Lawn ponding recognition method, device, equipment, medium and system

Abstract

The invention relates to the technical field of lawn maintenance, and discloses a lawn ponding recognition method, device, equipment, medium and system, which realize accurate and automatic distinguishing of surface ponding and underground ponding by fusing visual perception and physical interaction data analysis based on body vibration, and remarkably overcome the problem of high misjudgment rate of a traditional method. According to the invention, the intelligent autonomous mobile equipment has the capability of detecting'underground ponding 'which is difficult to find by naked eyes and common cameras, early diagnosis of the hidden ponding problem is realized, and the method can be directly integrated with the existing equipment and can be completed in daily operation without additional expensive equipment or human input; meanwhile, the method can dynamically adapt to different lawn physical characteristics through a preset classification model or a preset evaluation rule, and further improves the robustness and environment universality of the recognition method.

Description

Technical Field

[0001] This invention relates to the field of lawn maintenance technology, and in particular to a method, device, equipment, medium and system for identifying lawn water accumulation. Background Technology

[0002] Waterlogging in lawns is a common and highly detrimental problem. It can be categorized into surface water (temporarily present on the grass blades after rain) and groundwater (hidden beneath the turf and above the soil, where the lower soil layers are saturated with moisture, making it difficult to detect on the surface). The latter is more harmful, leading to root hypoxia, root rot, and the development of diseases. Currently, intelligent autonomous mowers cannot detect groundwater accumulation, making the lawn highly susceptible to damage during mowing operations in such areas. The mowers themselves can also be damaged by water overflowing from the groundwater.

[0003] Existing methods mainly rely on: manual inspection: inefficient, experience-dependent, and unable to detect groundwater accumulation in a timely manner; visual sensors: intelligent autonomous mobile devices are usually equipped with cameras, but they can only identify obvious surface water features such as specular reflection. For groundwater, its surface visual features (dark color, seemingly flat) are easily confused with healthy lawns or other problems, resulting in a high misjudgment rate; high-altitude remote sensing: costly, poor real-time performance, and unable to physically interact with and perceive the actual underground situation, thus having limited effectiveness in identifying surface water accumulation in the daily operation scenarios of autonomous mobile devices.

[0004] Therefore, existing methods for identifying lawn water accumulation cannot accurately perceive the actual situation on the ground and underground, and there is a particular lack of reliable and automated detection methods for groundwater accumulation, a "hidden killer". Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, medium, and system for identifying water accumulation in lawns, in order to overcome the deficiencies of the prior art.

[0006] The present invention provides a method for identifying water accumulation in lawns, comprising: Acquire images of the lawn area, real-time vibration data, and current location; Based on the captured images, determine whether the lawn area is visually suspected to be a waterlogged area; If the lawn area is visually suspected to be a waterlogged area, then the current body vibration data is obtained based on the real-time vibration data; Based on the current machine vibration data, the type of water accumulation at the current location is obtained through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0007] This invention provides a method for identifying water accumulation in lawns, wherein determining whether a lawn area is visually suspected to be a water accumulation area based on a captured image includes: The lawn area is divided into multiple grids; Extract the color and texture features corresponding to each grid; Based on the color and texture features, determine whether the corresponding grid is a visually suspected water accumulation area.

[0008] The present invention provides a method for identifying water accumulation in lawns, wherein determining whether a corresponding grid is a visually suspected water accumulation area based on the color and texture features includes: Calculate the average brightness and texture variance of each grid based on color and texture features; If the average brightness is lower than a first preset threshold and the texture variance is lower than a second preset threshold, then the corresponding grid is determined to be a visually suspected water accumulation area.

[0009] This invention provides a method for identifying water accumulation in lawns, wherein obtaining current machine vibration data based on real-time vibration data includes: The real-time vibration data is subjected to coordinate transformation and gravity compensation to obtain dynamic vibration acceleration data; The frequency domain features, time domain nonlinear features, and time domain statistical features of the dynamic vibration acceleration data are extracted and used as the current body vibration data.

[0010] The present invention provides a method for identifying water accumulation in lawns, wherein the frequency domain features include the dominant frequency component and the frequency band energy ratio, the time domain nonlinear features include the sample entropy, and the time domain statistical features include at least one of the root mean square value, peak factor, and waveform factor.

[0011] The present invention provides a method for identifying water accumulation in lawns. The method further includes establishing and updating dynamic vibration characteristic baselines, comprising: in an initial stage, collecting body vibration data in each grid of the lawn area; based on the initially collected body vibration data, calculating and storing at least one vibration characteristic baseline for each grid, the vibration characteristic baseline including at least one or more of a dominant frequency baseline, a sample entropy baseline, and an energy ratio baseline; and in subsequent operations, updating the vibration characteristic baselines based on newly collected body vibration data identified as being in a healthy state.

[0012] The present invention provides a method for identifying water accumulation in lawns, wherein the establishment and updating of the dynamic vibration characteristic baseline further includes: when the lawn area is identified as being in an unhealthy state, pausing the updating of the vibration characteristic baseline of the lawn area; and in subsequent operations, when the lawn area is identified as being restored to a healthy state, resuming the updating of the vibration characteristic baseline of the lawn area.

[0013] The present invention provides a method for identifying water accumulation in lawns, wherein updating the vibration feature baseline includes updating the vibration feature baseline using an exponentially weighted moving average algorithm.

[0014] This invention provides a method for identifying water accumulation in lawns, wherein the preset evaluation rules include: Calculate the deviation of the dominant frequency component, sample entropy, and frequency band energy ratio from the corresponding dominant frequency baseline, sample entropy baseline, and energy ratio baseline, respectively; If the deviation of the main frequency component is greater than a third preset threshold, and the deviation of the sample entropy is greater than a fourth preset threshold, or / and the deviation of the frequency band energy ratio is greater than a fifth preset threshold, then the current grid is groundwater accumulation; otherwise, the current grid is surface water accumulation.

[0015] The present invention provides a method for identifying lawn water accumulation. The preset classification model is trained in the following way: collecting multiple sets of machine vibration sample data and extracting the frequency domain features, time domain nonlinear features and time domain statistical features of the sample data; labeling the lawn water accumulation type corresponding to each set of sample data; and using machine learning algorithms to train the feature vectors and labels to obtain the preset classification model.

[0016] The present invention also provides a lawn water accumulation identification device, the device comprising: The data acquisition module is used to acquire images of the lawn area, real-time vibration data, and the current location. The visual judgment module is used to determine whether the lawn area is a visually suspected water accumulation area based on the captured image; The data processing module is used to obtain the current body vibration data based on the real-time vibration data if the lawn area is a visually suspected water accumulation area. The vibration recognition module is used to obtain the type of water accumulation at the current location based on the current machine vibration data through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0017] The present invention also provides an autonomous mobile device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the above-described lawn water accumulation identification method when executing the program.

[0018] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the above-described method for identifying water accumulation in lawns.

[0019] The present invention also provides a computer program product, the computer program product comprising a computer program that can be stored on a non-transitory computer-readable storage medium, and when the computer program is executed by a processor, the computer is able to execute any of the above-described lawn water accumulation identification methods.

[0020] The present invention also provides a lawn water accumulation identification system, characterized in that it includes the aforementioned autonomous mobile device, as well as a cloud server, a user terminal, and / or an autonomous water accumulation treatment device, wherein the cloud server or user terminal is used to receive and display the water accumulation identification results; and the autonomous water accumulation treatment device is used to perform autonomous water accumulation treatment operations based on the received water accumulation identification results and the location of the water accumulation.

[0021] The present invention provides a method, apparatus, device, medium, and system for identifying water accumulation in lawns, which can bring at least the following beneficial effects: By integrating visual perception with physical interaction data based on body vibration, this method effectively overcomes the problem of high misjudgment rate in identifying lawn water accumulation using traditional visual methods. It can accurately distinguish between surface water accumulation and groundwater accumulation, achieving automated and highly reliable identification of lawn water accumulation types.

[0022] By leveraging the physical contact between autonomous mobile devices and the lawn during operation, vibration response data can be acquired in real time, thereby enabling the detection of "underground water" that is difficult to detect with the naked eye and conventional cameras, and achieving early detection and diagnosis of potential water accumulation problems in lawns.

[0023] It can be directly deployed on existing smart autonomous mobile device platforms, reusing their existing sensors and computing resources to simultaneously complete data collection and lawn water accumulation type identification during daily lawn mowing operations, without the need for additional expensive dedicated equipment or reliance on manual inspection, significantly reducing deployment and maintenance costs.

[0024] By analyzing the vibration data of the machine body through a preset classification model or preset evaluation rules, the identification benchmark can be dynamically adjusted according to the physical characteristics of different lawns, adapting to various lawn working conditions and improving the adaptability and overall identification robustness of this method in different environments. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0026] Figure 1 This is a flowchart illustrating a method for identifying water accumulation in lawns provided by the present invention.

[0027] Figure 2 This is a schematic diagram of the structure of a lawn water accumulation identification device provided by the present invention.

[0028] Figure 3 This is a schematic diagram of the structure of the autonomous mobile device provided by the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, embodiments of this invention, and should not be construed as limiting the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention. In the description of this invention, it should be understood that the terminology used is for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0030] Figure 1 This is a flowchart illustrating a method for identifying water accumulation in lawns provided by the present invention. The execution subject of this method can be an autonomous mobile device suitable for lawn operations, such as an autonomous mobile lawnmower, automatic aerator, autonomous grass planter, or autonomous weeder. The autonomous mobile device can acquire actual data while operating, utilizing its own vision module (RGB camera or multispectral imaging system), motion sensing module (inertial measurement unit IMU), and positioning module (GPS, RTK-GPS, or ranging sensor), thereby achieving lawn water accumulation identification.

[0031] See Figure 1 The present invention provides a method for identifying water accumulation in lawns, which may include: S110: Acquire images of the lawn area, real-time vibration data, and current location.

[0032] In one embodiment, images of the lawn area are acquired using an RGB camera or multispectral imaging system on the autonomous mobile device. Real-time vibration data of the autonomous mobile device within the lawn area is acquired using an IMU (Inertial Measurement Unit, containing at least a three-axis accelerometer) on the autonomous mobile device. The current location of the autonomous mobile device within the lawn area is obtained using a GPS, RTK-GPS, or ranging sensor fusion positioning module on the autonomous mobile device. The RGB camera and IMU can be configured to acquire data synchronously, achieving synchronization through hardware triggering or high-precision software timestamps to ensure that image frames and IMU data segments are aligned in time, with a time synchronization error requirement of less than 1 millisecond. Simultaneously, accurate geographical location information is obtained from the positioning module.

[0033] S120. Based on the captured image, determine whether the lawn area is a visually suspected waterlogged area.

[0034] In one embodiment, S120 includes: dividing the lawn area into multiple grids; Extract the color and texture features corresponding to each grid; Based on the color and texture features, determine whether the corresponding grid is a visually suspected water accumulation area.

[0035] The step of determining whether the corresponding grid is a visually suspected water accumulation area based on the color and texture features includes: Calculate the average brightness and texture variance of each grid based on color and texture features; If the average brightness is lower than a first preset threshold and the texture variance is lower than a second preset threshold, then the corresponding grid is determined to be a visually suspected water accumulation area.

[0036] Preferably, the first preset threshold is set to 50, and the second preset threshold is set to 10.

[0037] S130. If the lawn area is a visually suspected waterlogged area, then obtain the current body vibration data based on the real-time vibration data.

[0038] In one embodiment, S130 includes: performing coordinate transformation and gravity compensation on the real-time vibration data to obtain dynamic vibration acceleration data; Specifically, based on the body vibration data of the autonomous mobile device in the visually suspected water accumulation area, complementary filtering or Kalman filtering is used to fuse the gyroscope data and accelerometer data of the sensors used by the autonomous mobile device to collect body vibration data, so as to obtain the pitch angle θ and roll angle φ of the autonomous mobile device. Based on the pitch angle θ and roll angle φ of the autonomous mobile device, the accelerometer data is rotated from the body coordinate system to the global coordinate system (NEU) using a rotation matrix. a_global=R(θ, φ)*a_body, where a_body represents the body coordinate system and R represents the rotation matrix. The gravitational acceleration g is subtracted from the acceleration in the vertical direction (skyward, Z-axis) to obtain the dynamic vibration acceleration a_z_vib(t) of the autonomous mobile device in the visually suspected water accumulation area, which is purely caused by the interaction with the ground. a_z_vib(t)= a_global-g.

[0039] By applying gravity compensation, the sample entropy can more accurately reflect the physical state of the soil (such as changes in viscoelasticity), thereby improving the accuracy of identifying lawn waterlogging types. For example, the sample entropy of groundwater accumulation areas is significantly higher (0.8-1.2), while the sample entropy of healthy lawns is lower (0.3-0.5), and this difference becomes more pronounced after gravity compensation.

[0040] The frequency domain features, time domain nonlinear features, and time domain statistical features of the dynamic vibration acceleration data are extracted as the current body vibration data. The frequency domain features include the dominant frequency component and the frequency band energy ratio; the time domain nonlinear features include sample entropy; and the time domain statistical features include at least one of the root mean square value, peak factor, and waveform factor.

[0041] Specifically, when extracting frequency domain features, power spectral density analysis can be performed, that is, the dynamic vibration acceleration a_z_vib(t) of the autonomous mobile device in the visually suspected water accumulation area, which is purely caused by ground interaction, is windowed (Hanning window) Fast Fourier Transform (FFT) to obtain the power spectral density (PSD). Perform main frequency component extraction, that is, extract the frequency component F_curr with the highest energy in the power spectral density range of 1-20Hz; The band energy ratio is calculated by calculating the ratio of the energy in the low-frequency band (1-5Hz) to the energy in the full-frequency band (1-50Hz) based on the power spectral density, thus obtaining the band energy ratio R_low.

[0042] Specifically, when extracting temporal nonlinear features, the dynamic vibration acceleration a_z_vib(t) of the autonomous mobile device in a visually suspected waterlogged area, purely caused by ground interaction, can be used to obtain the sample entropy, which serves as a core indicator for measuring the complexity of the vibration signal. The sample entropy expression is as follows: SampEn=-ln[A^m(r) / B^m(r)] In the formula, SampEn represents the sample entropy value, m represents the embedding dimension (2 in this embodiment), r represents the tolerance threshold, r = 0.2 * std(a_z_vib), and std represents the standard deviation. A^m(r) is the number of template pairs that remain similar in a higher dimension (m+1). It measures the persistence of signal regularity. Its ratio to B^m(r) jointly determines the final value of the sample entropy, thereby effectively distinguishing the different vibration patterns generated by autonomous mobile devices when driving on healthy lawns and waterlogged lawns. A low SampEn value indicates a regular and predictable signal, representing a healthy and compact lawn (regular vibration); a high SampEn value indicates a complex, random, and irregular signal, representing a waterlogged area (soil softening causes vibrations to become chaotic and irregular).

[0043] Specifically, when extracting time-domain statistical features, the root mean square value (reflecting vibration intensity), peak factor (identifying the impact component in the signal), and waveform factor (reflecting waveform characteristics) of the dynamic vibration acceleration a_z_vib(t) caused purely by ground interaction in the visually suspected waterlogged area of ​​the autonomous mobile device can be statistically obtained based on these features. The expression for the root mean square (RMS) value of dynamic vibration acceleration is: RMS=sqrt(mean(a_z_vib(t) 2 )); The expression for the peak factor CF of dynamic vibration acceleration is: CF = max|a_z_vib(t)| / RMS; The expression for the waveform factor SF of dynamic vibration acceleration is: SF=RMS / mean| a_z_vib(t)|.

[0044] All extracted features are used to construct a feature vector Feature_Vector=[F_curr,R_low,SampEn,RMS,CR,SF].

[0045] Preferably, the method further includes: establishing and updating a dynamic vibration characteristic baseline, including: In the initial phase, vibration data of the machine body was collected in each grid of the lawn area; Based on the initially acquired body vibration data, at least one vibration characteristic baseline is calculated and stored for each grid, and the vibration characteristic baseline includes at least one or more of the following: dominant frequency baseline, sample entropy baseline, and energy ratio baseline. In subsequent operations, the vibration characteristic baseline is updated based on newly collected vibration data of the organism that is identified as being in a healthy state.

[0046] Furthermore, the establishment and updating of the dynamic vibration characteristic baseline also includes: When the lawn area is identified as being in an unhealthy state, the updating of the vibration characteristic baseline of the lawn area is paused. In subsequent operations, when the lawn area is identified as having recovered to a healthy state, the vibration characteristic baseline of the lawn area is updated.

[0047] Preferably, updating the vibration characteristic baseline includes updating the vibration characteristic baseline using an exponentially weighted moving average algorithm.

[0048] S140. Based on the current machine vibration data, obtain the water accumulation type at the current location through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0049] In one embodiment, the preset classification model is trained by: collecting multiple sets of body vibration sample data and extracting the frequency domain features, time domain nonlinear features and time domain statistical features of the sample data; labeling the lawn water accumulation type corresponding to each set of sample data; and using machine learning algorithms to train the feature vectors and labels to obtain the preset classification model.

[0050] Machine learning algorithms include Support Vector Machine (SVM) or Random Forest models. Before learning, the sample data needs to be normalized and cross-validated.

[0051] Following the initial deployment, there will be a baseline learning phase (e.g., the first three full lawn mowings). During this period, all areas are assumed to be "healthy" by default, and IMU data is collected extensively. The lawn is divided into a fine grid (e.g., 10cm x 10cm), and a vibration characteristic baseline is established for each grid (or grid cluster): F_baseline_local: Baseline of dominant vibration frequency in the local area.

[0052] SampEn_baseline_local: Local region sample entropy baseline.

[0053] R_baseline_local: Local area low-frequency energy ratio baseline.

[0054] Wherein, F_baseline_local (local vibration dominant frequency baseline): Physical meaning: Represents the dominant frequency of the vertical vibration signal generated when an autonomous mobile device drives through a region with healthy, compacted soil. Unit: Hertz (Hz) Performance under healthy conditions: On firm, elastic, healthy soil, this value is higher (e.g., 12-15Hz) due to the high-frequency impact of wheels on grass blades and minor uneven surfaces.

[0055] Changes under abnormal conditions: When underground water accumulation causes soil softening (viscoelasticity), the vibration changes from high-frequency impact to low-frequency damped oscillation, resulting in the real-time measured F_curr being significantly lower than F_baseline_local.

[0056] Simply put, F_baseline_local is a healthy benchmark for measuring the "soil hardness" of the area.

[0057] SampEn_baseline_local (Local Sample Entropy Baseline): Physical meaning: Represents a typical value of the complexity and irregularity of vertical vibration signals in this region under healthy conditions. Unit: Dimensionless Performance under healthy conditions: On a healthy lawn, the vibration signal consists of relatively regular periodic impacts. The signal is relatively regular and predictable, so the sample entropy value is low (e.g., 0.3-0.5).

[0058] Changes under abnormal conditions: When underground water accumulates, the damping effect of the soil causes the vibration to become chaotic and irregular random oscillations, increasing the complexity of the signal and resulting in the real-time measured SampEn_curr being significantly higher than SampEn_baseline_local.

[0059] Simply put, SampEn_baseline_local is a health benchmark for measuring the "randomness of vibration signals" in a given area.

[0060] R_baseline_local (Local low-frequency energy ratio to baseline): Physical meaning: It indicates the proportion of vibrational energy in the low-frequency range (such as 1-5Hz) in a healthy state in this region.

[0061] Unit: Dimensionless (ratio, range 0-1) Performance under healthy conditions: High-frequency impact energy dominates in healthy soil, while low-frequency energy accounts for a smaller proportion, hence this value is low.

[0062] Changes under abnormal conditions: Groundwater absorbs high-frequency energy and excites low-frequency oscillations, resulting in a significant increase in the proportion of low-frequency energy, causing the real-time measured R_low to be higher than R_baseline_local.

[0063] Simply put, R_baseline_local is a health benchmark for measuring the "spectral distribution of vibration energy" in a given area.

[0064] Baseline update mechanism: In subsequent operations, if a region is judged as "healthy" by the model as a whole, its current feature values ​​will be used to smoothly update its individual baseline (e.g., using an exponentially weighted moving average). If a region is marked as abnormal (e.g., "waterlogged groundwater"), baseline updates for that region will be paused until it recovers to a healthy state.

[0065] In one embodiment, the preset evaluation rules include: Calculate the deviation of the dominant frequency component, sample entropy, and frequency band energy ratio from the corresponding dominant frequency baseline, sample entropy baseline, and energy ratio baseline, respectively; if the deviation of the dominant frequency component is greater than a third preset threshold, and the deviation of the sample entropy is greater than a fourth preset threshold, or / and the deviation of the frequency band energy ratio is greater than a fifth preset threshold, then the current grid is groundwater accumulation; otherwise, the current grid is surface water accumulation.

[0066] Wherein, the frequency deviation ΔF_ratio = (F_baseline_local - F_curr) / F_baseline_local; Sample entropy deviation ΔSampEn_ratio = (SampEn_curr - SampEn_baseline_local) / SampEn_baseline_local; Energy ratio deviation ΔR_ratio = (R_low - R_baseline_local) / R_baseline_local.

[0067] If the dominant frequency deviation is greater than a third preset threshold, the sample entropy deviation is greater than a fourth preset threshold, and the energy ratio deviation is greater than a fifth preset threshold, then the current grid is groundwater accumulation; otherwise, the current grid is surface water accumulation. Preferably, the third preset threshold is used to determine whether the dominant frequency deviation exceeds the normal range, the fourth preset threshold is used to determine whether the sample entropy deviation exceeds the normal range, and the fifth preset threshold is used to determine whether the energy ratio deviation exceeds the normal range. In practice, these thresholds can also be obtained based on actual experiments.

[0068] This invention provides a method for identifying lawn water accumulation. By fusing visual perception data from an autonomous mobile device with vibration signals recorded by an IMU (Inertial Measurement Unit), a dual-modal decision logic is constructed to achieve high-precision differentiation of lawn water accumulation types. First, preliminary screening is performed based on captured images to identify "visually suspected water accumulation areas" that are dark in color and have smooth surfaces. Then, the frequency domain characteristics of the vibration signals when the autonomous mobile device passes through these "visually suspected water accumulation areas" are analyzed to determine the physical load-bearing characteristics of the soil. Healthy, compact soil exhibits a higher dominant vibration frequency; while groundwater accumulation softens the soil, making it viscoelastic, which causes the autonomous mobile device to produce characteristic vibrations with a lower dominant frequency and increased low-frequency energy. Ultimately, this method achieves the identification of "surface water accumulation" and "groundwater accumulation."

[0069] The lawn water accumulation identification method provided by this invention can bring the following beneficial effects: It enables effective and automated differentiation between surface water and groundwater, solving the problem of high misjudgment rate in traditional visual methods.

[0070] Through physical interaction sensing, it has the ability to detect "underground water" that cannot be seen by the naked eye or ordinary cameras, thus achieving true early diagnosis.

[0071] The method can be directly integrated into existing smart autonomous mobile devices, allowing daily tasks to be completed seamlessly without the need for additional expensive equipment or manpower.

[0072] The dynamic vibration characteristic baseline management mechanism enables the system to adapt to the physical characteristics of different lawns, improving the robustness and universality of the method.

[0073] The lawn water accumulation identification device provided by the present invention is described below. The lawn water accumulation identification system described below can be referred to in correspondence with the lawn water accumulation identification method described above.

[0074] See Figure 2 The present invention provides a lawn water accumulation identification device, which may include: The data acquisition module is used to acquire images of the lawn area, real-time vibration data, and the current location. The visual judgment module is used to determine whether the lawn area is a visually suspected water accumulation area based on the captured image; The data processing module is used to obtain the current body vibration data based on the real-time vibration data if the lawn area is a visually suspected water accumulation area. The vibration recognition module is used to obtain the type of water accumulation at the current location based on the current machine vibration data through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0075] Furthermore, the visual judgment module is specifically used to divide the lawn area into multiple grids; Extract the color and texture features corresponding to each grid; Based on the color and texture features, determine whether the corresponding grid is a visually suspected water accumulation area.

[0076] Specifically, based on color and texture features, the average brightness and texture variance of each grid are calculated; If the average brightness is lower than a first preset threshold and the texture variance is lower than a second preset threshold, then the corresponding grid is determined to be a visually suspected water accumulation area.

[0077] Furthermore, the data processing module is specifically used to perform coordinate transformation and gravity compensation on the real-time vibration data to obtain dynamic vibration acceleration data; The frequency domain features, time domain nonlinear features, and time domain statistical features of the dynamic vibration acceleration data are extracted and used as the current body vibration data.

[0078] Furthermore, the frequency domain features include the dominant frequency component and the band energy ratio, the time domain nonlinear features include the sample entropy, and the time domain statistical features include at least one of the root mean square value, peak factor, and waveform factor.

[0079] Furthermore, the device also includes: a baseline management module for establishing and updating dynamic vibration characteristic baselines. Specifically, the baseline management module is used in the initial stage to collect body vibration data in each grid of the lawn area; based on the initially collected body vibration data, to calculate and store at least one vibration characteristic baseline for each grid, the vibration characteristic baseline including at least one or more of the dominant frequency baseline, sample entropy baseline, and energy ratio baseline; in subsequent operations, the vibration characteristic baseline is updated according to newly collected body vibration data identified as being in a healthy state.

[0080] Furthermore, the baseline management module is specifically used to pause updating the vibration characteristic baseline of the lawn area when the lawn area is identified as being in an unhealthy state; and to resume updating the vibration characteristic baseline of the lawn area in subsequent operations when the lawn area is identified as being in a healthy state.

[0081] Furthermore, the baseline management module is specifically used to update the vibration characteristic baseline using an exponentially weighted moving average algorithm.

[0082] Furthermore, the vibration identification module is specifically used to calculate the deviation of the dominant frequency component, sample entropy, and frequency band energy ratio from the corresponding dominant frequency baseline, sample entropy baseline, and energy ratio baseline, respectively; if the deviation of the dominant frequency component is greater than a third preset threshold, and the deviation of the sample entropy is greater than a fourth preset threshold, or / and the deviation of the frequency band energy ratio is greater than a fifth preset threshold, then the current grid is groundwater accumulation; otherwise, the current grid is surface water accumulation.

[0083] Furthermore, the preset classification model is trained in the following way: multiple sets of body vibration sample data are collected, and the frequency domain features, time domain nonlinear features and time domain statistical features of the sample data are extracted; the lawn water accumulation type corresponding to each set of sample data is labeled; and the feature vector and label are trained using machine learning algorithms to obtain the preset classification model.

[0084] Specific limitations regarding the lawn waterlogging detection device can be found in the limitations of the lawn waterlogging detection method described above, and will not be repeated here. Each module in the aforementioned lawn waterlogging detection device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of the electronic device in hardware form or independently of it, or stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of each module.

[0085] This invention also provides an autonomous mobile device, including a visual perception module, a physical perception module, a positioning module, a computing and control unit, and a communication module. The visual perception module acquires images of the autonomous mobile device in a lawn area; the physical perception module acquires vibration data of the autonomous mobile device in the lawn area; the positioning module acquires geographical location data of the autonomous mobile device in the lawn area; and the computing and control unit receives the images, vibration data, and geographical location data acquired by the visual perception module, physical perception module, and positioning module, executes any of the lawn water accumulation identification methods described above, obtains lawn water accumulation identification results (lawn water accumulation type, location, and confidence level), adjusts the autonomous mobile device's operating path based on the lawn water accumulation identification results, binds the lawn water accumulation identification results to the geographical location, updates the internal map of the autonomous mobile device, and sends it to a user terminal or the cloud via the communication module. The computing and control unit can also perform corresponding operations based on the lawn water accumulation identification results, such as marking "groundwater accumulation" areas as restricted areas to avoid further damage from trampling in the short term, and recording "surface water accumulation" areas for natural drying.

[0086] Example 1: Integration of Conventional Intelligent Autonomous Mobile Devices Scenario: Routine lawn mowing and health monitoring in a family backyard.

[0087] Process: The autonomous mobile device operates along a predetermined path. During its journey, the camera continuously records data. When a dark-colored area is detected, the system marks it as a visually suspected area of ​​water accumulation. As the autonomous mobile device passes through this area, the IMU records vibration data. Analysis reveals that the dominant vibration frequency is 8Hz, while the current healthy baseline is 12Hz, significantly lower than the threshold. The system immediately determines the area to be "groundwater accumulation" and marks it with a blue polygon on the app map, while also advising the user that "the soil in this area is waterlogged; it is recommended to suspend watering and consider aeration."

[0088] Example 2: Enhanced Diagnostics and Path Planning Scenario: Automated maintenance of a park lawn.

[0089] Process: The autonomous mobile device identifies a strip-shaped area of ​​"groundwater accumulation." The control unit not only marks it but also immediately and dynamically adjusts the subsequent mowing path to avoid the area, preventing damage to the turf from the autonomous mobile device's operation and subsequent damage to the device from overflowing groundwater. Simultaneously, the anomaly is reported, and the server can notify the lawn maintenance team or dispatch a smart aerator for targeted operations.

[0090] The present invention also provides a lawn water accumulation identification system, including the autonomous mobile device described above, as well as a cloud server, a user terminal and / or an autonomous water accumulation treatment device. The cloud server or user terminal is used to receive and display the water accumulation identification results; the autonomous water accumulation treatment device is used to perform autonomous water accumulation treatment operations based on the received water accumulation identification results and the location of the water accumulation.

[0091] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3 As shown, the electronic device may include a processor 810, a communications interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute the following steps: Acquire images of the lawn area, real-time vibration data, and current location; Based on the captured images, determine whether the lawn area is visually suspected to be a waterlogged area; If the lawn area is visually suspected to be a waterlogged area, then the current body vibration data is obtained based on the real-time vibration data; Based on the current machine vibration data, the type of water accumulation at the current location is obtained through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0092] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0093] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being able to be stored on a non-transitory computer-readable storage medium, and the computer program being executed by a processor, enabling the computer to perform the following steps: Acquire images of the lawn area, real-time vibration data, and current location; Based on the captured images, determine whether the lawn area is visually suspected to be a waterlogged area; If the lawn area is visually suspected to be a waterlogged area, then the current body vibration data is obtained based on the real-time vibration data; Based on the current machine vibration data, the type of water accumulation at the current location is obtained through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0094] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps: Acquire images of the lawn area, real-time vibration data, and current location; Based on the captured images, determine whether the lawn area is visually suspected to be a waterlogged area; If the lawn area is visually suspected to be a waterlogged area, then the current body vibration data is obtained based on the real-time vibration data; Based on the current machine vibration data, the type of water accumulation at the current location is obtained through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

[0095] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; 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; and these 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 method for identifying water accumulation in lawns, characterized in that, The method includes: Acquire images of the lawn area, real-time vibration data, and current location; Based on the captured images, determine whether the lawn area is visually suspected to be a waterlogged area; If the lawn area is visually suspected to be a waterlogged area, then the current body vibration data is obtained based on the real-time vibration data; Based on the current machine vibration data, the type of water accumulation at the current location is obtained through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

2. The lawn water accumulation identification method according to claim 1, characterized in that, The step of determining whether the lawn area is a visually suspected waterlogged area based on the captured image includes: The lawn area is divided into multiple grids; Extract the color and texture features corresponding to each grid; Based on the color and texture features, determine whether the corresponding grid is a visually suspected water accumulation area.

3. The lawn water accumulation identification method according to claim 2, characterized in that, The step of determining whether the corresponding grid is a visually suspected water accumulation area based on the color and texture features includes: Calculate the average brightness and texture variance of each grid based on color and texture features; If the average brightness is lower than a first preset threshold and the texture variance is lower than a second preset threshold, then the corresponding grid is determined to be a visually suspected water accumulation area.

4. The lawn water accumulation identification method according to claim 1, characterized in that, The step of obtaining current body vibration data based on the real-time vibration data includes: The real-time vibration data is subjected to coordinate transformation and gravity compensation to obtain dynamic vibration acceleration data; The frequency domain features, time domain nonlinear features, and time domain statistical features of the dynamic vibration acceleration data are extracted and used as the current body vibration data.

5. The lawn water accumulation identification method according to claim 4, characterized in that, The frequency domain characteristics include the dominant frequency component and the bandwidth energy ratio; The time-domain nonlinear characteristics include sample entropy; The time-domain statistical features include at least one of the root mean square value, peak factor, and waveform factor.

6. The lawn water accumulation identification method according to claim 5, characterized in that, The method further includes establishing and updating dynamic vibration characteristic baselines, including: in an initial stage, collecting body vibration data in each grid of the lawn area; based on the initially collected body vibration data, calculating and storing at least one vibration characteristic baseline for each grid, the vibration characteristic baseline including at least one or more of the dominant frequency baseline, sample entropy baseline, and energy ratio baseline; in subsequent operations, updating the vibration characteristic baselines based on newly collected body vibration data identified as being in a healthy state.

7. The lawn water accumulation identification method according to claim 6, characterized in that, The establishment and updating of the dynamic vibration characteristic baseline also includes: when the lawn area is identified as being in an unhealthy state, pausing the updating of the vibration characteristic baseline of the lawn area; in subsequent operations, when the lawn area is identified as having recovered to a healthy state, resuming the updating of the vibration characteristic baseline of the lawn area.

8. The lawn water accumulation identification method according to claim 6, characterized in that, The step of updating the vibration characteristic baseline includes updating the vibration characteristic baseline using an exponentially weighted moving average algorithm.

9. The lawn water accumulation identification method according to claim 6, characterized in that, The step of determining the type of water accumulation at the current location based on the current machine vibration data and through preset evaluation rules includes: Calculate the deviation of the dominant frequency component, sample entropy, and frequency band energy ratio from the corresponding dominant frequency baseline, sample entropy baseline, and energy ratio baseline, respectively; if the deviation of the dominant frequency component is greater than a third preset threshold, and the deviation of the sample entropy is greater than a fourth preset threshold, or / and the deviation of the frequency band energy ratio is greater than a fifth preset threshold, then the current grid is groundwater accumulation; otherwise, the current grid is surface water accumulation.

10. The method for identifying lawn water accumulation according to any one of claims 1 to 9, characterized in that, The preset classification model is trained in the following way: multiple sets of body vibration sample data are collected, and the frequency domain features, time domain nonlinear features and time domain statistical features of the sample data are extracted; the lawn water accumulation type corresponding to each set of sample data is labeled; and the feature vector and label are trained using machine learning algorithms to obtain the preset classification model.

11. A lawn water accumulation detection device, characterized in that, The device includes: The data acquisition module is used to acquire images of the lawn area, real-time vibration data, and the current location. The visual judgment module is used to determine whether the lawn area is a visually suspected water accumulation area based on the captured image; The data processing module is used to obtain the current body vibration data based on the real-time vibration data if the lawn area is a visually suspected water accumulation area. The vibration recognition module is used to obtain the type of water accumulation at the current location based on the current machine vibration data through a preset classification model or preset evaluation rules. The water accumulation type includes surface water accumulation on the lawn and groundwater accumulation on the lawn.

12. An autonomous mobile device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the lawn water accumulation identification method as described in any one of claims 1 to 10.

13. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the lawn water accumulation identification method as described in any one of claims 1 to 10.

14. A lawn water accumulation identification system, characterized in that, Includes the autonomous mobile device as described in claim 12, as well as a cloud server, a user terminal, and / or an autonomous water accumulation treatment device, wherein the cloud server or user terminal is used to receive and display water accumulation identification results; and the autonomous water accumulation treatment device is used to perform autonomous water accumulation treatment operations based on the received water accumulation identification results and the location of the water accumulation.

Images