or location, thus more accurately assisting the pool robot in performing obstacle avoidance or cleaning operations. For example, when identifying objects requiring cleaning such as leaves or fruit based on the image, the pool robot can determine the object's location based on the image, and then adjust its speed and / or posture to reach the object's location for targeted cleaning. When identifying obstacles such as floor drains, bar pillars, or roller shutters based on the image, the robot can combine at least one of the following: image, ultrasonic sensors, or laser rangefinders to determine the obstacle's location. The pool robot can then adjust its speed and / or posture during movement to precisely avoid obstacles based on the obstacle type and / or its location.
[0047] When the water turbidity is moderate, the water quality is relatively turbid, and the noise in the acquired images is relatively large, which increases the probability of misidentification. After the pool robot identifies the first object in at least one frame of the acquired image when it is far away from the first object, it can be controlled to move closer to the first position. This allows the image acquisition device to acquire images again during the movement of the pool robot or after moving a preset distance. Then, the object recognition result is reconfirmed based on the re-acquired image to reduce the probability of misidentification.
[0048] If the water turbidity is moderate, after identifying the first object based on at least one frame of the first image, the pool robot can be controlled to move towards the first location. This allows the image acquisition device to capture an image containing the first location during the movement or after moving a preset distance, serving as the second image. The second image is then used for object recognition, outputting a second recognition result. This second recognition result is used as the final recognition result, enabling the pool robot to perform cleaning or obstacle avoidance operations on the first object based on this result. The object identified in the second recognition result is referred to as the second object, and the location of the second object is referred to as the second location. The second recognition result may only include the identification of the second object. Alternatively, it may include the object type of the second object and / or the second location. Of course, the second location can also be obtained by combining the location of the second object with other sensors.
[0049] Based on the above solution, in one embodiment provided in this specification, such as Figure 2 As shown, the control method for the pool robot includes the following steps: Step 201: After determining that the first object has been identified based on at least one frame of the first image, if the turbidity of the pool water is moderate, control the pool robot to move towards the first position, so that the image acquisition device can acquire at least one frame of the second image containing the first position during the movement. The first position refers to the location of the first object when it is identified based on the first image.
[0050] Step 202: Based on the object recognition results of at least one frame of the second image, control the movement of the pool robot.
[0051] In one scenario example, a swimming pool robot can be controlled to perform AI navigation, following a bow-shaped path. During AI navigation, image acquisition equipment can be used to identify litter. Once litter is detected, the swimming pool robot can go to the location of the litter and clean it up.
[0052] In scenarios where the AI navigation path is a bow-shaped path, the pool robot can be controlled to move along the bow-shaped path in the pool to identify trash in the pool. After determining that the first object has been identified based on at least one frame of the first image and that the first object is trash, if the turbidity of the water is low, the pool robot can be controlled to move to the first position to clean the first object. Alternatively, if the first object is a pile of trash, a cleaning path can be planned based on the first position, and the pile of trash can be cleaned based on the planned cleaning path.
[0053] If the turbidity of the water is moderate, the pool robot can be controlled to move towards the first position first, so that during the process of the pool robot moving towards the first position or when it moves a preset distance, the image acquisition device acquires at least one frame of the second image containing the first position.
[0054] If a second object is identified based on at least one frame of the second image and the second object is trash, the pool robot is controlled to clean the trash based on a second position; the second position refers to the location of the second object when it is identified based on the second image. For example, the pool robot can be controlled to move to the second position to clean the first object, or, if the first object is a pile of trash, a cleaning path can be planned based on the second position, and the pile of trash can be cleaned based on the planned cleaning path.
[0055] Alternatively, if the second object is not detected based on at least one frame of the second image, the pool robot is controlled to return to the bow-shaped path. Alternatively, if the second object is detected based on at least one frame of the second image and the second object is an obstacle, the pool robot is controlled to return to the bow-shaped path. That is, if trash is not detected or an obstacle is detected based on at least one frame of the second image, the pool robot is controlled to return to the bow-shaped path.
[0056] For example, a swimming pool robot can be controlled to move along a bow-shaped path in a swimming pool to identify trash. If, based on at least one frame of a first image, a first object is identified and is indeed trash, and if the first position of the first object is not on the bow-shaped path, the swimming pool robot's direction of travel is adjusted from along the bow-shaped path to towards the first position; and the robot is then controlled to move towards the first position to clean the trash. If the first object is on the bow-shaped path, the robot can continue moving along the path to clean the trash. When the water turbidity is moderate, during the robot's movement towards the first position or after moving a preset distance, an image acquisition device is controlled to acquire at least one frame of a second image containing the first position to further verify the identification results. If trash is identified based on at least one frame of the second image, and the trash's position has not changed, there is no need to adjust the direction of travel or the speed of travel. If the detected location of the trash differs from that identified in the first image, the pool robot is controlled to adjust its speed and / or direction of travel to move towards the trash and clean it. If the first object is not identified or is identified as an obstacle based on at least one frame of the second image, the pool robot is controlled to adjust its speed and / or direction of travel, return to the zigzag path, and continue AI navigation.
[0057] In some embodiments, for AI navigation, if the object type is garbage when the water turbidity is moderate, the pool robot is controlled to move towards a first position; if the object type changes to an obstacle or no object is detected based on the second image during the movement of the pool robot or after moving a preset distance, the swimming speed of the pool robot is reduced and / or the posture is adjusted.
[0058] If the water turbidity is moderate, the water quality is relatively murky, resulting in higher noise levels in the acquired images and increasing the probability of misidentification. In this case, if the object type is trash, the pool robot can be controlled to move towards the first position. During the movement or after moving a preset distance, the trash can be identified again. If the identified object type changes or no object is identified when performing object recognition based on the second image, the pool robot's speed can be reduced and / or its posture adjusted to promptly change its direction of travel and return to the zigzag path. For example, the pool robot can be controlled to reduce its speed to zero, and then its yaw angle can be adjusted to align its direction of travel towards the zigzag path. Alternatively, if trash is identified, but its position has changed relative to the first image, the pool robot's direction of travel can be adjusted to face the changed position to perform cleaning.
[0059] Specifically, the preset distance can be set according to the distance between the initial position of the first object and the first object. For example, the preset distance can be 1 / 2 or 2 / 3 of the distance. No specific limitation is made here.
[0060] When the water turbidity is moderate, this method first identifies the object type and location. If the object is trash, the pool robot continues moving towards it. During movement or after traveling a preset distance, a second identification is performed. The result of this final identification determines whether further cleaning is necessary. Due to the turbidity of the water, the captured images have higher noise levels, increasing the probability of false identification. In this case, controlling the pool robot to move closer to the initial location or after traveling a preset distance before re-identifying reduces the impact of noise on the identification results and lowers the probability of false identification. When the detected object type changes to an obstacle or is absent, the pool robot's speed is reduced to facilitate timely adjustments to its direction. If trash is detected but its location changes, the pool robot's direction is adjusted to face the new location, preventing the robot from continuing in its original direction and failing to effectively clean the trash.
[0061] Alternatively, when the AI cruise uses a random path, if the object type is detected to have changed to an obstacle during the secondary recognition process, the recognition result of that second recognition will be used to control the pool robot to avoid the obstacle. After that, a random path will be selected to perform the cruise, so as to identify garbage during the cruise.
[0062] Alternatively, during a routine bow-shaped cleaning process, if an object is detected using image recognition and its type is trash, the pool robot will continue along its original bow-shaped path when the water turbidity is moderate. However, if the object type changes to an obstacle affecting the robot's movement during the robot's movement or after traveling a preset distance, the robot will initiate obstacle avoidance. If the object type changes to an obstacle during a secondary recognition process, the result of that secondary recognition will be used, meaning obstacle avoidance will be necessary.
[0063] In some embodiments, if it is determined that a first object has been identified based on at least one frame of a first image and that the first object is an obstacle, the pool robot is controlled to continue moving along the direction of travel when the first object is identified based on the first image, so that the image acquisition device can acquire at least one frame of a second image containing the first position during the movement or when the robot has moved a preset distance.
[0064] If no second object is detected based on at least one frame of the second image, the pool robot is controlled to continue moving in the direction in which the first object was detected based on the first image. Alternatively, if a second object is detected based on at least one frame of the second image and the second object does not affect the movement of the pool robot, the pool robot is controlled to continue moving in the direction in which the first object was detected based on the first image.
[0065] Alternatively, if a second object is identified based on at least one frame of the second image and the second object is an obstacle affecting the movement of the pool robot, then the pool robot is controlled to initiate obstacle avoidance operation based on the second position, where the second position refers to the location of the second object when it is identified based on the second image.
[0066] When the water turbidity is moderate, if the object is an obstacle affecting the pool robot's continued movement, the pool robot is controlled to continue moving in the current direction (i.e., the direction of movement of the pool robot when the first object is recognized based on the first image) to approach the first position and perform secondary recognition of the detected obstacle. It can maintain the original speed (first speed) or slow down to a second speed, which is less than the first speed. The first speed is the movement speed of the pool robot when the first object is recognized based on the first image.
[0067] If, during the movement of the pool robot or after moving a preset distance, no object is detected based on at least one frame of the second image, or if the detected object does not affect the movement of the pool robot, the pool robot is controlled to continue moving in the direction of travel when the first object was detected based on the first image. If an object is detected and it is an obstacle affecting the movement of the pool robot, the pool robot is controlled to initiate obstacle avoidance.
[0068] The water turbidity is moderate, indicating relatively murky water quality. This results in significant noise in the acquired images, increasing the probability of false recognition. If the object is an obstacle affecting the pool robot's continued movement, the robot can be controlled to continue moving at a low speed to approach the object for re-identification. If the object is identified again during the robot's movement or after traveling a preset distance, and if the identified object does not affect the robot's progress (e.g., the object is determined to be trash based on the second image; or the object is determined to be an obstacle based on the second image, but its position does not affect the robot's movement), or if no object is identified (meaning there are currently no factors affecting the robot's movement), the robot can be controlled to continue moving in the current direction.
[0069] If an object is identified again during the movement of the pool robot or after moving a preset distance, and this object affects the movement of the pool robot (e.g., the position of the object identified based on the second image has not changed compared to the object identified based on the first image, or has changed only slightly), that is, the object has affected the movement of the pool robot, then the pool robot is controlled to start obstacle avoidance operation.
[0070] When the water turbidity is moderate, this method first identifies the type and location of objects. Then, while the pool robot is moving forward or after traveling a preset distance, it performs a second identification to check if the impact of the first object on the robot has changed. Based on the final identification result, it determines whether to initiate obstacle avoidance or continue moving. Because the water is relatively turbid, the acquired images have significant noise, increasing the probability of false identification. This method, by performing a second identification during the pool robot's movement or after traveling a preset distance, reduces the impact of noise on the identification results, lowers the probability of false identification, and improves the accuracy of obstacle avoidance.
[0071] When the water turbidity is moderate, if the object is an obstacle, the pool robot will continue moving in the current direction. If, during movement or after traveling a preset distance, the object is detected as trash and continuing in the current direction will not cover it, the pool robot can adjust its direction and move towards the location of the trash. By performing secondary identification and cleaning only when the object is detected as trash in the second identification, missed cleaning can be further reduced.
[0072] In other embodiments, when the water turbidity exceeds a second threshold, the images acquired by the image acquisition device have low reliability, and the images are no longer used for object recognition. That is, when the water turbidity is severe, after determining that a first object has been identified based on at least one frame of the first image, the identification result is ignored, and the pool robot continues to move along the direction in which the first object was identified based on the first image.
[0073] The water turbidity level is considered severe when the water turbidity exceeds a second threshold. For example, the second threshold could be 2 NTU.
[0074] By classifying water turbidity into low, medium, and high levels, the pool robot can be controlled accordingly based on the different turbidity levels, ensuring the accuracy of the pool robot's actions.
[0075] In some embodiments, the water turbidity is sent to the client corresponding to the pool robot so that the water turbidity can be displayed periodically or in real time on the client.
[0076] Specifically, the turbidity of the water can be displayed periodically or in real-time on the client side. For example, periodic display could be every 2 hours or every 4 hours, and users can set it according to their needs.
[0077] Optionally, you can display only the severely turbid material.
[0078] In some embodiments, when the turbidity of the water is low, the location of the pool robot is jointly identified by the water depth data collected by the water depth sensor installed on the pool robot and the image collected by the image acquisition device, and the location is either below or above the water surface. When the turbidity of the pool water is detected to be moderate or severe, the location of the pool robot is identified by the water depth data collected by the water depth sensor installed on the pool robot.
[0079] Specifically, when the turbidity of the water is low, the recognition accuracy of the pool robot is less affected by the water quality. Water depth data can be collected by the water depth sensor installed on the pool robot, and images can be collected by the image acquisition device. Then, the processor can identify whether the pool robot is below or above the water surface based on the water depth data and the images.
[0080] Optionally, if the location of the pool robot is identified as a first location through water depth data and a second location through image recognition; if the first and second locations are the same (e.g., both are below or above the water surface), then the first or second location is determined as the final identification location; if the first and second locations are different (e.g., the first location is below the water surface and the second location is above the water surface), then the location of the pool robot is identified at least twice by the water depth sensor and the camera, resulting in at least two sets of identification results; if the number of different identification results in the at least two sets of identification results is greater than a preset number, then an error message is issued, indicating that the water depth sensor or camera may be malfunctioning; if the number of different identification results in the at least two sets of identification results is less than a preset number, then the same identification result is taken as the final identification result.
[0081] If the turbidity of the water is moderate or severe, the image acquired by the image acquisition device is considered to be of low reliability. In this case, the location of the pool robot can be identified solely by the water depth data to ensure the accuracy of the identification.
[0082] In some embodiments, adjusting the swimming pool robot's speed and / or posture according to the type of the first object includes: When the object type is garbage, the swimming pool robot's speed is reduced to a first preset speed, and it moves to the first object at the first preset speed to clean the first object; When the object is an obstacle, the swimming pool robot's speed is reduced to a second preset speed, and obstacle avoidance is initiated; based on the obstacle avoidance path corresponding to the obstacle avoidance operation, the swimming pool robot's posture is adjusted.
[0083] Specifically, the first preset speed can be 20% or 30% lower than the current travel speed. By controlling the pool robot to move towards the first object at the reduced first preset speed to clean the first object, the pool robot can approach the first object at a low speed, minimizing the problem of large water flow fluctuations caused by the pool robot's excessively fast travel speed, which would cause the trash to drift away from the pool robot. This reduces the probability of trash drifting away and improves the accuracy of the cleaning action.
[0084] Of course, optionally, when the object type is garbage, the speed of movement can remain unchanged to ensure cleaning efficiency.
[0085] In addition, the second preset speed can be 20% lower than the current speed or 40% lower than the current speed. The second preset speed can be the same as or different from the first preset speed.
[0086] When the object is an obstacle, the swimming pool robot's speed is reduced to a second preset speed and obstacle avoidance is initiated. This allows the swimming pool robot to avoid obstacles at a low speed, minimizing the possibility of a violent collision between the swimming pool robot and the first object if the robot misidentifies the distance or boundary of the first object.
[0087] Optionally, when the pool robot is moving at low speed, it can identify the first object in real time or multiple times. Based on the identification results, the speed and posture of the pool robot can be adjusted in a timely manner to further reduce the probability of the above-mentioned violent collisions or debris drifting away. The speed and posture of the pool robot can also be adjusted in a timely manner based on the identification results to accurately switch the pool robot's actions to obstacle avoidance or cleaning actions, thereby improving the accuracy of the machine's action execution.
[0088] In some embodiments, when the water turbidity is high, the cleaning time per cycle of the pool robot can be reduced for any cleaning mode. For example, if the cleaning time is three hours in normal cleaning mode, it can be reduced to two hours when the water turbidity is high. Since high turbidity may indicate that the pool contains a lot of debris or suspended matter, reducing the cleaning time per cycle allows users or the equipment used to clean the pool robot to empty the filter box more promptly, reducing the risk of filter box clogging and overflow, thereby ensuring the effectiveness of the pool robot's cleaning.
[0089] In some embodiments, the water purification module can be activated based on received user instructions or based on water turbidity to perform water purification actions.
[0090] Specifically, the water purification module can be installed on the pool robot or on the corresponding base station of the pool robot.
[0091] User commands can be triggered by the user, such as when the user sees the pushed information about water turbidity and chooses to trigger the command; or when the user has a need to activate the water purification module, such as water purification after cleaning up a small area of garbage unrelated to water turbidity; or when the user triggers the command based on water turbidity, such as when the water turbidity is moderate or severe.
[0092] In addition, the pool robot can also autonomously activate the water purification module based on the turbidity of the water. For example, when the pool robot detects that the turbidity of the water is moderate or severe, it will autonomously activate the water purification module.
[0093] When performing water purification, the device can move along a planned path within the pool, move randomly, or stop at a preset location.
[0094] Optionally, the execution priority of user commands and water turbidity can be set. If it is determined that the water purification module does not need to be started based on the water turbidity, but a user command is received, the water purification module will be started based on the user command.
[0095] Optionally, after receiving a user's command, the pool robot can determine whether the water purification module is turned on. If it is turned on, it will perform water purification actions. If it detects that the water purification module is not turned on, it will send a reminder message to the client corresponding to the pool robot to remind the user that the water purification has malfunctioned or to remind the user to replace the purification module.
[0096] Specifically, the reminder message can also remind users to clean, for example, prompting them to turn on the pool circulation pump, perform water purification, or replace the encrypted filter box for cleaning (at this time there may not be much trash, but there is more flocculent matter), etc.
[0097] In this way, the pool robot can activate the water purification module in different ways, achieving diversity in the timing of water purification activation.
[0098] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.
[0099] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.
Claims
1. A control method for a swimming pool robot, characterized in that, The pool robot moves on a cleaning surface to perform cleaning operations on the pool. The cleaning surface includes one or more of the pool wall, pool bottom, and water surface. The pool robot is equipped with at least one image acquisition device, which is used to acquire images of the pool to identify objects within the pool. The method includes: After identifying a first object based on at least one frame of a first image, if the turbidity of the pool water is moderate, the pool robot is controlled to move towards a first position close to the first object, so that the image acquisition device acquires at least one frame of a second image containing the first position during the movement; wherein, the turbidity of the water corresponding to the moderate level is greater than or equal to a first threshold and less than or equal to a second threshold; the first position refers to the position of the first object when it is identified based on the first image. The movement of the pool robot is controlled based on the object recognition result of at least one frame of the second image.
2. The method according to claim 1, characterized in that, The images acquired by the image acquisition device are also used to identify object types, including garbage that needs to be cleaned and obstacles that need to be avoided.
3. The method according to claim 2, characterized in that, The method further includes: controlling the pool robot to move along a bow-shaped path in the pool to identify trash in the pool; If, based on at least one frame of a first image, a first object is identified and the first object is trash, controlling the pool robot to move towards a first position closer to the first object, so that the image acquisition device acquires at least one frame of a second image containing the first position during the movement, includes: The pool robot is controlled to move toward the first position, such that during the movement of the pool robot toward the first position or when it moves a preset distance, the image acquisition device acquires at least one frame of a second image containing the first position.
4. The method according to claim 3, characterized in that, The control of the pool robot's movement based on the object recognition result of the at least one frame of the second image includes: If a second object is identified based on at least one frame of the second image and the second object is trash, then the pool robot is controlled to clean the trash based on a second position; the second position refers to the location of the second object when it is identified based on the second image. Alternatively, if the second object is not identified based on at least one frame of the second image, the pool robot is controlled to return to the bow-shaped path; Alternatively, if a second object is identified based on the at least one frame of the second image and the second object is an obstacle, then the pool robot is controlled to return to the bow-shaped path.
5. The method according to claim 2, characterized in that, If, based on at least one frame of a first image, a first object is identified and the first object is an obstacle, the pool robot is controlled to move towards a first position closer to the first object, so that the image acquisition device acquires at least one frame of a second image containing the first position during the movement, including: The pool robot is controlled to continue moving along the direction of travel when the first object is identified based on the first image, so that the image acquisition device can acquire at least one frame of a second image containing the first position during the movement or when the robot has moved a preset distance.
6. The method according to claim 5, characterized in that, The control of the pool robot's movement based on the object recognition result of the at least one frame of the second image includes: If no second object is detected based on at least one frame of the second image, the pool robot is controlled to continue moving in the direction of travel when the first object was detected based on the first image. Alternatively, if a second object is identified based on the at least one frame of the second image and the second object does not affect the movement of the pool robot, then the pool robot is controlled to continue moving along the direction of travel when the first object was identified based on the first image; Alternatively, if a second object is identified based on the at least one frame of the second image and the second object is an obstacle affecting the movement of the pool robot, then the pool robot is controlled to initiate an obstacle avoidance operation based on the second position, where the second position refers to the location of the second object when it is identified based on the second image.
7. The method according to claim 5, characterized in that, The control of the pool robot to continue moving along the direction of travel when the first object is identified based on the first image includes: The pool robot is controlled to continue moving at a second speed along the direction of travel when the first object is recognized based on the first image. The second speed is less than the first speed, where the first speed is the moving speed of the pool robot when the first object is recognized based on the first image.
8. The method according to claim 1, characterized in that, The water turbidity also includes low degree, which corresponds to the water turbidity being less than a first threshold; when the water turbidity of the pool is low, after determining that a first object is identified based on at least one frame of the first image, the pool robot is controlled to execute a movement strategy for the first object. And / or, the water turbidity also includes severe turbidity, where severe turbidity corresponds to water turbidity greater than the second threshold; in the case where the water turbidity also includes severe turbidity, after determining that the first object has been identified based on at least one frame of the first image, the pool robot is controlled to continue moving along the direction of travel when the first object was identified based on the first image.
9. The method according to claim 1 or 8, characterized in that, The method further includes: Based on the received user command or based on the turbidity of the water body, the water purification module is activated to perform water purification actions; And / or, send the water turbidity to the client corresponding to the pool robot, so as to display the water turbidity periodically or in real time on the client; And / or, the turbidity of the pool water can be identified through images acquired by the image acquisition device.
10. The method according to claim 8, characterized in that, Also includes: When the turbidity of the water is low, the location of the swimming pool robot is jointly identified by the water depth data collected by the water depth sensor installed on the swimming pool robot and the image collected by the image acquisition device. The location is either below or above the water surface. If the turbidity of the pool water is detected to be moderate or severe, the location of the pool robot is identified by the water depth data collected by the water depth sensor installed on the pool robot.