Water level prediction model construction method based on image recognition

Abstract

The invention discloses a water level prediction model construction method based on image recognition, and relates to the technical field of water level monitoring. Comprising the following steps: establishing a visual reference system of a region, constructing a multispectral image data fusion module, constructing a water surface edge detection module, designing an anomaly detection mechanism based on self-supervised learning, constructing a water level height estimation model based on image semantic segmentation, and constructing a water level distribution prediction model combined with geographic information. A data feedback model based on meteorological dynamics is constructed, and a water level trend estimation module based on short-term time sequence prediction is introduced. According to the method, a water level prediction model is constructed by introducing technologies of multispectral image fusion, semantic segmentation and dynamic height estimation, and accurate extraction and prediction of water level information are realized under various environmental conditions. The robustness of the model under complex weather conditions is remarkably improved, and meanwhile, the problem of deviation caused by different visual angles and optical distortion is solved through dynamic adjustment factors and a geometric correction mechanism.

Description

Technical Field

[0001] The present invention relates to the technical field of water level monitoring, and in particular to a method for constructing a water level prediction model based on image recognition. Background Art

[0002] Currently, water level monitoring systems usually rely on IoT sensors, radar or ultrasonic ranging devices. However, these technologies face many challenges in practical applications: hardware equipment is easily affected by the external environment (such as heavy rainfall, floating objects), and the maintenance cost is high when deployed over a large area. In recent years, water level monitoring technology based on image recognition has gradually emerged, which uses cameras to collect images and combines algorithms to achieve water level monitoring and prediction. Compared with traditional equipment, this method has the advantages of non-contact, high flexibility and low cost.

[0003] In the existing technology, the ability of traditional single-spectrum imaging methods to distinguish between water surface and landmarks is significantly reduced under conditions of insufficient light or strong reflection (such as at night, heavy rain, and haze). Many methods rely on traditional edge detection or simple segmentation algorithms, which are difficult to cope with the interference caused by complex backgrounds (such as floating objects and reflective water surfaces). Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a method for constructing a water level prediction model based on image recognition to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for constructing a water level prediction model based on image recognition, comprising the following steps:

[0006] S1. Establish a regional visual reference system in the target waters;

[0007] S2. In the visual benchmark system, a multispectral image data fusion module is constructed to fuse the image data collected from the target water area;

[0008] S3. In the visual reference system, a water surface edge detection module is constructed to locate the water surface edge of the target water area;

[0009] S4. Design an anomaly detection mechanism based on self-supervised learning to perform image anomaly detection on the image data collected in the target water area;

[0010] S5. Construct a water level estimation model based on image semantic segmentation to separate the water surface, landmarks and other background areas of the target water area to estimate the water level;

[0011] S6. Integrate the geographic information of the surrounding areas of the target waters and construct a water level distribution prediction model combining the geographic information;

[0012] S7. Construct a data feedback model based on meteorological dynamics and access meteorological data to calculate water level prediction;

[0013] S8. Introduce the water level trend estimation module of short-term time series prediction to complete the construction of the water level prediction model.

[0014] To further optimize the technical solution, in step S1, a visual reference system is designed by analyzing the topography of the target water area and known water level markers, including bridge piers and water level columns;

[0015] The benchmark system determines the specific features of water level markers in the image, including highly segmented color changes, shape markings, and is established in three-dimensional space by combining multi-angle image data.

[0016] To further optimize the technical solution, in step S2, the multispectral image data fusion module has a built-in adaptive weight allocation algorithm, which dynamically adjusts the credibility of different spectral data according to the acquisition environment. The spectral data includes visible light, infrared and ultraviolet data, and the fused image can always reflect the real changes in the water level.

[0017] To further optimize the technical solution, the adaptive weight distribution algorithm dynamically adjusts the credibility of visible light, infrared and ultraviolet data to ensure that the fused image can accurately reflect the real changes in water level;

[0018] The adaptive weight allocation algorithm includes the following processes:

[0019] Input multispectral image data;

[0020] Calculate the image signal-to-noise ratio;

[0021] Calculate the weights of environmental factors;

[0022] Normalized weights;

[0023] Multispectral image fusion.

[0024] To further optimize the technical solution, in step S3, the water surface edge detection module is based on a deep reinforcement learning algorithm, and the model learns how to detect the water surface edge in a complex background from a large amount of training data, and dynamically adjusts the strategy during detection to improve the robustness of water surface detection and provide accurate basic data for water level prediction;

[0025] When the water surface has large ripples, the model prioritizes smooth areas;

[0026] When light reflections are strong, the model chooses to avoid the bright areas.

[0027] To further optimize the technical solution, in step S4, the anomaly detection mechanism, when used, includes:

[0028] When the water level markers are covered by plants or the camera is blurred by raindrops, the anomaly detection mechanism learns the intrinsic distribution characteristics of the data from normal water level images, actively issues warnings when abnormal inputs are detected, and completes the missing image information through the generative adversarial network (GAN).

[0029] To further optimize the technical solution, in step S5, the water level estimation model includes:

[0030] Define water level :

[0031] ;

[0032] in,

[0033] : The vertical pixel coordinate of the water surface area, the pixel row number from the top of the image to the water surface boundary area;

[0034] : The vertical pixel coordinate of the landmark reference point, the pixel row number from the top of the image to the bottom of the landmark;

[0035] : The mapping coefficient between the pixel of image height and the real height;

[0036] : Dynamic environment adjustment factor, used to correct errors caused by optical distortion or environmental interference;

[0037] Split Mapping:

[0038] The semantic segmentation network divides the image into three regions: water surface, landmarks, and background, and outputs a map;

[0039] Geometry Correction:

[0040] Automatically adapt images of different viewing angles and camera positions to ensure the accuracy of unit conversion;

[0041] Dynamic Environment Adjustment Factor:

[0042] In complex environments including strong light reflection and rain blur, segmentation may have errors, and the factor is adjusted dynamically to correct the water level estimation.

[0043] To further optimize the technical solution, in the segmentation mapping, the following mapping is output:

[0044] : Pixels The classification probability map of represents the probability of water surface, represents the probability of a marker;

[0045] Water surface boundary extraction:

[0046] ;

[0047] Among them, the threshold Used to determine whether a pixel belongs to the water surface;

[0048] Marker Benchmark Extraction:

[0049] ;

[0050] Segmentation mapping is used to ensure that the model can accurately extract the boundary area between the water surface and the landmark from the complex background.

[0051] To further optimize the technical solution, in step S7, the data feedback model is used to introduce the factors of precipitation, wind speed and temperature into the calculation of water level prediction;

[0052] The model inputs meteorological data in real time, affects the final water level prediction results in a weighted manner, and automatically optimizes the degree of dependence on meteorological data based on environmental changes or seasonal precipitation patterns.

[0053] To further optimize the technical solution, in step S8, the water level trend estimation module uses the water level data output by the image recognition as input, combines it with the historical water level data, and predicts the future water level change trend;

[0054] During training, the model combines sliding window technology to extract the short-term dynamic characteristics of the water level and provide accurate trend forecast results, thereby predicting the water level trend in the next 1-3 hours.

[0055] Compared with the prior art, the present invention provides a method for constructing a water level prediction model based on image recognition, which has the following beneficial effects:

[0056] This method of constructing a water level prediction model based on image recognition introduces multispectral image fusion, semantic segmentation, and dynamic height estimation technologies to construct a water level prediction model, and achieves accurate extraction and prediction of water level information under a variety of environmental conditions. Compared with existing technologies, the robustness of the model under complex weather conditions is significantly improved, and the deviation problem caused by different viewing angles and optical distortion is solved through dynamic adjustment factors and geometric correction mechanisms. The model has higher applicability, accuracy, and environmental adaptability, and is suitable for intelligent water resources management, flood warning, and other fields, effectively reducing hardware maintenance costs and improving monitoring efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 A schematic diagram of a flow chart of a method for constructing a water level prediction model based on image recognition proposed by the present invention;

[0058] Figure 2 A schematic flow chart of an adaptive weight allocation algorithm in a water level prediction model construction method based on image recognition proposed by the present invention;

[0059] Figure 3 This is a flow chart of a water level height estimation model in a water level prediction model construction method based on image recognition proposed in the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Embodiment 1:

[0062] See also Figure 1 , a method for constructing a water level prediction model based on image recognition, comprising the following steps:

[0063] S1. Establish a regional visual reference system

[0064] In this embodiment, a visual reference system is designed by analyzing the topography of the target water area and known water level markers, including bridge piers and water level columns;

[0065] The benchmark system determines the specific features of water level markers in the image, including highly segmented color changes, shape markings, and is established in three-dimensional space by combining multi-angle image data.

[0066] Water level prediction requires accurate water level reference markers, and the markers may be different for each target area. By establishing a visual benchmark system, it can be ensured that the model can identify the basis for water level changes in the area. For example, the red, yellow, and green segmented markings on the piers can be marked as key features, and possible visual interference such as algae and dirt need to be eliminated. The process of building the benchmark system includes taking panoramic and close-up images of the water area, and applying deep learning technology to feature and classify the water level markers. This system will serve as a reference framework for the entire model to ensure the regional specificity and accuracy of subsequent predictions.

[0067] S2. Construct multispectral image data fusion module

[0068] Since light conditions (such as at night, foggy days, or heavy rain) can cause the quality of visible light images to degrade, multispectral data acquisition technology is introduced. Infrared images can detect heat signals in low-light or no-light conditions, while ultraviolet light is sensitive to reflections from organic matter or floating objects in the water. By fusing these spectral data with visible light images, it is possible to ensure stable water level information in various climate conditions.

[0069] In this embodiment, the multispectral image data fusion module has a built-in adaptive weight distribution algorithm, which dynamically adjusts the credibility of different spectral data according to the acquisition environment. The spectral data includes visible light, infrared and ultraviolet data. The fused image can always reflect the real changes in the water level.

[0070] The adaptive weight distribution algorithm dynamically adjusts the credibility of visible light, infrared and ultraviolet data to ensure that the fused image can accurately reflect the real changes in water level.

[0071] S3. Build water surface edge detection module

[0072] In this embodiment, the location of the edge of the water surface is a key step in predicting the water level. However, due to interference from reflections, ripples, and floating objects, traditional edge detection algorithms (such as the Canny algorithm) may not be reliable enough. The water surface edge detection module is based on a deep reinforcement learning algorithm. The model learns how to detect the edge of the water surface in a complex background from a large amount of training data, and dynamically adjusts the strategy during detection to improve the robustness of water surface detection and provide accurate basic data for water level prediction;

[0073] When the water surface is rippled, the model prioritizes smooth areas.

[0074] If the standard deviation of the grayscale or texture changes in the water surface area in the image exceeds a preset threshold (for example, the standard deviation is greater than 50), the water surface ripples are considered to be large. At the same time, in the frequency domain, if the amplitude of the ripples exceeds the set threshold (for example, the frequency distribution of the ripples is higher than a certain standard value, such as 5Hz), it can also be considered to be large ripples.

[0075] When light reflections are strong, the model chooses to avoid the bright areas.

[0076] In an RGB image, if the pixel brightness (e.g., grayscale value or the maximum value of the three channels R, G, and B) is greater than a preset threshold (e.g., 200), the area can be considered to have strong reflection. If the brightness value of a certain area exceeds the brightness of 99% of the pixels in the image (e.g., exceeds the 95% percentile), the light reflection can also be considered to be strong.

[0077] S4. Design anomaly detection mechanism based on self-supervised learning

[0078] During actual operation, the model may be affected by equipment obstruction, contamination, and other issues, resulting in abnormal data input.

[0079] In this embodiment, the anomaly detection mechanism, when used, includes:

[0080] When the water level markers are covered by plants or the camera is blurred by raindrops, the anomaly detection mechanism learns the intrinsic distribution characteristics of the data from normal water level images, actively issues warnings when abnormal inputs are detected, and completes the missing image information through the generative adversarial network (GAN).

[0081] S5. Construct a water level estimation model based on image semantic segmentation

[0082] Calculating the water level directly in the image requires accurate segmentation of the water surface area. Perform pixel-level segmentation on the input image to separate the water surface, landmarks, and other background areas. The segmentation results can be directly used for quantitative calculation of the water level, such as calculating the actual water level from the height mark on the landmark. The model needs to be trained on a variety of data under different lighting and weather conditions to ensure the stability and universality of the segmentation results.

[0083] S6. Construct a water level distribution prediction model combining geographic information

[0084] In this embodiment, water level changes are not only affected by the local environment, but also closely related to the topography of the surrounding area. For example, precipitation in a certain upstream area may cause the water level to rise downstream. Therefore, by integrating GIS data, including river networks, terrain height differences, historical precipitation records, etc., a water level distribution prediction model is established to predict the water level based on the topography information.

[0085] S7. Construct a data feedback model based on meteorological dynamics

[0086] In this embodiment, water level prediction not only depends on image data, but is also closely related to meteorological factors. For example, heavy rain will significantly increase the water level, while continued high temperatures may cause the water level to drop. The data feedback model is used to introduce factors such as precipitation, wind speed, and temperature into the calculation of water level prediction;

[0087] The model inputs meteorological data in real time, affects the final water level prediction results in a weighted manner, and automatically optimizes the degree of dependence on meteorological data based on environmental changes or seasonal precipitation patterns.

[0088] S8, introduce the water level trend estimation module of short-term time series prediction

[0089] In this embodiment, the water level trend estimation module takes the water level data output by image recognition as input, combines it with the historical water level data, and predicts the future water level change trend, thereby completing the construction process of the water level prediction model.

[0090] The water level prediction model combines the sliding window technology in training to extract the short-term dynamic characteristics of the water level and provide accurate trend prediction results, thereby predicting the water level trend in the next 1-3 hours.

[0091] Embodiment 2:

[0092] See also Figure 2 and Figure 3 Based on the water level prediction model construction method based on image recognition described in Example 1, in step S2, an adaptive weight allocation algorithm is designed.

[0093] Step 1: Input multispectral image data

[0094] set up:

[0095] : Visible light image;

[0096] : Infrared light image;

[0097] : Ultraviolet image.

[0098] Collect environmental information for , including the following environmental features:

[0099] Light intensity ;

[0100] Weather conditions (such as sunny, rainy, foggy);

[0101] Time period (Day or night).

[0102] Step 2: Calculate the image signal-to-noise ratio

[0103] For each spectral image, calculate the signal-to-noise ratio , to quantify the credibility of each image:

[0104]

[0105] Effective pixel intensity: refers to the pixel intensity of the water surface or water level marker area in the image (achieved through preliminary segmentation).

[0106] Noise intensity: refers to the standard deviation of pixel intensity in the background area of ​​the image.

[0107] Get .

[0108] They are the image signal-to-noise ratio of visible light, the image signal-to-noise ratio of infrared light, and the image signal-to-noise ratio of ultraviolet light.

[0109] Step 3: Calculate environmental factor weights

[0110] According to environmental characteristics , determine the initial weight of the spectral type :

[0111] Visible light weight :

[0112]

[0113] when Higher (daytime, strong light environment), visible light is more important, and the weight tends to 1.

[0114] when The lower the value (night, low light environment), the lower the weight.

[0115] Infrared light weight :

[0116]

[0117] In low light, rainy days or at night, the infrared weighting is increased.

[0118] and is a gain function based on weather and time of day.

[0119] UV Weight :

[0120]

[0121] On foggy days, the UV weight decreases; on sunny days, the weight increases.

[0122] is the gain function according to the visibility (haze) in the environment.

[0123] in, is a tuning parameter.

[0124] Step 4: Normalize weights

[0125] The signal-to-noise ratio and the environmental weighting are combined.

[0126] Step 5: Multispectral Image Fusion

[0127] According to the calculated weights, the three spectral images are weighted and fused into the final image. :

[0128]

[0129] The model is based on environmental characteristics and signal-to-noise ratio, and the weights are adjusted in real time to adapt to different light, weather and time conditions. It has strong adaptability to complex conditions such as low light, fog and rain. It combines the signal-to-noise ratio and environmental factor weights to avoid direct reliance on complex deep learning calculations, greatly improving real-time performance.

[0130] When applied, the model:

[0131] On a clear day: It has a heavy weight and mainly relies on visible light images.

[0132] At night: The weight is greater, and the infrared image contributes more.

[0133] In foggy weather: Increase the weight of the UV image to capture the reflection characteristics of the water surface.

[0134] The water level estimation model in step S5 is designed.

[0135] The water level estimation model includes the following processes:

[0136] Define water level

[0137] ;

[0138] in,

[0139] : The vertical pixel coordinate of the water surface area, the pixel row number from the top of the image to the water surface boundary area;

[0140] : The vertical pixel coordinate of the landmark reference point, the pixel row number from the top of the image to the bottom of the landmark;

[0141] : The mapping coefficient between the pixel of image height and the real height;

[0142] : Dynamic environment adjustment factor, used to correct errors caused by optical distortion or environmental interference.

[0143] Segmentation Map

[0144] The semantic segmentation network divides the image into three regions: water surface, landmark, and background, and outputs the following mapping:

[0145] : Pixels The classification probability map of represents the probability of water surface, represents the probability of a marker;

[0146] Water surface boundary extraction:

[0147] ;

[0148] Among them, the threshold Used to determine whether a pixel belongs to the water surface;

[0149] Marker Benchmark Extraction:

[0150] ;

[0151] Segmentation mapping is used to ensure that the model can accurately extract the boundary area between the water surface and the landmark from the complex background.

[0152] Geometry Correction

[0153] It is the conversion coefficient between image pixel height and real height, which can be dynamically calculated by the following formula:

[0154]

[0155] in,

[0156] : The true height of the marker (known).

[0157] : The vertical pixel coordinate of the top of the marker;

[0158] : The vertical pixel coordinate of the landmark reference point.

[0159] Automatically adapt images to different viewing angles and camera positions to ensure the accuracy of unit conversion.

[0160] Dynamic Environment Adjustment Factor

[0161] In complex environments including strong light reflection and rain blur, segmentation has errors and the dynamic adjustment factor to correct the water level estimate.

[0162] When used, the model includes:

[0163] Step 1: Semantic segmentation to extract water surface and landmark areas

[0164] Perform pixel-level classification on the input image and generate a classification probability map . Extract water surface boundary points by threshold method and landmark reference points .

[0165] Step 2: Determine the pixel and true height mapping of the landmark

[0166] Based on the known true height of the landmark , calculate the conversion factor , used to deduce the true height from the pixel height.

[0167] Step 3: Calculate the base water level

[0168] Divide the difference between the water surface pixel position and the marker reference pixel position by , get the basic water level height.

[0169] Step 4: Dynamically adjust the water level

[0170] Calculate the adjustment factor based on image brightness (ambient light interference) and edge uncertainty , and correct the base water level to get the final result.

[0171] The model uses geometric correction and dynamic adjustment factors to adapt to a variety of acquisition environments, including different viewing angles, lighting conditions, and weather conditions. With the help of a semantic segmentation network, the boundary area between the water surface and the landmark can be accurately extracted, avoiding the errors of traditional edge detection methods.

[0172] When applied, the model:

[0173] The actual height of the marker is 2 meters;

[0174] The pixel positions of the top and bottom of the marker are 100 and 300, respectively;

[0175] The water surface pixel position is 299.95;

[0176] Dynamic Adjustment Factor -0.05 (correction of the water surface's downward offset caused by light reflection).

[0177] calculate :

[0178]

[0179] Calculate the water level:

[0180]

[0181] The final water level was -5.05 meters (the water level was 5.05 meters downward relative to the marker reference point).

[0182] The beneficial effects of the present invention are:

[0183] This method of constructing a water level prediction model based on image recognition introduces multispectral image fusion, semantic segmentation, and dynamic height estimation technologies to construct a water level prediction model, and achieves accurate extraction and prediction of water level information under a variety of environmental conditions. Compared with existing technologies, the robustness of the model under complex weather conditions is significantly improved, and the deviation problem caused by different viewing angles and optical distortion is solved through dynamic adjustment factors and geometric correction mechanisms. The model has higher applicability, accuracy, and environmental adaptability, and is suitable for intelligent water resources management, flood warning, and other fields, effectively reducing hardware maintenance costs and improving monitoring efficiency.

[0184] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples and different implementations described in this specification without contradiction.

[0185] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for constructing a water level prediction model based on image recognition, characterized in that: The following steps are involved: S1. Establish a regional visual reference system in the target waters; S2. In the visual benchmark system, a multispectral image data fusion module is constructed to fuse the image data collected from the target water area; S3. In the visual reference system, a water surface edge detection module is constructed to locate the water surface edge of the target water area; S4. Design an anomaly detection mechanism based on self-supervised learning to perform image anomaly detection on the image data collected from the target water area; S5. Construct a water level estimation model based on image semantic segmentation to separate the water surface, landmarks and other background areas of the target water area to estimate the water level; S6. Integrate the geographic information of the surrounding areas of the target waters and construct a water level distribution prediction model combining the geographic information; S7. Construct a data feedback model based on meteorological dynamics and access meteorological data to calculate water level prediction; S8. Introduce the water level trend estimation module of short-term time series prediction to complete the construction of the water level prediction model.

2. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In the step S1, a visual reference system is designed by analyzing the topography of the target water area and known water level markers, including bridge piers and water level columns; The benchmark system determines the specific features of water level markers in the image, including highly segmented color changes, shape markings, and is established in three-dimensional space by combining multi-angle image data.

3. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In step S2, the multispectral image data fusion module has a built-in adaptive weight distribution algorithm, which dynamically adjusts the credibility of different spectral data according to the acquisition environment. The spectral data includes visible light, infrared and ultraviolet data. The fused image can always reflect the real changes in the water level.

4. The method for constructing a water level prediction model based on image recognition according to claim 3 is characterized in that: The adaptive weight distribution algorithm dynamically adjusts the credibility of visible light, infrared and ultraviolet data to ensure that the fused image can accurately reflect the real changes in water level; The adaptive weight allocation algorithm includes the following processes: Input multispectral image data; Calculate the image signal-to-noise ratio; Calculate the weights of environmental factors; Normalized weights; Multispectral image fusion.

5. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In step S3, the water surface edge detection module is based on a deep reinforcement learning algorithm. The model learns how to detect the water surface edge in a complex background from a large amount of training data, and dynamically adjusts the strategy during detection to improve the robustness of water surface detection and provide accurate basic data for water level prediction; When the water surface has large ripples, the model focuses on the smooth areas; When light reflections are strong, the model chooses to avoid the bright areas.

6. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In step S4, when the anomaly detection mechanism is used, it includes: When the water level markers are covered by plants or the camera is blurred by raindrops, the anomaly detection mechanism learns the intrinsic distribution characteristics of the data from normal water level images, actively issues warnings when abnormal inputs are detected, and completes the missing image information through the generative adversarial network (GAN).

7. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In step S5, the water level estimation model includes: Define water level : ； in, : The vertical pixel coordinate of the water surface area, the pixel row number from the top of the image to the water surface boundary area; : The vertical pixel coordinate of the landmark reference point, the pixel row number from the top of the image to the bottom of the landmark; : The mapping coefficient between the pixel of image height and the real height; : Dynamic environment adjustment factor, used to correct errors caused by optical distortion or environmental interference; Split Mapping: The semantic segmentation network divides the image into three regions: water surface, landmarks, and background, and outputs a map; Geometry Correction: Automatically adapt images of different viewing angles and camera positions to ensure the accuracy of unit conversion; Dynamic Environment Adjustment Factor: In complex environments including strong light reflection and rain blur, segmentation may have errors, and the factor is adjusted dynamically to correct the water level estimation.

8. The method for constructing a water level prediction model based on image recognition according to claim 7, characterized in that: In the segmentation mapping, the following mapping is output: : Pixels The classification probability map of represents the probability of water surface, represents the probability of a marker; Water surface boundary extraction: ； Among them, the threshold Used to determine whether a pixel belongs to the water surface; Marker Benchmark Extraction: ； Segmentation mapping is used to ensure that the model can accurately extract the boundary area between the water surface and the landmark from the complex background.

9. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In step S7, the data feedback model is used to introduce factors such as precipitation, wind speed, and temperature into the calculation of water level prediction; The model accesses meteorological data in real time, affects the final water level prediction results through weighted methods, and automatically optimizes the degree of dependence on meteorological data according to environmental changes or seasonal precipitation patterns.

10. The method for constructing a water level prediction model based on image recognition according to claim 1, characterized in that: In step S8, the water level trend estimation module uses the water level data output by the image recognition as input, combines it with the historical water level data, and predicts the future water level change trend; During training, the model combines sliding window technology to extract the short-term dynamic characteristics of the water level and provide accurate trend forecast results, thereby predicting the water level trend in the next 1-3 hours.

Images