Urban small and micro wetland digital evaluation and parameterization combination construction method and system

Abstract

The application discloses a city small and micro wetland digital evaluation and parameterization combination construction method and system, and the method comprises the following steps: collecting site multi-source remote sensing data by a drone and preprocessing; carrying out digital evaluation on ecological factors of the site to obtain site characteristics; constructing a standard ecological factor coding label set according to a preset evaluation standard, and distributing corresponding codes to the site based on the site characteristics; splitting the ecological function of the ecological factor into different splicable function components to form a component library, each function component has an adaptation rule and a function component code associated with the standard ecological factor code; judging the site scene type and the construction target and determining the adaptation condition; calling the parameterized function component library based on the standard ecological factor code, matching the corresponding function component according to the adaptation condition and the adaptation rule, outputting the function component code combination configuration scheme through parameterized configuration to generate a city small and micro wetland design scheme. The application can improve the design efficiency of the city small and micro wetland design scheme.

Description

Technical Field

[0001] This application relates to the field of urban ecological infrastructure construction technology, and in particular to a method and system for digital assessment and parametric combination construction of urban micro-wetlands. Background Technology

[0002] With the deepening implementation of concepts such as "sponge city" and "ecological city," urban micro-wetlands, as an important component of green infrastructure, are receiving increasing attention for their role in regulating rainwater, purifying water quality, and enhancing biodiversity. However, the construction of urban micro-wetlands currently faces the following challenges:

[0003] 1. Lack of rapid site assessment methods: Most small-scale wetland construction sites lack basic ecological data, such as topographic elevation, hydrological conditions, soil characteristics, and biological habitat conditions. Government departments or owners find it difficult to invest significant human and financial resources in conducting detailed data surveys and analyses, leading to a heavy reliance on experience-based judgment during the design process, which affects the scientific validity and effectiveness of achieving the functions of small-scale wetlands.

[0004] 2. Insufficient ability to nest multiple scenarios: Traditional wetland design relies on customized solutions, which makes it difficult to quickly adapt to different urban spaces, such as old residential areas, urban waterways, parks and green spaces, and industrial parks. It lacks a unified design logic and a portable structure, which restricts its large-scale application in urban systems.

[0005] 3. Lack of modular construction mechanism, resulting in low efficiency in solution construction: At present, the wetland construction process relies on case-by-case design, which leads to long design cycles, high costs, and difficulty in forming a path for large-scale promotion.

[0006] Therefore, there is an urgent need to develop a construction method and system with digital evaluation capabilities, parameterized combination capabilities, and high adaptability to improve the construction efficiency of urban micro-wetlands. Summary of the Invention

[0007] The purpose of this invention is to address the problems of insufficient rapid site assessment methods, inadequate multi-scenario nesting capabilities, lack of parameterized construction mechanisms, and low construction efficiency in existing urban micro-wetland design scheme construction methods. This invention provides a digital assessment and parameterized combination construction method and system for urban micro-wetlands that possesses digital assessment methods, parameterized combination capabilities, and high adaptability.

[0008] To achieve the above-mentioned objectives, the first aspect of this invention provides a method for digital assessment and parameterized combination construction of urban micro-wetlands, comprising:

[0009] Multi-source remote sensing data of the target site is collected by drone and preprocessed based on image processing technology;

[0010] The ecological factors of the target site are digitally evaluated based on the preprocessed multi-source remote sensing data to obtain site characteristics. The ecological factors include at least hydrological conditions, vegetation conditions, soil conditions, micro-topography and drainage paths, animal activity traces and water quality conditions.

[0011] A standard ecological factor coding label set is constructed based on the preset evaluation criteria of each ecological factor, and the standard ecological factor codes corresponding to the label set are assigned to the target site based on the site characteristics.

[0012] The key ecological functions of urban micro-wetlands related to the ecological factors are broken down into different splicable functional components and a parameterized functional component library is formed. Each functional component has an adaptation rule and a functional component code associated with the standard ecological factor code.

[0013] Determine the scene type and construction goals of the target site to determine the adaptation conditions;

[0014] Based on the standard ecological factor encoding, the parameterized functional component library is invoked;

[0015] Match the corresponding functional components according to the adaptation conditions and adaptation rules, output the combined configuration scheme of functional component codes through parameterized configuration, and generate a design scheme for urban micro-wetlands with scene adaptability and functional coupling logic based on the combined configuration scheme.

[0016] Preferably, the acquisition of multi-source remote sensing data of the target site via UAV remote sensing specifically includes:

[0017] Using drones, we conduct aerial data collection over the target area to acquire raw remote sensing images, including high-resolution visible light images, near-infrared images, hyperspectral images, and multi-angle oblique photographic images. During the drone flight data collection, we ensure that the acquired images have overlap through flight path planning, which is beneficial for later stitching and modeling.

[0018] Preferably, the preprocessing of the multi-source remote sensing data based on image processing technology specifically includes:

[0019] Image denoising and enhancement: Multi-scale guided filtering algorithm and Retinex image enhancement algorithm are used on the original remote sensing image to suppress noise and optimize brightness and contrast, respectively, to improve edge sharpness and color reproduction. This process preserves ecological feature details.

[0020] Image registration and stitching: Multiple UAVs collect images and perform geometric registration through ground control points (GCPs) to ensure spatiotemporal consistency. The structured light reconstruction algorithm (SfM) and the multi-view stereo reconstruction algorithm (MVS) are fused and processed to automatically stitch multi-angle images, generating a continuous and seamless orthophoto map (DOM) and a high-density 3D point cloud.

[0021] Digital surface model (DSM) and digital elevation model (DEM) are generated. Based on the stitched orthophoto map (DOM) and high-density 3D point cloud, the ground and ground features are further distinguished. The influence of tall objects, including buildings and tree canopies, is removed by Boolean filtering algorithm to generate DSM and DEM data.

[0022] Hyperspectral image data inversion integration fuses hyperspectral images with high-density 3D point cloud projection results, and uses typical spectral inversion models to invert water quality parameters, including chlorophyll concentration, total suspended solids, and COD in water bodies. At the same time, it helps to identify different plant types, exposed soil, and algae distribution, forming a multi-level ecological factor map.

[0023] Preferably, the digital assessment includes:

[0024] Combining image recognition algorithms, a digital assessment of ecological factors is performed based on the preprocessed results, outputting site characteristics, including:

[0025] Hydrological information, including DEM, DSM and preprocessed multispectral images, is used to automatically detect waterlogged areas using a deep learning-based terrain segmentation model. The water bodies and dark surfaces are distinguished by combining image spectral features, and the extent and depth of waterlogging are estimated. Through elevation raster analysis, the interference of low-lying flood-prone areas and artificial structures on drainage paths is identified, and the vector boundary and elevation zone distribution map of the waterlogged area are output.

[0026] Vegetation identification and coverage: YOLOv8 and a self-trained convolutional neural network (CNN) model were jointly applied to preprocessed multi-source remote sensing images to perform multi-level vegetation identification, distinguishing between herbaceous, shrub, and tree vegetation, and generating vegetation distribution maps through pixel classification statistics; at the same time, normalized vegetation index (NDVI) and DEM data were fused to estimate vegetation coverage and label different community types, including sparse, dense, and degraded communities.

[0027] Soil exposure assessment utilizes red edge and shortwave infrared (SWIR) band features and image texture features from multispectral images to identify exposed surface areas. Combined with slope data derived from DEM data, the erosion risk level of the exposed areas is assessed, and the results are output as a soil erosion sensitivity map.

[0028] Micro-topography and drainage path analysis: Based on the difference between DSM and DEM data, the micro-topographic undulations are calculated to identify naturally formed small-scale topographic units, including depressions, terraces, and gentle slopes; D8 or D∞ water flow direction algorithms are applied to simulate the direction of rainfall runoff, extract potential drainage paths, and overlay existing drainage facility layers to analyze water catchment obstacles and areas for improvement.

[0029] Animal activity trace identification combines high-resolution visible light images, nighttime thermal infrared images from near-infrared images, and supplementary observation data from artificial ground. Using an animal activity identification model trained on YOLOv5 and a thermal signal classifier, animal activity traces are identified to determine ecologically sensitive areas of the site.

[0030] Water quality analysis combines hyperspectral imagery data with artificial ground water sample testing data. Based on a regression inversion model, multiple indicators of water bodies are estimated, including COD, TN, TP, Chl-a, and TSS. Finally, a water quality classification map is generated, and cross-analysis with vegetation and hydrological factors is performed to output the labeling of potential purification priority areas and heavily polluted areas.

[0031] Preferably, the evaluation criteria specifically include:

[0032] The criteria for assessing hydrological conditions include significant water accumulation (area > 100m² and depth > 10cm), slight water accumulation, and no water accumulation.

[0033] The criteria for judging plant condition include: rich community (multiple types and coverage >70%), sparse vegetation (coverage <30%) or single type, vegetation degradation or bareness;

[0034] Soil condition assessment criteria include high-risk erosion areas, bare but low-risk areas, and areas covered by vegetation.

[0035] The criteria for evaluating micro-topography and drainage pathways include natural catchment depressions, hills / no-catchment features, and clearly defined drainage channels;

[0036] The criteria for judging animal activity traces include the presence of rare animals, the presence of common animals but with abundant populations, and the presence of rare animals.

[0037] Water quality assessment criteria include good water quality, moderate water quality, and significant water pollution.

[0038] Preferably, the combined configuration scheme of the output functional components through parameterized configuration includes:

[0039] Based on the dynamic matching relationship between the evaluation criteria of standard ecological factor coding and the adaptation rules of functional component coding, automatic combination recommendations are made, and the priority of each functional component in the functional component combination configuration scheme is optimized based on the adaptation conditions.

[0040] Designers can adjust the combination of functional components according to actual needs.

[0041] Preferably, the design scheme for the urban micro-wetland includes:

[0042] Component configuration recommendations: Based on the site's ecological characteristics and design objectives, list suitable functional components and their configuration logic.

[0043] The deployment logic is explained, along with the ecological action pathways of the component combinations and their correspondence with site ecological issues.

[0044] The parameter configuration table lists the key parameters of each functional component in the combined configuration scheme.

[0045] To achieve the objectives of this invention, a second aspect provides a system for the digital assessment and parameterized combination construction of urban micro-wetlands, applying the method for the digital assessment and parameterized combination construction of urban micro-wetlands described in the above technical solution. The system includes:

[0046] The remote sensing data acquisition module is used to acquire multi-source remote sensing data of the target site via UAV remote sensing.

[0047] The ecological factor identification and digital assessment module is used to preprocess multi-source remote sensing data and digitally assess ecological factors.

[0048] A parametric functional component library for storing combinable functional components;

[0049] The scene adaptation unit is used to determine the adaptation conditions based on the scene type and construction goals.

[0050] The component recommendation and combination configuration engine is used to call functional components in the parameterized functional component library based on the standard ecological factor coding, and output the combination configuration scheme of functional component coding through parameterized configuration according to the adaptation conditions and adaptation rules.

[0051] The design scheme output module is used to generate urban micro-wetland design schemes based on combined configuration schemes.

[0052] Preferably, the component recommendation and combination configuration engine performs the following operations:

[0053] Based on the ecological factor coding and adaptation conditions, the system automatically recommends combinations and optimizes the priority of each functional component in the functional component combination configuration scheme based on the adaptation conditions.

[0054] Preferably, the design scheme generated by the design scheme output module includes:

[0055] Component configuration recommendations: Based on the site's ecological characteristics and design objectives, list suitable functional components and their configuration logic.

[0056] The deployment logic is explained, along with the ecological action pathways of the component combinations and their correspondence with site ecological issues.

[0057] The parameter configuration table lists the key parameters of each functional component in the combined configuration scheme.

[0058] Compared with the prior art, the beneficial effects of this invention are:

[0059] This invention utilizes UAV remote sensing image acquisition and image processing technology to extract, analyze, and digitally evaluate the basic ecological data of a pre-designed small wetland site. It deconstructs the ecological functions of the small wetland into several functional components, and establishes a correspondence between ecological factor codes and parameterized components to achieve dynamic linkage between the digital evaluation results and modular design. Based on standardized ecological factor codes and adaptation conditions, it intelligently recommends the most suitable functional component combination configuration scheme from a parameterized functional component library to generate an urban small wetland design scheme. The final output scheme possesses scene adaptability and functional coupling logic. This invention features digital evaluation methods, parameterized combination capabilities, and high adaptability, thereby improving the design efficiency of urban small wetland design schemes. Attached Figure Description

[0060] Figure 1 This is a flowchart illustrating the steps of a method for digital assessment and parameterized combination construction of urban micro-wetlands in one embodiment.

[0061] Figure 2 A flowchart illustrating the steps for standard ecological factor coding of a target site in one embodiment;

[0062] Figure 3 This is a flowchart of parameterized construction for multi-scenario adaptation in one embodiment. Detailed Implementation

[0063] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0064] Example 1

[0065] like Figure 1 As shown, the steps of the digital assessment and parameterized combination construction method for urban micro-wetlands in this embodiment 1 include:

[0066] S1: Collect multi-source remote sensing data of the target site using UAV remote sensing and preprocess it based on image processing technology;

[0067] S2: Based on the preprocessed multi-source remote sensing data, the ecological factors of the target site are digitally evaluated to obtain site characteristics. The ecological factors include at least hydrological conditions, vegetation conditions, soil conditions, micro-topography and drainage paths, animal activity traces and water quality conditions.

[0068] S3: Construct a standard ecological factor coding label set according to the preset evaluation criteria of each ecological factor, and assign the standard ecological factor code corresponding to the label set to the target site based on the site characteristics.

[0069] S4: Decompose the key ecological functions of urban micro-wetlands related to the ecological factors into different splicable functional components and form a parameterized functional component library. Each functional component has an adaptation rule and a functional component code associated with the standard ecological factor code.

[0070] S5: Determine the scene type and construction goals of the target site and determine the adaptation conditions;

[0071] S6: Based on the standard ecological factor encoding, call the parameterized functional component library;

[0072] S7: Match corresponding functional components according to adaptation conditions and rules, output a combination configuration scheme of functional component codes through parameterized configuration, and generate a design scheme for urban micro-wetlands with scene adaptability and functional coupling logic based on the combination configuration scheme. Scene adaptability refers to the mutual suitability of adaptation conditions and rules, and functional coupling logic refers to the dynamic matching relationship between standard ecological factor codes and functional component codes.

[0073] Furthermore, the above steps S1-S3 specifically include the following steps, and the flowchart for the standard ecological factor coding of the target site is as follows: Figure 2 As shown:

[0074] (1) Data collection

[0075] Use consumer-grade or industrial-grade drones to fly at low altitudes over the target area to acquire high-resolution RGB images and visible light videos. If possible, use infrared or multispectral sensors to collect data. During the drone flight, ensure image overlap through flight path planning to facilitate later stitching and modeling.

[0076] (2) Data preprocessing

[0077] The primary processing targets are multi-source remote sensing data acquired via UAV platforms, including high-resolution visible light imagery, near-infrared imagery, hyperspectral imagery, and multi-angle oblique photographic images. The goal of preprocessing is to generate high-precision 3D terrain models and image feature enhancement results from the raw image data for subsequent ecological factor identification, providing a data foundation for the accurate extraction of ecological information.

[0078] The specific preprocessing steps are as follows:

[0079] Image denoising and enhancement:

[0080] Multi-scale guided filtering and Retinex image enhancement algorithms are applied to the original remote sensing images to suppress noise and optimize brightness and contrast, thereby improving edge sharpness and color reproduction. This process preserves detailed ecological features, such as vegetation texture, water surface reflection, and soil cracking, providing higher accuracy for subsequent classification and recognition.

[0081] Image registration and stitching:

[0082] Multiple UAVs acquire images and perform geometric registration using ground control points (GCPs) to ensure spatiotemporal consistency. The images are then fused using a structured light reconstruction algorithm (SfM) and a multi-view stereo reconstruction algorithm (MVS) to automatically stitch together multi-angle images, generating a continuous, seamless orthophoto map (DOM) and a high-density 3D point cloud.

[0083] Generating Digital Surface Models (DSM) and Digital Elevation Models (DEM):

[0084] Based on the stitched point cloud data, the ground and ground features are further distinguished. The influence of tall objects (such as buildings and tree canopies) is removed by Boolean filtering algorithm to generate DSM and DEM data, and micro-topographic information such as terrain undulation, depressions, and drainage paths is extracted from them.

[0085] Hyperspectral data inversion integration:

[0086] By fusing hyperspectral images with point cloud projection results and using material identification algorithms, this embodiment employs typical spectral inversion models, such as the continuous spectral ratio index and random forest regression method, to invert water quality parameters such as chlorophyll concentration, total suspended matter, and COD in water bodies. At the same time, it assists in identifying different plant types, exposed soil, and algae distribution, forming a multi-level ecological factor map.

[0087] (3) Digital assessment of ecological factors

[0088] Combining image recognition algorithms such as YOLO, UNet, or self-trained CNN models, ecological factors are identified and classified in images, outputting key site features, including:

[0089] Hydrological information:

[0090] Based on DEM / DSM, DOM, and optical images, a deep learning-based terrain segmentation model (such as a variant of UNet) is used to automatically detect waterlogged areas. Combined with image spectral features, water bodies are distinguished from dark surfaces, and the extent and depth of waterlogging are estimated. Through elevation raster analysis, low-lying, flood-prone areas and the interference of man-made structures on drainage paths are identified, and vector boundaries and elevation zone distribution maps of waterlogged areas are output.

[0091] Vegetation identification and coverage calculation:

[0092] YOLOv8 and a self-trained convolutional neural network (CNN) model were jointly applied to multi-source images for multi-level vegetation recognition. The model distinguishes between herbaceous, shrub, and tree vegetation, and generates vegetation distribution maps through pixel classification statistics. Simultaneously, NDVI (Normalized Difference Vegetation Index), DOM (Domain of Degradation) data, and DSM (Digital Standard Model) elevation data were integrated to estimate vegetation cover and label different community types such as sparse, dense, and degraded vegetation.

[0093] Soil exposure assessment:

[0094] Exposed surface areas were identified using red-edge and shortwave infrared (SWIR) bands and image texture features from multispectral imagery. The erosion risk level of these exposed areas was then assessed using slope data (derived from DEM) and DOM data. The output is a "Soil Erosion Sensitivity Map," used as a reference for erosion resistance design in component configurations.

[0095] Micro-topography and drainage path analysis:

[0096] Micro-topographic relief is calculated based on the difference between DSM and DEM data to identify naturally formed small-scale topographic units such as depressions, terraces, and gentle slopes. The D8 or D∞ flow direction algorithm is applied to simulate rainfall runoff direction, extract potential drainage paths, and overlay existing drainage facility layers to analyze water catchment obstacles and areas requiring modification.

[0097] Animal activity traces:

[0098] By combining visible light imagery, nighttime thermal infrared imagery, and supplementary ground observation data, a trained animal activity recognition model (based on YOLOv5 + thermal signal classifier) ​​is used to identify signs of habitat or passage for small mammals, birds, and amphibians, such as animal passages, drinking points, and trampling marks, in order to determine ecologically sensitive areas of the site.

[0099] Water quality analysis:

[0100] Integrating hyperspectral remote sensing data with ground water sample monitoring data, this study estimates multiple indicators of water bodies based on regression inversion models (such as PLSR partial least squares regression and random forest RF), including COD, TN, TP, Chl-a, and TSS. Finally, a water quality classification map is generated, and cross-analysis with vegetation and hydrological factors is performed to output labels for potential remediation priority areas and heavily polluted areas.

[0101] (4) Parameterized output of site ecological factors (standard ecological factor coding)

[0102] Based on preset evaluation criteria, a standard ecological factor coding label set is constructed, as shown in Table 1. The target site is assigned the corresponding standard ecological factor codes from the label set based on the site characteristics, and the output coding supports linkage with subsequent functional component coding.

[0103]

[0104] Table 1: Ecological Factors and Their Evaluation Criteria

[0105] Steps S4-S7 provide a parameterized construction mechanism for multi-scenario adaptation, as illustrated in the flowchart below. Figure 3 As shown, the details are as follows:

[0106] Key ecological functions of wetlands (such as topography shaping, hydrological regulation, wastewater purification, plant configuration, and habitat construction) are pre-divided into multiple standard functional components that can be assembled. Each component has an independent functional component code, design parameters, and adaptation rules (such as plant community modules, water purification modules, habitat modules, and gentle slope modules). Multiple components can be combined and deployed. Furthermore, the parameterized functional component library constructed in step S4 is shown in Table 2.

[0107]

[0108] Table 2: Modular Library and Coding of Functional Components

[0109] Further, step S5 includes:

[0110] The system determines the applicable scenarios based on site type (such as old residential areas, parks and green spaces, industrial parks, and urban waterways) and construction goals (such as rainwater storage, biological habitat, and enhanced purification functions). Site type and construction goals can be input by the user or automatically identified based on the built-in matching scheme logic. The built-in matching scheme is shown in Table 3.

[0111]

[0112] Table 3: Description of Scenario Types, Characteristics, and Development Goals of Built-in Matching Schemes

[0113] Further, steps S6-S7 include:

[0114] (1) Based on the assigned standard ecological factor codes, the parameterized functional component library is invoked. Corresponding functional components are matched according to adaptation conditions and rules. A combination configuration scheme for the functional component codes is output through parameterized configuration. Based on the dynamic matching relationship between the evaluation criteria of the standard ecological factor codes and the adaptation rules of the functional component codes, automatic combination recommendations are made. The priority of each functional component in the combination configuration scheme is optimized based on adaptation conditions. Table 4 shows a built-in multi-scenario functional component combination configuration scheme. Designers can adjust the functional component combinations according to actual needs and generate component deployment suggestions and design schemes in real time based on the adjustments.

[0115]

[0116] Table 4: Parametric Combination Configuration Scheme for Multi-Scenario Functional Components

[0117] (2) Based on the above combined configuration scheme, generate a design scheme for urban micro-wetlands:

[0118] After completing the digital assessment of ecological factors and the recommendation of parameterized component combinations, a design scheme with clear guiding significance is generated. The final output scheme has scene adaptability and functional coupling logic, which can support designers to carry out wetland construction work under different site types.

[0119] Example 2

[0120] This embodiment 2 further illustrates the urban micro-wetland design scheme generated in step S7 of embodiment 1.

[0121] The design scheme for urban micro-wetlands specifically includes:

[0122] Component configuration recommendations: Based on the site digital assessment results and construction goals, list the combination configuration scheme, including a list of suitable functional components and their configuration logic (such as site characteristics, layout location, combination sequence, main function, etc.).

[0123] The deployment logic is explained, along with the ecological action pathways of the component combinations and their correspondence with site ecological issues.

[0124] Based on the parameter configuration table of the preset template, list the key parameters of each recommended component (such as size, storage capacity, planting density, plant variety, etc.) to facilitate further optimization and adjustment by designers.

[0125] This design solution output mechanism eliminates the need for designers to start from scratch with functional analysis and logical planning. Based on existing evaluation data and recommended combinations, designers can conduct more targeted and adaptable designs, establishing a clear logical connection between design concepts, component selection, and site response. This enhances the systematic nature and feasibility of the design solution and improves design efficiency. Compared to traditional design processes that rely on manual experience, this design solution features automatic analysis, module recommendation, and parameter configuration functions, shortening the design cycle and improving the scientific rigor and adaptability of the design.

[0126] Example 3

[0127] This embodiment 3, based on embodiment 1, provides a system for the digital assessment and parametric combination construction of urban micro-wetlands. It applies the methods for the digital assessment and parametric combination construction of urban micro-wetlands described in embodiments 1 and 2. The system includes:

[0128] The remote sensing data acquisition module is used to acquire multi-source remote sensing data of the target site via UAV remote sensing.

[0129] The ecological factor identification and digital assessment module is used to preprocess multi-source remote sensing data and digitally assess ecological factors.

[0130] A parametric functional component library for storing combinable functional components;

[0131] The scene adaptation unit is used to determine the adaptation conditions based on the scene type and construction goals.

[0132] The component recommendation and combination configuration engine is used to call functional components in the parameterized functional component library based on the standard ecological factor coding, and output the combination configuration scheme of functional component coding through parameterized configuration according to the adaptation conditions and adaptation rules.

[0133] The design scheme output module is used to generate urban micro-wetland design schemes based on combined configuration schemes.

[0134] Furthermore, the component recommendation and combination configuration engine performs the following operations:

[0135] Based on the ecological factor coding and adaptation conditions, the system automatically recommends combinations and optimizes the priority of each functional component in the functional component combination configuration scheme based on the adaptation conditions.

[0136] Furthermore, the design scheme generated by the design scheme output module includes:

[0137] Component configuration recommendations: Based on the site's ecological characteristics and design objectives, list suitable functional components and their configuration logic.

[0138] The deployment logic is explained, along with the ecological action pathways of the component combinations and their correspondence with site ecological issues.

[0139] Based on the parameter configuration table configured in the preset template, list the key parameters of each functional component in the combined configuration scheme.

[0140] In summary, the present invention has the following advantages and effects compared to the prior art:

[0141] The above embodiments of the present invention provide a low-cost and rapidly deployable digital site ecological factor assessment mechanism. It utilizes UAV remote sensing images and image recognition algorithms to extract key ecological data of the target site, including site waterlogging conditions (whether there is short-term waterlogging and water body elevation), hydrological characteristics, surface vegetation cover and plant identification, soil exposure degree, topographic micro-elevation differences and drainage paths, and animal habitat traces, thereby enabling quantitative analysis of the site's basic ecological conditions and judgment of design schemes.

[0142] Building upon this foundation, the system constructs a parametric module library composed of various ecological functional components. The core functions of small wetlands (such as hydrological regulation, topographic shaping, vegetation configuration, wastewater purification, and habitat creation) are broken down into standardized "plug-in units." Designers can use these parametric control modules to make targeted improvements and enhancements based on assessment results and construction goals, addressing site-specific issues. This enables efficient integration and rapid deployment across multiple scenarios, including parks, urban waterways, old residential areas, and industrial parks.

[0143] The system boasts the combined advantages of "data-driven approach + modular integration + multi-scenario adaptation," enhancing both the intelligence and responsiveness of the design process, and strengthening the feasibility, replicability, and ecosystem service effectiveness of small wetlands as green infrastructure in complex urban spaces. Compared to traditional design processes relying on manual experience, this design scheme features automatic analysis, module recommendation, and parameter configuration functions, shortening the design cycle and improving the scientific rigor and adaptability of the design. This method can be widely applied in areas such as urban ecological restoration, habitat improvement, and low-impact development, demonstrating significant ecological and engineering value and potential for widespread adoption.

[0144] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for digital assessment and parameterized combination construction of urban micro-wetlands, characterized in that, Includes the following steps: Multi-source remote sensing data of the target site is collected by drone and preprocessed based on image processing technology; The ecological factors of the target site are digitally evaluated based on the preprocessed multi-source remote sensing data to obtain site characteristics. The ecological factors include at least hydrological conditions, vegetation conditions, soil conditions, micro-topography and drainage paths, animal activity traces and water quality conditions. A standard ecological factor coding label set is constructed based on the preset evaluation criteria of each ecological factor, and the standard ecological factor codes corresponding to the label set are assigned to the target site based on the site characteristics. The key ecological functions of urban micro-wetlands related to the ecological factors are broken down into different splicable functional components according to the scenario requirements and a parameterized functional component library is formed. Each functional component has an adaptation rule and a functional component code coupled with the standard ecological factor coding function. Determine the scene type and construction goals of the target site and determine the adaptation conditions; Based on the assigned standard ecological factor codes, the parameterized functional component library is invoked; Match the corresponding functional components according to the adaptation conditions and adaptation rules, output the combined configuration scheme of functional component codes through parameterized configuration, and generate a design scheme for urban micro-wetlands with scene adaptability and functional coupling logic based on the combined configuration scheme.

2. The method according to claim 1, characterized in that, The acquisition of multi-source remote sensing data of the target site via UAV remote sensing specifically includes: Using drones, we conduct aerial data collection over the target area to acquire raw remote sensing images, including high-resolution visible light images, near-infrared images, hyperspectral images, and multi-angle oblique photographic images. During the drone flight data collection, we ensure that the acquired images have overlap through flight path planning, which is beneficial for later stitching and modeling.

3. The method according to claim 2, characterized in that, The preprocessing of the multi-source remote sensing data based on image processing technology specifically includes: Image denoising and enhancement: Multi-scale guided filtering algorithm and Retinex image enhancement algorithm are used on the original remote sensing image to suppress noise and optimize brightness and contrast, respectively, to improve edge sharpness and color reproduction. This process preserves ecological feature details. Image registration and stitching: Multiple UAVs collect images and perform geometric registration through ground control points (GCPs) to ensure spatiotemporal consistency. The structured light reconstruction algorithm (SfM) and the multi-view stereo reconstruction algorithm (MVS) are fused and processed to automatically stitch multi-angle images, generating a continuous and seamless orthophoto map (DOM) and a high-density 3D point cloud. Digital surface model (DSM) and digital elevation model (DEM) are generated. Based on the stitched orthophoto map (DOM) and high-density 3D point cloud, the ground and ground features are further distinguished. The influence of tall objects, including buildings and tree canopies, is removed by Boolean filtering algorithm to generate DSM and DEM data. Hyperspectral image data inversion integration fuses hyperspectral images with high-density 3D point cloud projection results, and uses typical spectral inversion models to invert water quality parameters, including chlorophyll concentration, total suspended solids, and COD in water bodies. At the same time, it helps to identify different plant types, exposed soil, and algae distribution, forming a multi-level ecological factor map.

4. The method according to claim 3, characterized in that, The digital assessment includes: Combining image recognition algorithms, a digital assessment of ecological factors is performed based on the preprocessed results, outputting site characteristics, including: Hydrological information, including DEM, DSM and preprocessed multispectral images, is used to automatically detect waterlogged areas using a deep learning-based terrain segmentation model. The water bodies and dark surfaces are distinguished by combining image spectral features, and the extent and depth of waterlogging are estimated. Through elevation raster analysis, the interference of low-lying flood-prone areas and artificial structures on drainage paths is identified, and the vector boundary and elevation zone distribution map of the waterlogged area are output. Vegetation identification and coverage: YOLOv8 and a self-trained convolutional neural network (CNN) model were jointly applied to preprocessed multi-source remote sensing images to perform multi-level vegetation identification, distinguishing between herbaceous, shrub, and tree vegetation, and generating vegetation distribution maps through pixel classification statistics; at the same time, normalized vegetation index (NDVI) and DEM data were fused to estimate vegetation coverage and label different community types, including sparse, dense, and degraded communities. Soil exposure assessment utilizes red edge and shortwave infrared (SWIR) band features and image texture features from multispectral images to identify exposed surface areas. Combined with slope data derived from DEM data, the erosion risk level of the exposed areas is assessed, and the results are output as a soil erosion sensitivity map. Micro-topography and drainage path analysis: Based on the difference between DSM and DEM data, the micro-topographic undulations are calculated to identify naturally formed small-scale topographic units, including depressions, terraces, and gentle slopes; D8 or D∞ water flow direction algorithms are applied to simulate the direction of rainfall runoff, extract potential drainage paths, and overlay existing drainage facility layers to analyze water catchment obstacles and areas for improvement. Animal activity trace identification combines high-resolution visible light images, nighttime thermal infrared images from near-infrared images, and supplementary observation data from artificial ground. Using an animal activity identification model trained on YOLOv5 and a thermal signal classifier, animal activity traces are identified to determine ecologically sensitive areas of the site. Water quality analysis combines hyperspectral imagery data with artificial ground water sample testing data. Based on a regression inversion model, multiple indicators of water bodies are estimated, including COD, TN, TP, Chl-a, and TSS. Finally, a water quality classification map is generated, and cross-analysis with vegetation and hydrological factors is performed to output the labeling of potential purification priority areas and heavily polluted areas.

5. The method according to claim 1, characterized in that, The evaluation criteria specifically include: The criteria for assessing hydrological conditions include significant water accumulation (area > 100m² and depth > 10cm), slight water accumulation, and no water accumulation. The criteria for judging plant condition include: rich community (multiple types and coverage >70%), sparse vegetation (coverage <30%) or single type, vegetation degradation or bareness; Soil condition assessment criteria include high-risk erosion areas, bare but low-risk areas, and areas covered by vegetation. The criteria for evaluating micro-topography and drainage pathways include natural catchment depressions, hills / no-catchment features, and clearly defined drainage channels; The criteria for judging animal activity traces include the presence of rare animals, the presence of common animals but with abundant populations, and the presence of rare animals. Water quality assessment criteria include good water quality, moderate water quality, and significant water pollution.

6. The method according to claim 5, characterized in that, The combined configuration scheme of output functional components through parameterized configuration includes: Based on the dynamic matching relationship between the evaluation criteria of standard ecological factor coding and the adaptation rules of functional component coding, automatic combination recommendations are made, and the priority of each functional component in the functional component combination configuration scheme is optimized based on the adaptation conditions. Designers can adjust the combination of functional components according to actual needs.

7. The method according to claim 1, characterized in that, The urban micro-wetland design scheme includes: Component configuration recommendations: Based on the site digital assessment results and construction goals, list suitable functional components and their configuration logic. The deployment logic is explained, along with the ecological action pathways of the component combinations and their correspondence with site ecological issues. Based on the parameter configuration table configured in the preset template, list the key parameters of each functional component in the combined configuration scheme.

8. A system for digital assessment and parameterized combination construction of urban micro-wetlands, employing the method for digital assessment and parameterized combination construction of urban micro-wetlands as described in claims 1-7, characterized in that: The system includes: The remote sensing data acquisition module is used to acquire multi-source remote sensing data of the target site via UAV remote sensing. The ecological factor identification and digital assessment module is used to preprocess multi-source remote sensing data and digitally assess ecological factors. A parametric functional component library for storing combinable functional components; The scene adaptation unit is used to determine the adaptation conditions based on the scene type and construction goals. The component recommendation and combination configuration engine is used to call functional components in the parameterized functional component library based on the standard ecological factor coding, and output the combination configuration scheme of functional component coding through parameterized configuration according to the adaptation conditions and adaptation rules. The design scheme output module is used to generate urban micro-wetland design schemes based on combined configuration schemes.

9. The system according to claim 8, characterized in that, The component recommendation and combination configuration engine performs the following operations: Based on the ecological factor coding and adaptation conditions, the system automatically recommends combinations and optimizes the priority of each functional component in the functional component combination configuration scheme based on the adaptation conditions.

10. The system according to claim 8, characterized in that, The design schemes generated by the design scheme output module include: Component configuration recommendations: Based on the site's ecological characteristics and design objectives, list suitable functional components and their configuration logic. The deployment logic is explained, along with the ecological action pathways of the component combinations and their correspondence with site ecological issues. Based on the parameter configuration table configured in the preset template, list the key parameters of each functional component in the combined configuration scheme.

Images