Laminar flow intelligent control method and system for edgeless swimming pool

Abstract

This invention discloses a laminar flow intelligent control method and system for infinity pools. The method includes: acquiring image data within the edge area of ​​the infinity pool; identifying human motion information corresponding to the image data based on a human motion recognition model; identifying water flow disturbance information corresponding to the image data based on a water flow parameter recognition model; determining laminar flow control commands for the infinity pool based on a preset motion disturbance analysis model, according to the human motion information and the water flow disturbance information; the laminar flow control commands are used to control the laminar flow propellers of the infinity pool. Therefore, this invention can achieve intelligent laminar flow control based on image perception and joint analysis of human motion and water flow disturbance, improving the water flow comfort and swimming experience of infinity pools, and providing users with a better swimming experience.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a laminar flow intelligent control method and system for infinity pools. Background Technology

[0002] With the rapid popularization of infinity pools in high-end resorts, private villas, and upscale leisure venues, users and operators are increasingly focusing on improving swimming comfort and overall experience through intelligent water flow control. A key technical challenge is achieving precise laminar flow regulation to avoid discomfort. Existing technologies typically adjust laminar flow by acquiring information from fixed sensors in the pool and using fixed water flow speed control rules or user commands. However, existing solutions lack joint dynamic analysis of human motion and water flow disturbance information, as well as the generation of intelligent laminar flow control commands driven by motion disturbance analysis models. This makes it difficult to achieve adaptive adjustment based on swimmers' real-time behavior. Commonly used static or single-dimensional control strategies cannot adapt to different swimming movements and multi-user scenarios, resulting in insufficient water flow comfort and swimming experience. Furthermore, uncorrected water flow disturbances can easily lead to user discomfort or safety hazards, limiting the level of intelligence and user satisfaction in infinity pools. Therefore, existing technologies have shortcomings that urgently need to be addressed. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a laminar flow intelligent control method and system for infinity pools, which can realize intelligent laminar flow control based on image perception and joint analysis of human movement and water flow disturbance, improve the water flow comfort and swimming experience of infinity pools, and provide users with a better swimming experience.

[0004] To address the aforementioned technical problems, the first aspect of this invention discloses a laminar flow intelligent control method for infinity pools, the method comprising: Acquire image data within the edge region of an infinity pool; Based on the human motion recognition model, the human motion information corresponding to the image data is identified; Based on the water flow parameter recognition model, the water flow disturbance information corresponding to the image data is identified; Based on a preset motion disturbance analysis model, laminar flow control commands for the infinity pool are determined according to the human motion information and the water flow disturbance information; the laminar flow control commands are used to control the laminar flow propellers of the infinity pool.

[0005] As an optional implementation, in the first aspect of the present invention, the step of identifying human motion information corresponding to the image data based on a human motion recognition model includes: The image data is input into a human image segmentation model to obtain at least one human image region in the image data; Determine the human action corresponding to each of the aforementioned human image regions; By statistically analyzing the human actions corresponding to all the human image regions, the human action information corresponding to the image data is obtained.

[0006] As an optional implementation, in the first aspect of the invention, determining the human action corresponding to each of the human image regions includes: For each human image region, the human image region is input into a trained pose classifier to obtain the pose type of the human image region. When the posture type is swimming stroke, the human image region is input into the trained swimming stroke recognition model to obtain the output predicted swimming stroke, which is determined to be the human action corresponding to the human image region. When the posture type is a static posture, the human image region is input into the trained pool resting action recognition model to obtain the output resting action, which is determined as the human action corresponding to the human image region; the resting action is floating, sitting by the pool, leaning against the pool, or standing in the pool.

[0007] As an optional implementation, in the first aspect of the present invention, the step of statistically analyzing the human actions corresponding to all the human image regions to obtain the human action information corresponding to the image data includes: For each preset human action type, calculate the number of times that human action type appears in all the human actions corresponding to all the human image regions; Calculate the average distance between the regions of the human image where the human action type has appeared, and obtain the density parameter corresponding to the human action type; Calculate the product of the occurrence count and the density parameter to obtain the type priority corresponding to the human action type; The occurrence count and type priority of all human action types with a occurrence count greater than zero are determined as the human action information corresponding to the image data.

[0008] As an optional implementation, in the first aspect of the present invention, the step of identifying the water flow disturbance information corresponding to the image data based on the water flow parameter identification model includes: The image data is input into a trained poolside image segmentation model to obtain multiple image parts that are equally spaced along the predicted pool edge. Each of the aforementioned image portions is input into the trained pool water level recognition model to obtain the predicted water level height corresponding to each of the aforementioned image portions. Based on the predicted water level height corresponding to all the image portions, calculate the water level fluctuation parameters; The image data is input into a trained pool flow velocity recognition model to obtain the output predicted flow velocity. Based on the water level fluctuation parameters and the predicted flow velocity, and using a preset mathematical relationship model for water flow disturbance, water flow disturbance information is calculated.

[0009] As an optional implementation, in the first aspect of the invention, calculating the water level fluctuation parameters based on the predicted water level height corresponding to all of the image portions includes: For any two image portions whose image positions are adjacent in the image data, calculate the height difference between the predicted water level heights of the two image portions; Calculate the distance between the image positions of the two image portions; Calculate the ratio of the height difference to the distance to obtain the water level fluctuation parameters corresponding to the two image portions; The average value of the water level fluctuation parameters corresponding to all pairs of adjacent image portions is calculated to obtain the water level fluctuation parameters.

[0010] As an optional implementation, in the first aspect of the present invention, the mathematical relationship model of water flow disturbance is obtained by fitting through the following steps: Obtain water level fluctuation data and flow velocity data corresponding to multiple historical perceptual fluctuation data of the infinity pool; Based on the iterative fitting algorithm, the coefficients of the polynomial mathematical expressions between the water level fluctuation data and the flow velocity data corresponding to all the historical perceptual fluctuation data are fitted to obtain the mathematical relationship model of the water flow disturbance.

[0011] As an optional implementation, in the first aspect of the present invention, determining the laminar flow control command of the infinity pool based on a preset motion disturbance analysis model, according to the human motion information and the water flow disturbance information, includes: Based on a preset disturbance power correspondence, a first power parameter that is inversely proportional to the water flow disturbance information is determined; The human motion information is input into a trained pool swimming demand prediction model to obtain the output swimming demand; the swimming demand is used to characterize the overall swimming demand of people in the infinity pool. Based on a preset power demand correspondence, a second power parameter that is proportional to the swimming demand is determined. The laminar output power of the infinity pool is obtained by calculating the average value of the first power parameter and the second power parameter. Generate laminar flow control commands that include the laminar flow output power.

[0012] A second aspect of this invention discloses a laminar flow intelligent control system for an infinity pool, the system comprising: The acquisition module is used to acquire image data within the edge area of ​​the infinity pool; The first recognition module is used to recognize human action information corresponding to the image data based on a human action recognition model. The second recognition module is used to identify the water flow disturbance information corresponding to the image data based on the water flow parameter recognition model. The control module is used to determine the laminar flow control command of the infinity pool based on a preset motion disturbance analysis model, according to the human motion information and the water flow disturbance information; the laminar flow control command is used to control the laminar flow propeller of the infinity pool.

[0013] As an optional implementation, in a second aspect of the present invention, the specific method by which the first recognition module identifies human motion information corresponding to the image data based on a human motion recognition model includes: The image data is input into a human image segmentation model to obtain at least one human image region in the image data; Determine the human action corresponding to each of the aforementioned human image regions; By statistically analyzing the human actions corresponding to all the human image regions, the human action information corresponding to the image data is obtained.

[0014] As an optional implementation, in a second aspect of the invention, the first recognition module determines the specific manner in which it determines the human action corresponding to each of the human image regions, including: For each human image region, the human image region is input into a trained pose classifier to obtain the pose type of the human image region. When the posture type is swimming stroke, the human image region is input into the trained swimming stroke recognition model to obtain the output predicted swimming stroke, which is determined to be the human action corresponding to the human image region. When the posture type is a static posture, the human image region is input into the trained pool resting action recognition model to obtain the output resting action, which is determined as the human action corresponding to the human image region; the resting action is floating, sitting by the pool, leaning against the pool, or standing in the pool.

[0015] As an optional implementation, in a second aspect of the present invention, the specific method by which the first recognition module statistically analyzes human actions corresponding to all the human image regions to obtain human action information corresponding to the image data includes: For each preset human action type, calculate the number of times that human action type appears in all the human actions corresponding to all the human image regions; Calculate the average distance between the regions of the human image where the human action type has appeared, and obtain the density parameter corresponding to the human action type; Calculate the product of the occurrence count and the density parameter to obtain the type priority corresponding to the human action type; The occurrence count and type priority of all human action types with a occurrence count greater than zero are determined as the human action information corresponding to the image data.

[0016] As an optional implementation, in a second aspect of the invention, the second identification module identifies the water flow disturbance information corresponding to the image data based on a water flow parameter identification model in the following specific ways: The image data is input into a trained poolside image segmentation model to obtain multiple image parts that are equally spaced along the predicted pool edge. Each of the aforementioned image portions is input into the trained pool water level recognition model to obtain the predicted water level height corresponding to each of the aforementioned image portions. Based on the predicted water level height corresponding to all the image portions, calculate the water level fluctuation parameters; The image data is input into a trained pool flow velocity recognition model to obtain the output predicted flow velocity. Based on the water level fluctuation parameters and the predicted flow velocity, and using a preset mathematical relationship model for water flow disturbance, water flow disturbance information is calculated.

[0017] As an optional implementation, in a second aspect of the invention, the second identification module calculates water level fluctuation parameters based on the predicted water level height corresponding to all the image portions in a specific manner, including: For any two image portions whose image positions are adjacent in the image data, calculate the height difference between the predicted water level heights of the two image portions; Calculate the distance between the image positions of the two image portions; Calculate the ratio of the height difference to the distance to obtain the water level fluctuation parameters corresponding to the two image portions; The average value of the water level fluctuation parameters corresponding to all pairs of adjacent image portions is calculated to obtain the water level fluctuation parameters.

[0018] As an optional implementation, in a second aspect of the invention, the mathematical relationship model of water flow disturbance is obtained by fitting through the following steps: Obtain water level fluctuation data and flow velocity data corresponding to multiple historical perceptual fluctuation data of the infinity pool; Based on the iterative fitting algorithm, the coefficients of the polynomial mathematical expressions between the water level fluctuation data and the flow velocity data corresponding to all the historical perceptual fluctuation data are fitted to obtain the mathematical relationship model of the water flow disturbance.

[0019] As an optional implementation, in a second aspect of the invention, the control module, based on a preset motion disturbance analysis model, determines the specific method of the laminar flow control command for the infinity pool according to the human motion information and the water flow disturbance information, including: Based on a preset disturbance power correspondence, a first power parameter that is inversely proportional to the water flow disturbance information is determined; The human motion information is input into a trained pool swimming demand prediction model to obtain the output swimming demand; the swimming demand is used to characterize the overall swimming demand of people in the infinity pool. Based on a preset power demand correspondence, a second power parameter that is proportional to the swimming demand is determined. The laminar output power of the infinity pool is obtained by calculating the average value of the first power parameter and the second power parameter. Generate laminar flow control commands that include the laminar flow output power.

[0020] A third aspect of this invention discloses another intelligent laminar flow control system for infinity pools, the system comprising: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute some or all of the steps in the laminar flow intelligent control method for infinity pools disclosed in the first aspect of the present invention.

[0021] The fourth aspect of the present invention discloses a computer storage medium storing computer instructions, which, when invoked, are used to execute some or all of the steps in the laminar flow intelligent control method for an infinity pool disclosed in the first aspect of the present invention.

[0022] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: This invention acquires image data within the edge area of ​​an infinity pool, identifies human motion information based on a human motion recognition model, identifies water flow disturbance information based on a water flow parameter recognition model, and determines laminar flow control commands based on a motion disturbance analysis model. This enables intelligent laminar flow control based on image perception and joint analysis of human motion and water flow disturbance, improving the water flow comfort and swimming experience of infinity pools and providing users with a better swimming experience. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating a laminar flow intelligent control method for an infinity pool disclosed in an embodiment of the present invention.

[0025] Figure 2 This is a schematic diagram of the structure of a laminar flow intelligent control system for an infinity pool disclosed in an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of another laminar flow intelligent control system for an infinity pool disclosed in an embodiment of the present invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0029] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0030] This invention discloses a laminar flow intelligent control method and system for infinity pools. By acquiring image data within the edge area of ​​the infinity pool, it identifies human motion information based on a human motion recognition model, identifies water flow disturbance information based on a water flow parameter recognition model, and determines laminar flow control commands based on a motion disturbance analysis model. This enables intelligent laminar flow control based on image perception and joint analysis of human motion and water flow disturbances, improving the water flow comfort and swimming experience of the infinity pool and providing users with a better swimming experience. Detailed descriptions follow.

[0031] Example 1 Please see Figure 1 , Figure 1 This is a flowchart illustrating a laminar flow intelligent control method for an infinity pool disclosed in an embodiment of the present invention. Figure 1 The described laminar flow intelligent control method for infinity pools can be applied to data processing systems / data processing equipment / data processing servers (wherein, the server includes local processing servers or cloud processing servers). For example... Figure 1 As shown, the laminar flow intelligent control method for this infinity pool can include the following operations: 101. Obtain image data within the edge area of ​​the infinity pool.

[0032] Optionally, the edge area of ​​the infinity pool can be the overflow channel area of ​​the four walls of the infinity pool, the buffer zone outside the transparent acrylic water barrier of the pool wall, or the physical boundary line where the top of the pool wall meets the drainage grid. This invention does not limit the scope of the invention.

[0033] Optionally, the image data may be a 2D high-definition video frame sequence acquired by a high-resolution industrial color RGB camera, an infrared three-dimensional depth image data stream generated based on a Time-of-File (ToF) sensor, or a cross-medium photoelectric image frame combined with an underwater infrared detector; the present invention does not limit the scope of the data.

[0034] 102. Based on the human motion recognition model, identify the human motion information corresponding to image data. Optionally, the human motion recognition model can be a multi-skeleton node pose behavior inference network based on the spatiotemporal graph convolutional network ST-GCN, a video-level action classification model based on 3D-ResNet, or an end-to-end action sequence recognition architecture of TimeSformer with integrated self-attention mechanism. This invention does not limit the model.

[0035] 103. Based on the water flow parameter identification model, identify the water flow disturbance information corresponding to the image data. Optionally, the water flow parameter identification model can be a dynamic quantization network for water surface ripples that integrates dense optical flow method and multi-scale residual feature extraction operator, or a fluid velocity regression network that tracks the pixel displacement field of floating bubbles and water mist on the water surface. This invention does not limit the model.

[0036] 104. Based on the preset motion disturbance analysis model, determine the laminar flow control command for the infinity pool according to human motion information and water flow disturbance information.

[0037] Optionally, laminar flow control commands are used to control the laminar flow propellers of the infinity pool.

[0038] Optionally, the motion disturbance analysis model can be a dynamic allocation decision network based on a multilayer perceptron regressor, an adaptive fuzzy expert control system, or a flow field stability configuration strategy network based on a deep reinforcement learning (DQN) architecture; this invention does not limit the model.

[0039] Optionally, the laminar flow propulsion device can be a variable frequency submersible axial flow pump, a hydraulic impeller propulsion device driven by a permanent magnet synchronous motor, or a multi-stage water jet drive pump set with a rectifier grid; the present invention does not limit the type of propulsion device.

[0040] As can be seen, the above-described embodiments of the invention acquire image data within the edge area of ​​an infinity pool, identify human motion information based on a human motion recognition model, identify water flow disturbance information based on a water flow parameter recognition model, and determine laminar flow control commands based on a motion disturbance analysis model. This enables intelligent laminar flow control based on image perception and joint analysis of human motion and water flow disturbance, improving the water flow comfort and swimming experience of the infinity pool and providing users with a better swimming experience.

[0041] As an optional embodiment, the step above, identifying human motion information corresponding to image data based on a human motion recognition model, includes: The image data is input into the human image segmentation model to obtain at least one human image region in the image data; Determine the human action corresponding to each human image region; By statistically analyzing the human movements corresponding to all human image regions, we can obtain the human movement information corresponding to the image data.

[0042] Optionally, the human image segmentation model can be a DeepLabv3+ network based on dilated convolution and feature pyramids, a Mask R-CNN architecture with instance segmentation capabilities, or a lightweight real-time human segmentation network PP-HumanSeg specifically optimized and fine-tuned for water refraction environments. This invention does not limit the specific network.

[0043] Optionally, the human image region can be a pixel-level mask region containing the edge of the complete human torso, or a two-dimensional bounding box rectangle that tightly encloses the swimmer's body; the present invention does not limit this.

[0044] As can be seen, through the above optional embodiments, by inputting image data into the human image segmentation model to obtain human image regions, determining the human actions corresponding to each region and statistically obtaining human action information, accurate statistics of human actions based on image segmentation are achieved, improving the accuracy and comprehensiveness of action information extraction, and reducing the risk of action recognition deviation caused by multiple people mixed in the image.

[0045] As an optional embodiment, the step of determining the human action corresponding to each human image region in the above steps includes: For each human image region, the human image region is input into the trained pose classifier to obtain the pose type of the human image region. When the posture type is swimming stroke, the human image region is input into the trained swimming stroke recognition model to obtain the output predicted swimming stroke, which is determined to be the human action corresponding to the human image region. When the pose type is static, the human image region is input into the trained pool resting action recognition model to obtain the output resting action, which is determined to be the human action corresponding to the human image region.

[0046] Optional resting postures include floating, sitting by the pool, leaning against the pool, or standing in the pool.

[0047] Optionally, the pose classifier can be a support vector machine (SVM) classifier based on the joint features of the human skeleton aspect ratio and movement speed, or a multilayer linear discriminant component based on the Softmax function; the present invention does not limit this.

[0048] Optionally, the swimming stroke recognition model can be a temporal convolutional network (TCN) pre-trained for specific skeletal motion trajectories for breaststroke, freestyle, backstroke, and butterfly, but this invention does not limit it.

[0049] Optionally, the pool resting action recognition model can be a heuristic classifier that combines the vertical height threshold of the human body's center of mass with the variance analysis of the motion displacement, or a behavior pause determination regression model based on the Long Short-Term Memory (LSTM) network. This invention does not limit the model.

[0050] As can be seen, through the above optional embodiments, by inputting each human image region into the posture classifier to obtain the posture type, inputting it into the swimming posture recognition model to obtain the predicted swimming posture when swimming, and inputting it into the resting posture recognition model to obtain the resting posture when in a static posture, it is possible to realize the classification of human motion based on posture classification, improve the targeting and accuracy of motion recognition, and reduce the risk of recognition errors caused by confusion of motion types.

[0051] As an optional embodiment, the step above, which involves statistically analyzing the human actions corresponding to all human image regions to obtain human action information corresponding to the image data, includes: For each preset human action type, calculate the number of times that human action type appears in the corresponding human actions of all human image regions; Calculate the average distance between human image regions where this human action type has appeared, and obtain the density parameters corresponding to this human action type; Calculate the product of the frequency of occurrence and the dense parameters to obtain the type priority corresponding to the human action type; The occurrence count and type priority of all human action types with a greater than zero occurrence count are determined as the human action information corresponding to the image data.

[0052] Specifically, the types of human movements can be the swimming strokes or resting movements mentioned in the above scheme.

[0053] As can be seen, through the above optional embodiments, the type priority is obtained by calculating the product of the occurrence frequency and density parameters of each human action type, and the occurrence frequency and priority of all action types with a occurrence frequency greater than zero are determined as human action information. This realizes the quantitative evaluation of action information based on frequency and spatial density, improves the rationality and representativeness of action information representation, and reduces the risk of information bias caused by single statistics.

[0054] As an optional embodiment, the step above, identifying the water flow disturbance information corresponding to the image data based on the water flow parameter identification model, includes: The image data is input into a trained poolside image segmentation model to obtain multiple image parts that are equally spaced along the predicted pool edge. Each image portion is input into the trained pool water level recognition model to obtain the predicted water level height corresponding to each output image portion; Calculate water level fluctuation parameters based on the predicted water level height corresponding to all image portions; Image data is input into a trained pool flow velocity recognition model to obtain the output predicted flow velocity; Based on the water level fluctuation parameters and predicted flow velocity, and using a pre-defined mathematical model of water flow disturbance, the water flow disturbance information is calculated.

[0055] Optionally, the poolside image segmentation model can be a U-Net semantic segmentation network that integrates edge extraction operators, a Hough transform block model established for the boundary features of water and land, or a pre-trained SAM (Segment Anything Model) lightweight derivative architecture. This invention does not limit the model.

[0056] Optionally, the water level identification model in the pool can be a temporal convolutional regression network that captures the relative height changes of overflow scale pixels on the pool wall, or an absolute water level regression mapping model based on the overflow water flow shadow boundary features. This invention does not limit the model.

[0057] Optionally, the flow velocity identification model in the pool can be a particle image velocimetry (PIV) depth feature grid, or an optical flow regression network that estimates velocity vectors based on pixel-level displacement fields of bubbles and disturbance ripples on the water surface. This invention does not limit the model.

[0058] As can be seen, through the above optional embodiments, by inputting image data into the poolside image segmentation model to obtain multiple image parts, inputting them into the pool water level identification model to obtain the predicted water level height and calculate the water level fluctuation parameters, and inputting them into the pool flow velocity identification model to obtain the predicted flow velocity, and calculating the flow disturbance information based on the mathematical relationship model of water flow disturbance, the accurate identification of water flow disturbance jointly by water level and flow velocity is achieved, improving the scientificity and completeness of disturbance information calculation, and reducing the risk of disturbance assessment deviation caused by single parameter analysis.

[0059] As an optional embodiment, the step above, calculating the water level fluctuation parameters based on the predicted water level height corresponding to all image portions, includes: For any two image portions that are adjacent in image data, calculate the height difference between the predicted water level heights of the two image portions; Calculate the distance between the image locations of the two image parts; Calculate the ratio of height difference to distance to obtain the water level fluctuation parameters corresponding to the two image portions; The average value of the water level fluctuation parameters corresponding to all pairs of adjacent image portions is calculated to obtain the water level fluctuation parameters.

[0060] As can be seen, through the above optional embodiments, the water level fluctuation parameter is obtained by calculating the ratio of the height difference to the distance of the predicted water level height in adjacent image parts, and the average value is calculated to obtain the water level turbulence parameter. This realizes the dynamic water level disturbance quantification based on the water level change in adjacent areas, improves the sensitivity and accuracy of the water level turbulence parameter, and reduces the risk of ignoring local disturbances caused by global averaging.

[0061] As an optional embodiment, the mathematical relationship model of water flow disturbance in the above steps is obtained by fitting through the following steps: Obtain water level fluctuation data and flow velocity data corresponding to multiple historical perceptual fluctuation data of the infinity pool; Based on the iterative fitting algorithm, the coefficients of the polynomial mathematical expressions between the water level fluctuation data and the flow velocity data corresponding to all historical perceptual fluctuation data are fitted to obtain the mathematical relationship model of water flow disturbance.

[0062] Optionally, the historical somatosensory fluctuation data can be the scalar of water flow impact force collected by multi-axis accelerometers deployed inside the infinity pool, the dynamic data of surge period pressure recorded by high-precision pressure gauges, or the discrete grading label of somatosensory flow velocity comfort manually entered by professional swimmers. This invention does not limit the data.

[0063] Optionally, the iterative fitting algorithm can be a nonlinear least squares feedback algorithm, a Gauss-Newton iterative convergence method, or a ridge regression polynomial fitting strategy with regularization constraints; this invention does not limit the specific algorithm.

[0064] As can be seen, through the above optional embodiments, by obtaining water level fluctuation data and flow velocity data corresponding to historical perceptual fluctuation data, and fitting a polynomial mathematical expression based on an iterative fitting algorithm to obtain a mathematical relationship model of water flow disturbance, accurate water flow disturbance relationship modeling based on historical data is achieved, improving the model's fitting accuracy and generalization ability, and reducing the risk of relationship deviation caused by empirical formulas.

[0065] As an optional embodiment, the step described above, determining the laminar flow control command for the infinity pool based on a preset motion disturbance analysis model and according to human motion information and water flow disturbance information, includes: Based on the preset disturbance power correspondence, a first power parameter that is inversely proportional to the water flow disturbance information is determined; Human motion information is input into a trained pool swimming demand prediction model to obtain the output swimming demand; optionally, the swimming demand is used to characterize the overall swimming demand of people in an infinity pool. Based on the preset demand-power correspondence, a second power parameter that is proportional to the swimming demand is determined. The laminar output power of the infinity pool is obtained by calculating the average value of the first power parameter and the second power parameter. Generate laminar flow control commands that include laminar flow output power.

[0066] Optionally, the swimming demand prediction model can be a feedforward neural network based on multi-feature weighted splicing, an intent recognition and perception classifier that integrates graph network nodes, or a decision network that performs multi-dimensional weighted scoring on the current dynamic total number of people in the pool and the high-density priority posture distribution. This invention does not limit the model.

[0067] As can be seen, through the above optional embodiments, by determining the inverse first power parameter based on water flow disturbance information, obtaining the swimming demand prediction model based on human motion information to obtain the swimming demand determination second power parameter, and calculating the average of the two to obtain the laminar flow output power to generate laminar flow control commands, dynamic power optimization based on human motion and water flow disturbance is realized, improving the intelligence and comfort of laminar flow control, and reducing the risk of water flow discomfort caused by fixed power.

[0068] Example 2 Please see Figure 2 , Figure 2 This is a schematic diagram of the structure of a laminar flow intelligent control system for an infinity pool disclosed in an embodiment of the present invention. Figure 2 The described laminar flow intelligent control system for infinity pools can be applied to data processing systems / data processing equipment / data processing servers (wherein, the server includes local processing servers or cloud processing servers). For example... Figure 2 As shown, the laminar flow intelligent control system of this infinity pool may include: The acquisition module 201 is used to acquire image data within the edge area of ​​the infinity pool.

[0069] The first recognition module 202 is used to recognize human motion information corresponding to image data based on a human motion recognition model. The second recognition module 203 is used to recognize the water flow disturbance information corresponding to the image data based on the water flow parameter recognition model. The control module 204 is used to determine the laminar flow control command of the infinity pool based on the preset motion disturbance analysis model and the human motion information and water flow disturbance information.

[0070] Optionally, laminar flow control commands are used to control the laminar flow propellers of the infinity pool.

[0071] As can be seen, the above-described embodiments of the invention acquire image data within the edge area of ​​an infinity pool, identify human motion information based on a human motion recognition model, identify water flow disturbance information based on a water flow parameter recognition model, and determine laminar flow control commands based on a motion disturbance analysis model. This enables intelligent laminar flow control based on image perception and joint analysis of human motion and water flow disturbance, improving the water flow comfort and swimming experience of the infinity pool and providing users with a better swimming experience.

[0072] As an optional embodiment, the first recognition module, based on a human motion recognition model, identifies the specific method by which it recognizes the human motion information corresponding to the image data, including: The image data is input into the human image segmentation model to obtain at least one human image region in the image data; Determine the human action corresponding to each human image region; By statistically analyzing the human movements corresponding to all human image regions, we can obtain the human movement information corresponding to the image data.

[0073] As can be seen, through the above optional embodiments, by inputting image data into the human image segmentation model to obtain human image regions, determining the human actions corresponding to each region and statistically obtaining human action information, accurate statistics of human actions based on image segmentation are achieved, improving the accuracy and comprehensiveness of action information extraction, and reducing the risk of action recognition deviation caused by multiple people mixed in the image.

[0074] As an optional embodiment, the first recognition module determines the specific method of human action corresponding to each human image region, including: For each human image region, the human image region is input into the trained pose classifier to obtain the pose type of the human image region. When the posture type is swimming stroke, the human image region is input into the trained swimming stroke recognition model to obtain the output predicted swimming stroke, which is determined to be the human action corresponding to the human image region. When the posture type is static posture, the human image region is input into the trained pool resting action recognition model to obtain the output resting action, which is determined to be the human action corresponding to the human image region; the resting action is floating, sitting by the pool, leaning against the pool, or standing in the pool.

[0075] As can be seen, through the above optional embodiments, by inputting each human image region into the posture classifier to obtain the posture type, inputting it into the swimming posture recognition model to obtain the predicted swimming posture when swimming, and inputting it into the resting posture recognition model to obtain the resting posture when in a static posture, it is possible to realize the classification of human motion based on posture classification, improve the targeting and accuracy of motion recognition, and reduce the risk of recognition errors caused by confusion of motion types.

[0076] As an optional embodiment, the specific method by which the first recognition module statistically analyzes the human actions corresponding to all human image regions to obtain the human action information corresponding to the image data includes: For each preset human action type, calculate the number of times that human action type appears in the corresponding human actions of all human image regions; Calculate the average distance between human image regions where this human action type has appeared, and obtain the density parameters corresponding to this human action type; Calculate the product of the frequency of occurrence and the dense parameters to obtain the type priority corresponding to the human action type; The occurrence count and type priority of all human action types with a greater than zero occurrence count are determined as the human action information corresponding to the image data.

[0077] As can be seen, through the above optional embodiments, the type priority is obtained by calculating the product of the occurrence frequency and density parameters of each human action type, and the occurrence frequency and priority of all action types with a occurrence frequency greater than zero are determined as human action information. This realizes the quantitative evaluation of action information based on frequency and spatial density, improves the rationality and representativeness of action information representation, and reduces the risk of information bias caused by single statistics.

[0078] As an optional embodiment, the second recognition module identifies the specific method by which it identifies the water flow disturbance information corresponding to the image data based on the water flow parameter recognition model, including: The image data is input into a trained poolside image segmentation model to obtain multiple image parts that are equally spaced along the predicted pool edge. Each image portion is input into the trained pool water level recognition model to obtain the predicted water level height corresponding to each output image portion; Calculate water level fluctuation parameters based on the predicted water level height corresponding to all image portions; Image data is input into a trained pool flow velocity recognition model to obtain the output predicted flow velocity; Based on the water level fluctuation parameters and predicted flow velocity, and using a pre-defined mathematical model of water flow disturbance, the water flow disturbance information is calculated.

[0079] As can be seen, through the above optional embodiments, by inputting image data into the poolside image segmentation model to obtain multiple image parts, inputting them into the pool water level identification model to obtain the predicted water level height and calculate the water level fluctuation parameters, and inputting them into the pool flow velocity identification model to obtain the predicted flow velocity, and calculating the flow disturbance information based on the mathematical relationship model of water flow disturbance, the accurate identification of water flow disturbance jointly by water level and flow velocity is achieved, improving the scientificity and completeness of disturbance information calculation, and reducing the risk of disturbance assessment deviation caused by single parameter analysis.

[0080] As an optional embodiment, the second recognition module calculates the water level fluctuation parameters based on the predicted water level height corresponding to all image portions in the following specific ways: For any two image portions that are adjacent in image data, calculate the height difference between the predicted water level heights of the two image portions; Calculate the distance between the image locations of the two image parts; Calculate the ratio of height difference to distance to obtain the water level fluctuation parameters corresponding to the two image portions; The average value of the water level fluctuation parameters corresponding to all pairs of adjacent image portions is calculated to obtain the water level fluctuation parameters.

[0081] As can be seen, through the above optional embodiments, the water level fluctuation parameter is obtained by calculating the ratio of the height difference to the distance of the predicted water level height in adjacent image parts, and the average value is calculated to obtain the water level turbulence parameter. This realizes the dynamic water level disturbance quantification based on the water level change in adjacent areas, improves the sensitivity and accuracy of the water level turbulence parameter, and reduces the risk of ignoring local disturbances caused by global averaging.

[0082] As an optional embodiment, the mathematical relationship model of water flow disturbance is obtained by fitting through the following steps: Obtain water level fluctuation data and flow velocity data corresponding to multiple historical perceptual fluctuation data of the infinity pool; Based on the iterative fitting algorithm, the coefficients of the polynomial mathematical expressions between the water level fluctuation data and the flow velocity data corresponding to all historical perceptual fluctuation data are fitted to obtain the mathematical relationship model of water flow disturbance.

[0083] As can be seen, through the above optional embodiments, by obtaining water level fluctuation data and flow velocity data corresponding to historical perceptual fluctuation data, and fitting a polynomial mathematical expression based on an iterative fitting algorithm to obtain a mathematical relationship model of water flow disturbance, accurate water flow disturbance relationship modeling based on historical data is achieved, improving the model's fitting accuracy and generalization ability, and reducing the risk of relationship deviation caused by empirical formulas.

[0084] As an optional embodiment, the control module, based on a preset motion disturbance analysis model, determines the specific method of laminar flow control commands for the infinity pool according to human motion information and water flow disturbance information, including: Based on the preset disturbance power correspondence, a first power parameter that is inversely proportional to the water flow disturbance information is determined; Human motion information is input into a trained pool swimming demand prediction model to obtain the output swimming demand; optionally, the swimming demand is used to characterize the overall swimming demand of people in an infinity pool. Based on the preset demand-power correspondence, a second power parameter that is proportional to the swimming demand is determined. The laminar output power of the infinity pool is obtained by calculating the average value of the first power parameter and the second power parameter. Generate laminar flow control commands that include laminar flow output power.

[0085] As can be seen, through the above optional embodiments, by determining the inverse first power parameter based on water flow disturbance information, obtaining the swimming demand prediction model based on human motion information to obtain the swimming demand determination second power parameter, and calculating the average of the two to obtain the laminar flow output power to generate laminar flow control commands, dynamic power optimization based on human motion and water flow disturbance is realized, improving the intelligence and comfort of laminar flow control, and reducing the risk of water flow discomfort caused by fixed power.

[0086] Example 3 Please see Figure 3 , Figure 3 This is yet another laminar flow intelligent control system for an infinity pool disclosed in this invention. Figure 3 The described laminar flow intelligent control system for the infinity pool is applied in a data processing system / data processing equipment / data processing server (wherein, the server includes a local processing server or a cloud processing server). For example... Figure 3 As shown, the laminar flow intelligent control system of this infinity pool may include: Memory 301 storing executable program code; Processor 302 coupled to memory 301; The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the laminar flow intelligent control method for the infinity pool described in Embodiment 1.

[0087] Example 4 This invention discloses a computer read storage medium that stores a computer program for electronic data interchange, wherein the computer program causes a computer to execute the steps of the laminar flow intelligent control method for an infinity pool described in Embodiment 1.

[0088] Example 5 This invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform the steps of the laminar flow intelligent control method for an infinity pool described in Embodiment 1.

[0089] The foregoing has described specific embodiments of this specification; other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than those shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily have to follow the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0090] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0091] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware components.

[0092] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0093] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0094] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0095] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0096] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0097] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0098] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0099] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0100] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0101] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0102] Finally, it should be noted that the laminar flow intelligent control method and system for an infinity pool disclosed in the embodiments of the present invention are merely preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, not to limit it. Although the present invention 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 the present invention.

Claims

1. A laminar flow intelligent control method for an infinity pool, characterized in that, The method includes: Acquire image data within the edge region of an infinity pool; Based on the human motion recognition model, the human motion information corresponding to the image data is identified; Based on the water flow parameter recognition model, the water flow disturbance information corresponding to the image data is identified; Based on a preset motion disturbance analysis model, laminar flow control commands for the infinity pool are determined according to the human motion information and the water flow disturbance information; the laminar flow control commands are used to control the laminar flow propellers of the infinity pool.

2. The laminar flow intelligent control method for an infinity pool according to claim 1, characterized in that, The method of identifying human motion information corresponding to the image data based on the human motion recognition model includes: The image data is input into a human image segmentation model to obtain at least one human image region in the image data; Determine the human action corresponding to each of the aforementioned human image regions; By statistically analyzing the human actions corresponding to all the human image regions, the human action information corresponding to the image data is obtained.

3. The laminar flow intelligent control method for an infinity pool according to claim 2, characterized in that, Determining the human action corresponding to each of the human image regions includes: For each human image region, the human image region is input into a trained pose classifier to obtain the pose type of the human image region. When the posture type is swimming stroke, the human image region is input into the trained swimming stroke recognition model to obtain the output predicted swimming stroke, which is determined to be the human action corresponding to the human image region. When the posture type is a static posture, the human image region is input into the trained pool resting action recognition model to obtain the output resting action, which is determined as the human action corresponding to the human image region; the resting action is floating, sitting by the pool, leaning against the pool, or standing in the pool.

4. The laminar flow intelligent control method for an infinity pool according to claim 2, characterized in that, The step of statistically analyzing the human actions corresponding to all the human image regions to obtain the human action information corresponding to the image data includes: For each preset human action type, calculate the number of times that human action type appears in all the human actions corresponding to all the human image regions; Calculate the average distance between the regions of the human image where the human action type has appeared, and obtain the density parameter corresponding to the human action type; Calculate the product of the occurrence count and the density parameter to obtain the type priority corresponding to the human action type; The occurrence count and type priority of all human action types with a occurrence count greater than zero are determined as the human action information corresponding to the image data.

5. The laminar flow intelligent control method for an infinity pool according to claim 1, characterized in that, The water flow parameter recognition model identifies water flow disturbance information corresponding to the image data, including: The image data is input into a trained poolside image segmentation model to obtain multiple image parts that are equally spaced along the predicted pool edge. Each of the aforementioned image portions is input into the trained pool water level recognition model to obtain the predicted water level height corresponding to each of the aforementioned image portions. Based on the predicted water level height corresponding to all the image portions, calculate the water level fluctuation parameters; The image data is input into a trained pool flow velocity recognition model to obtain the output predicted flow velocity. Based on the water level fluctuation parameters and the predicted flow velocity, and using a preset mathematical relationship model for water flow disturbance, water flow disturbance information is calculated.

6. The laminar flow intelligent control method for an infinity pool according to claim 5, characterized in that, The calculation of water level fluctuation parameters based on the predicted water level height corresponding to all the image portions includes: For any two image portions whose image positions are adjacent in the image data, calculate the height difference between the predicted water level heights of the two image portions; Calculate the distance between the image positions of the two image portions; Calculate the ratio of the height difference to the distance to obtain the water level fluctuation parameters corresponding to the two image portions; The average value of the water level fluctuation parameters corresponding to all pairs of adjacent image portions is calculated to obtain the water level fluctuation parameters.

7. The laminar flow intelligent control method for an infinity pool according to claim 5, characterized in that, The mathematical relationship model of water flow disturbance is obtained by fitting through the following steps: Obtain water level fluctuation data and flow velocity data corresponding to multiple historical perceptual fluctuation data of the infinity pool; Based on the iterative fitting algorithm, the coefficients of the polynomial mathematical expressions between the water level fluctuation data and the flow velocity data corresponding to all the historical perceptual fluctuation data are fitted to obtain the mathematical relationship model of the water flow disturbance.

8. The laminar flow intelligent control method for an infinity pool according to claim 1, characterized in that, The pre-set motion disturbance analysis model determines the laminar flow control commands for the infinity pool based on the human motion information and the water flow disturbance information, including: Based on a preset disturbance power correspondence, a first power parameter that is inversely proportional to the water flow disturbance information is determined; The human motion information is input into a trained pool swimming demand prediction model to obtain the output swimming demand; the swimming demand is used to characterize the overall swimming demand of people in the infinity pool. Based on a preset power demand correspondence, a second power parameter that is proportional to the swimming demand is determined. The laminar output power of the infinity pool is obtained by calculating the average value of the first power parameter and the second power parameter. Generate laminar flow control commands that include the laminar flow output power.

9. A laminar flow intelligent control system for an infinity pool, characterized in that, The system includes: The acquisition module is used to acquire image data within the edge area of ​​the infinity pool; The first recognition module is used to recognize human action information corresponding to the image data based on a human action recognition model. The second recognition module is used to identify the water flow disturbance information corresponding to the image data based on the water flow parameter recognition model. The control module is used to determine the laminar flow control command of the infinity pool based on a preset motion disturbance analysis model, according to the human motion information and the water flow disturbance information; the laminar flow control command is used to control the laminar flow propeller of the infinity pool.

10. A laminar flow intelligent control system for an infinity pool, characterized in that, The system includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the laminar flow intelligent control method for an infinity pool as described in any one of claims 1-8.

Images