Method and system for extracting peat bog information based on GIS remote sensing image

Abstract

The present application relates to the technical field of image processing, in particular to a peat bog information extraction method and system based on GIS remote sensing image. The method collects multi-temporal optical remote sensing image, DEM data and pixel precipitation, pre-processes the multi-temporal optical remote sensing image, and performs semantic segmentation to identify the region of interest; the DEM data is analyzed for elevation characteristics to extract pixel slope and pixel convergence accumulation, the precipitation uniformity is calculated based on the pixel precipitation, and the pixel slope and pixel convergence accumulation are weighted and fused to screen the peat formation potential area; the coupling delay index is calculated based on the time delay of the vegetation green peak and the water supply trend peak value to screen the peat bog suspected area; the area threshold is screened to obtain the target peat bog area and output the result. The present application combines the differences between peat bog and ordinary bog in multi-dimensional characteristics, extracts patches with peat layer characteristics, and avoids misjudgment of peat bog relying on spectral feature recognition.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, specifically to a method and system for extracting peat bog information based on GIS remote sensing imagery. Background Technology

[0002] Peat bogs are a type of wetland in terrestrial ecosystems formed by the incomplete decomposition and accumulation of plant remains under long-term waterlogging, low temperature, and oxygen-deficient conditions. They possess ecological functions such as water conservation and climate regulation. Peat bogs represent an advanced stage of evolution from ordinary bogs under specific environmental conditions. The core difference between the two lies in whether or not a large-scale peat layer is formed. Functionally, peat bogs have a higher priority for ecological protection and climate regulation due to their carbon pool characteristics.

[0003] Current peat bog extraction techniques primarily rely on remote sensing image interpretation and field surveys. Remote sensing utilizes satellite imagery, employing band combinations and National Distance Water Index (NDWI) enhancement for identification, combined with neural network models to achieve automated extraction. Subsequent field surveys validate and correct the remote sensing image interpretation results. However, existing peat bog remote sensing extraction techniques heavily depend on spectral features to identify surface vegetation and humidity. Because peat bogs and ordinary bogs exhibit significant spectral overlap in the RGB and near-infrared bands, surface information alone cannot distinguish whether a peat layer exists. This spectral confusion leads to a large number of ordinary bogs being misidentified as peat bogs. Summary of the Invention

[0004] To address the technical problem of misclassifying ordinary swamps as peat swamps due to reliance on spectral feature recognition in remote sensing image interpretation, this invention aims to provide a method and system for extracting peat swamp information based on GIS remote sensing imagery. The specific technical solution adopted is as follows: This invention proposes a method for extracting peat bog information based on GIS remote sensing imagery, the method comprising: Acquire multi-temporal optical remote sensing images of the area to be extracted at different sampling times, digital elevation model (DEM) data, precipitation and vegetation index of different pixels in the DEM data; preprocess and semantically segment the multi-temporal optical remote sensing images to identify the region of interest in peat bogs; Elevation feature analysis was performed on the DEM data to extract cell slope and cell runoff accumulation. The precipitation uniformity of each region of interest was determined based on the temporal distribution of precipitation. Based on the precipitation uniformity, cell slope and cell runoff accumulation were weighted and fused to obtain the peat bog formation potential index. The peat bog formation potential index was used to screen the regions of interest to obtain peat formation potential areas. The coupling hysteresis index is calculated based on the time delay of the vegetation green peak and the water replenishment trend peak in the peat formation potential area. Based on the coupling hysteresis index, the peat formation potential area is screened to obtain the suspected peat swamp area. The suspected peat bog area is screened by area threshold to obtain the target peat bog area; the result of the target peat bog area is then output.

[0005] Furthermore, the preprocessing of the multi-temporal optical remote sensing images includes: The sampling time span covers at least two complete growth periods, and radiometric calibration, atmospheric correction, geometric fine correction, and cloud masking are performed on the multi-temporal optical remote sensing images.

[0006] Furthermore, the semantic segmentation method includes: Based on U-Net, a semantic segmentation model was trained on the labeled remote sensing image dataset to obtain the semantic segmentation model; the semantic segmentation model was then used to classify the preprocessed multi-temporal optical remote sensing images pixel by pixel to identify all swamp areas as regions of interest.

[0007] Furthermore, the method for obtaining the uniformity of precipitation includes: At each sampling time, the average precipitation of all pixels in different regions of interest is calculated as the regional precipitation; the coefficient of variation of regional precipitation at all sampling times for each region of interest is calculated; the regional coefficient of variation is normalized by the maximum and minimum values ​​to obtain the precipitation uniformity of each region of interest.

[0008] Furthermore, the method for obtaining the pixel slope and pixel confluence accumulation includes: Geometric registration of remote sensing images is performed based on DEM data. The maximum slope change rate is calculated for each pixel in the DEM data using the maximum slope gradient algorithm to obtain the pixel slope. The flow direction of each pixel is calculated based on the DEM data using the D8 single flow direction algorithm. The number of grids flowing into each pixel from upstream is counted through the flow direction grid to obtain the pixel confluence accumulation.

[0009] Furthermore, the method for obtaining the peat bog formation potential index includes: The adjustment amplitude coefficient is preset based on the characteristics of different regions of interest; the slope factor weight and the runoff accumulation factor weight of the region are calculated based on the precipitation uniformity and the adjustment amplitude coefficient of different regions of interest; the slope factor weight is positively correlated with the precipitation uniformity; the runoff accumulation factor weight is negatively correlated with the precipitation uniformity.

[0010] The slope factor is obtained by normalizing the pixel slope to its maximum and minimum values ​​and then inverting the values. The pooling accumulation factor is obtained by normalizing the pixel pooling accumulation to its maximum and minimum values. The slope factor and its weights are multiplied together to obtain the slope index. The pooling accumulation factor and its weights are multiplied together to obtain the pooling accumulation index. The slope index and the pooling accumulation index are summed together to obtain the peat bog formation potential index.

[0011] Furthermore, the method for obtaining the peat formation potential zone includes: Regions of interest with a peat bog formation potential index greater than or equal to a preset potential threshold are designated as peat bog formation potential zones.

[0012] Furthermore, the method for obtaining the suspected peat bog area includes: For each pixel, the time series curve of the normalized vegetation index for each region is extracted and smoothed. The time point of the maximum value during the growing season is taken as the date of vegetation green peak. The daily precipitation time series is subjected to sliding smoothing processing with a preset time window, and the peak date of the curve is extracted to obtain the peak date of water replenishment trend. The time interval between the vegetation green peak date and the peak date of the water replenishment trend was calculated to obtain the lag days for each pixel. The lag days were normalized based on the plant growth period to obtain the lag coefficient. The lag coefficient was multiplied by the peat bog formation potential index to obtain the coupled lag index for each pixel. The average coupled lag index of each pixel in the peat bog formation potential area was taken to obtain the regional lag index for the peat bog formation potential area. The regional lag indices of all peat bog formation potential areas were processed using the Otsu method to obtain the suspected threshold. By comparing the suspected threshold and the regional lag index, suspected peat bog areas were screened out.

[0013] Furthermore, the method for obtaining the target peat bog area includes: Independent patches were extracted from suspected peat bog areas using the eight-neighbor connectivity criterion. The number of pixels in each patch was counted and the area was calculated. An area threshold was set based on the empirical value of the smallest contiguous unit in the peat bog region. The area threshold was compared with the area to obtain contiguous suspected areas. Extract the regional contours of contiguous suspected areas at the initial time during the first two growth periods, extract the internal overlapping area of ​​the two contours and the area of ​​the union region of the patch boundaries, calculate their intersection-union ratio, and obtain the boundary drift index; set the boundary drift threshold by the actual hydrological fluctuation intensity of the region, and compare it with the boundary drift index to obtain the target peat swamp region.

[0014] The present invention also proposes a peat bog information extraction system based on GIS remote sensing imagery. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the method described in any of the foregoing descriptions.

[0015] The present invention has the following beneficial effects: This invention identifies regions of interest (ROIs) in peat bogs through preprocessing and semantic segmentation of multi-temporal optical remote sensing images. Based on a digital elevation model (DEM), it extracts pixel slope and pixel runoff accumulation. Combining this with precipitation uniformity, a weighted fusion method is used to calculate a peat bog formation potential index. This assesses the potential conditions for peat formation from two dimensions: the terrain's ability to collect and retain water. The topographic constraints required for peat layer formation are transformed into quantifiable evaluation indicators, solving the problem that traditional methods cannot accurately identify peat bogs relying solely on surface spectral information. Furthermore, based on the peat formation potential zone, the time delay between the vegetation green peak and the peak of the water recharge trend is calculated to obtain... The coupling hysteresis index, which addresses the lag in vegetation response caused by the water-retention and heat-insulating effects of peat layers, transforms the characteristics of peat layers into calculable data. This index differentiates swamps based on their phenological response mechanisms. Furthermore, by applying area thresholds to suspected peat swamp areas and utilizing the contiguous spatial distribution of peat swamps, non-compliant patches are eliminated. This invention effectively distinguishes between peat swamps and ordinary swamps by constructing a multi-dimensional progressive identification framework. Combined with multi-level indicators, it solves the problem of misjudgment caused by spectral confusion, effectively improving the accuracy and stability of peat swamp information extraction. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments of the present 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a method for extracting peat bog information based on GIS remote sensing imagery, provided as an embodiment of the present invention. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a peat bog information extraction method and system based on GIS remote sensing imagery proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] The following description, in conjunction with the accompanying drawings, details the specific scheme of the peat bog information extraction method and system based on GIS remote sensing imagery provided by this invention.

[0021] Please see Figure 1 The diagram illustrates a flowchart of a peat bog information extraction method based on GIS remote sensing imagery, according to an embodiment of the present invention. The method includes: Step S1: Acquire multi-temporal optical remote sensing images of the area to be extracted at different sampling times, digital elevation model (DEM) data, precipitation and vegetation index of different pixels in the DEM data; preprocess and semantically segment the multi-temporal optical remote sensing images to identify the region of interest in the peat bog.

[0022] This invention addresses complex geographical scenarios such as alpine wetlands and peat bogs. Because peat bogs and ordinary bogs exhibit highly similar spectral characteristics in the visible and near-infrared bands, the essential differences in peat layers cannot be distinguished during the interpretation of single-temporal optical remote sensing images, easily leading to misclassification of ordinary bogs as peat bogs. Therefore, this invention acquires multi-temporal optical remote sensing images of the area to be extracted at different sampling times, performs preprocessing and semantic segmentation, identifies regions of interest (ROIs) within the peat bog, and combines this with digital elevation model (DEM) data and precipitation data from different pixels of the DEM data to screen for potential peat formation areas in subsequent processes. The acquired vegetation index serves as an influencing factor for extracting the vegetation green peak date in subsequent processes.

[0023] Preferably, in some implementations of the present invention, the method for acquiring multi-temporal optical remote sensing images of the area to be extracted at different sampling times, preprocessing and semantically segmenting the multi-temporal optical remote sensing images, and identifying the region of interest in peat bogs includes: Multi-temporal optical remote sensing images of the target area to be extracted, acquired at different sampling times using optical satellites, are used to identify peat bogs. The sampling time span covers at least two complete growing seasons. Considering the phenological characteristics of vegetation in high-altitude regions, whose growing season is generally 5 to 7 months, the sampling time span is set to 16 months to ensure that the identification results of the acquired multi-temporal optical remote sensing images are not affected by time factors. The acquired raw multi-temporal optical remote sensing images are usually affected by noise, atmosphere, and cloud shadows, requiring preprocessing. Preprocessing includes radiometric calibration, atmospheric correction, geometrical fine correction, and cloud masking. Radiometric calibration unifies the dimensions of images from different time phases; atmospheric correction removes atmospheric effects and restores the color and reflectance characteristics of the images; geometrical fine correction accurately aligns the images spatially, ensuring that each pixel corresponds to the actual geographical location; and cloud masking identifies and marks areas covered by clouds and cloud shadows as invalid. Through preprocessing, the raw multi-temporal optical remote sensing images are transformed into reliable image data with noise reduction, accurate spectrum, precise location, and marked invalid areas. We acquired an open-source semantic segmentation dataset of remote sensing images of peat bogs, and then trained a semantic segmentation model on the labeled dataset using the U-Net algorithm to obtain a semantic segmentation model that can be used for peat bog identification. We then used the semantic segmentation model to classify the preprocessed multi-temporal optical remote sensing images pixel by pixel, and preliminarily identified all bog areas as regions of interest (ROIs) by filtering the classification results. However, the identified ROIs may contain both peat bogs and ordinary bogs, so further filtering is needed.

[0024] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining digital elevation model (DEM) data, precipitation data, and vegetation index of different pixels in the DEM data includes: Digital elevation model (DEM) data with the same spatial range as the multi-temporal optical remote sensing imagery is acquired. The DEM data is extracted by photogrammetric processing of high-resolution optical satellite imagery and adopts the same geographic coordinate system and spatial resolution as the multi-temporal optical remote sensing imagery.

[0025] Precipitation data for different pixels in the DEM data are obtained. The precipitation data is derived from observation data from meteorological stations. Data with the same geographic coordinate system and spatial resolution as the DEM data is generated by spatial interpolation.

[0026] The vegetation index NDVI corresponding to the multi-temporal optical remote sensing image is obtained. The vegetation index is calculated by normalizing the multi-temporal optical remote sensing image after radiometric calibration and atmospheric correction.

[0027] Step S2: Perform elevation feature analysis on the DEM data to extract cell slope and cell runoff accumulation; determine the precipitation uniformity of each region of interest based on the temporal distribution of precipitation; perform weighted fusion of cell slope and cell runoff accumulation based on precipitation uniformity to obtain the peat bog formation potential index; screen regions of interest using the peat bog formation potential index to obtain peat formation potential zones.

[0028] This invention focuses on the extraction of information about peat bogs. The formation of peat bogs depends on a long-term stable water accumulation environment, which requires a continuous and stable water supply and the effective retention of water by local topography. The continuous and stable water supply mainly comes from regional precipitation. The precipitation uniformity, determined by the temporal distribution of precipitation in each region of interest, can characterize the stability of the regional water supply. The greater the precipitation uniformity, the more stable the regional water supply. The effective retention of water mainly depends on the slope and runoff accumulation in the local topography. Slope characterizes the degree of inclination of the local topography. A steeper topography indicates water loss, and a steeper slope indicates a weaker water retention capacity. Runoff accumulation quantifies the topography's ability to collect water; a larger runoff accumulation indicates a stronger water collection capacity.

[0029] By analyzing the elevation features of the DEM (Digital Elevation Model) data, pixel slope and pixel runoff accumulation were extracted. Based on precipitation uniformity, these two parameters were weighted and fused to obtain a peat bog formation potential index. The formation potential index is a combined indicator of water supply and the effective retention of water by local topography, representing the stability of the waterlogged environment. A higher value indicates a more stable waterlogged environment and a greater potential for peat bog formation. Therefore, the formation potential index is used to assess the formation potential of peat bogs and screen potential peat bog formation areas.

[0030] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the peat bog formation potential index includes: Using the same geographic coordinate system for the DEM data and based on the digital elevation model (DEM), geometric registration was performed on multi-temporal optical remote sensing images to achieve pixel-level spatial alignment. This geometric registration employed a polynomial correction method, calculating correction parameters by selecting ground control points uniformly distributed within the multi-temporal optical remote sensing image range. These control points were derived from high-resolution satellite imagery. The DEM data underwent depression filling, and the D8 flow direction algorithm was used to calculate the flow direction of each pixel (x, y). This involved determining the direction of maximum elevation difference between each pixel and its eight neighbors as the outflow direction. Then, based on the flow direction raster, the number of upstream raster cells flowing into each pixel was counted to obtain the pixel's cumulative runoff. The depression-filling process is a well-known technique in this regard and will not be elaborated here. For each pixel in the DEM data, the maximum slope change rate of the pixel is calculated using its elevation value and that of its surrounding 3×3 neighboring pixels, and is taken as the pixel slope. .

[0031] For precipitation data of different pixels in the DEM data, the average precipitation of all pixels in different regions of interest is calculated at each sampling time as the regional precipitation. Then, the average regional precipitation of each region of interest is calculated at all sampling times. and standard deviation Calculate the regional coefficient of variation The precipitation uniformity was calculated by normalizing the regional variation coefficient to its maximum and minimum values. The formula for obtaining the uniformity of precipitation is expressed as: , The value ranges from 0 to 1. A larger value indicates better uniformity of precipitation across months. If the regional precipitation average is calculated... If the area of ​​interest receives no precipitation during a certain period, it is directly determined that the area does not have the potential to form a peat bog, and a peat bog formation potential index is set. And skip the subsequent steps of calculating weights and potential indices.

[0032] For the average precipitation For regions of interest, an adjustment amplitude coefficient k is preset based on the characteristics of different regions of interest. The value of k ranges from 0 to 0.5, which is used to control the sensitivity of precipitation uniformity to the adjustment of weights. The larger the value of k, the greater the change in weights with precipitation uniformity, ensuring that the slope factor and runoff accumulation factor always maintain a synergistic effect; then, based on the precipitation uniformity of different regions of interest... Set the baseline value for slope factor weight. and the benchmark value of the cumulative flow factor The benchmark value allows for symmetrical adjustment of the weights, ensuring that the sum of the two weights is always 1. In a specific implementation of this invention, the slope factor weight... and the weight of the cumulative flow factor The formula for obtaining it is: ; Based on the uniformity of precipitation Dynamically adjust the weights of slope factor and runoff accumulation factor to improve precipitation uniformity. The smaller the value, the higher the slope factor weight. The smaller the weight, the greater the cumulative flow factor. The greater the weight, the more uniform the precipitation. The larger the slope factor weight, the greater the slope factor weight. The larger the value, the greater the cumulative flow factor. The smaller the weight.

[0033] To eliminate quantization differences, pixel slope and pixel confluence cumulative volume Perform maximum and minimum value normalization to obtain the confluence cumulant factor. Since swamps mostly form on gentle slopes, the gentler the slope, the more conducive it is to the formation of a waterlogged environment. Therefore, the slope factor is obtained by inverting the normalized pixel slope. .

[0034] In some implementations of this invention, the peat bog formation potential index The formula for obtaining it is expressed as: Combined with slope factor weights and the weight of the cumulative flow factor Multiply the slope factor and its weight to obtain the slope index; then multiply the runoff accumulation factor and its weight to obtain the runoff accumulation index; finally, sum the slope index and the runoff accumulation index to obtain the peat bog formation potential index. ; The value ranges from 0 to 1, with higher values ​​indicating a greater potential for peat bog formation in the area.

[0035] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the peat formation potential zone includes: Based on the distribution characteristics of the peat bog formation potential index, a potential threshold T=0.6 is preset. T is used to transform the continuously distributed peat bog formation potential index into a binary classification result, enabling quantitative screening of peat bog formation potential areas. The peat bog formation potential index of all regions of interest is compared with the potential threshold. ROI as a potential peat formation zone .

[0036] Step S3: Calculate the coupling hysteresis index based on the time delay of the vegetation green peak and the water replenishment trend peak in the peat formation potential area, and screen the peat formation potential area based on the coupling hysteresis index to obtain the suspected peat swamp area.

[0037] This invention focuses on the extraction of information from peat bogs. The peat layer in peat bogs has water-retention and heat-insulating effects, and the slow growth of bog vegetation results in a lag in the vegetation's response to short-term environmental fluctuations. Specifically, the vegetation green peak appears later than the peak of the water replenishment trend. The vegetation green peak represents the moment of most vigorous vegetation growth, while the peak of the water replenishment trend represents the moment of most abundant regional water replenishment. In contrast, ordinary bogs lack the influence of a peat layer, resulting in rapid vegetation growth and a quick response to temperature and water level fluctuations. The vegetation green peak and the peak of the water replenishment trend are highly synchronized. Based on this difference between peat bogs and ordinary bogs, a coupling hysteresis index is calculated by measuring the time delay between the vegetation green peak and the peak of the water replenishment trend. This index serves as an indicator of the degree to which the date of the vegetation green peak lags behind the peak of the water replenishment trend. Using this index, potential peat bog areas can be further screened within peat formation potential zones.

[0038] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the coupling hysteresis index includes: Based on the Normalized Difference Vegetation Index (NDVI) of each pixel (x, y) in the peat formation potential zone, the time-series curve of the vegetation index for each pixel is smoothed and filtered to remove curve noise, resulting in a smoothed curve. The smoothed curve restores the true vegetation growth trend and can intuitively display the vegetation growth status. The maximum value within the growing season is extracted from the smoothed curve on a daily basis as the vegetation green peak date. The time series of daily precipitation is pre-defined within a time window. Then, a moving average is applied to the series within the window to remove high-frequency fluctuations. The peak date of the smoothed curve is extracted on a daily basis as the peak date of the water replenishment trend. The time window is based on Based on this, a 30-day extension is applied forward and backward. This time window is used to focus precipitation analysis on the period before and after the vegetation green peak, ensuring the temporal matching between the peak of the water replenishment trend and the vegetation green peak.

[0039] Considering the phenomenon that the vegetation green peak in peat bogs appears later than the peak of the water recharge trend, the difference between the vegetation green peak date and the peak date of the water recharge trend for each pixel is defined as the lag days for each pixel. A longer lag period indicates a later peak in vegetation growth, which is more consistent with the water retention and heat insulation effects of peat layers. In a specific implementation of this invention, the formula for obtaining the lag period for each pixel is expressed as: The max function is used to filter out invalid leading responses. The number of lag days represents the lag in vegetation's response to short-term environmental fluctuations. The larger the value, the more obvious the lag.

[0040] Define the coupling hysteresis index for each cell. This is used to quantify the synergistic relationship between hysteresis effects and topographic water accumulation potential. In a specific implementation of this invention, the formula for obtaining the coupling hysteresis index of each pixel is expressed as: × To standardize the units of measurement, we first consider the number of days of lag based on the vegetation growing season. Perform maximum-minimum normalization to obtain the hysteresis coefficient. Then calculate the hysteresis coefficient and the peat bog formation potential index for each pixel. The product of these factors yields the coupling hysteresis index for each pixel.

[0041] Preferably, in some implementations of the present invention, the method for obtaining suspected peat bog areas includes: Coupling hysteresis index for each cell in the peat formation potential zone The average value was taken to obtain the regional hysteresis index L of the peat formation potential area. The regional coupling hysteresis index reflects the overall hysteresis effect and topographic water accumulation potential of the area. The regional hysteresis index of all peat formation potential areas was processed using the Otsu method to obtain the suspected threshold. The Otsu method is a well-known technique in the art and will not be described in detail here.

[0042] Because peat bog vegetation exhibits a significant lag in its response to water changes, displaying a high regional lag index, therefore... The area is considered a suspected peat bog area. .

[0043] Step S4: Filter the suspected peat bog area using area thresholds to obtain the target peat bog area; output the results for the target peat bog area.

[0044] This invention focuses on extracting information about peat bogs. Peat bogs, due to their thick and stable peat layers, can self-sustain a waterlogged environment, thus exhibiting a continuous spatial distribution with high overlap of patch boundaries at different times. Ordinary swamps, lacking the support of a peat layer, rely on short-term surface wet conditions and are easily affected by hydrological fluctuations, resulting in a fragmented spatial distribution and significant interannual drift of patch boundaries. Based on this difference between peat bogs and ordinary swamps, this invention uses a pre-defined area threshold based on the continuous distribution characteristic and then filters and marks suspected peat bog areas using this area threshold. The region is characterized by contiguous distribution; then, patches with high overlap at different times are identified to obtain the target peat bog region.

[0045] Preferably, in some implementations of the embodiments of the present invention, the method for obtaining the target peat bog area includes: suspected peat bog area Connectivity analysis was performed, and the eight-neighborhood connectivity criterion was used to extract independent patches. The area S of each patch P was calculated. The minimum contiguous unit area threshold for medium peat bogs is set using empirical values. This area threshold corresponds to a sufficient number of pixels to meet the statistical reliability requirements for patch identification, while also conforming to the spatial characteristics of contiguous peat bog distribution. The patches were marked as contiguous suspected areas. .

[0046] For the selected contiguous suspected areas Boundary stability analysis was performed, and the initial values ​​during the first two growth phases were extracted. Vector graphic outline and Extract the overlapping area inside the contours at these two moments. Area of ​​the union region with the patch boundary In one specific implementation of this invention, the formula for obtaining the boundary drift index is expressed as: Among them, the intersection and union ratio The boundary drift index is obtained by subtracting the crossover ratio from 1, which reflects the interannual overlap of patch boundaries. Boundary drift index The value ranges from 0 to 1. A smaller value indicates a higher degree of interannual overlap of patch boundaries and better stability; a larger value indicates more significant boundary drift and less stable spatial distribution of patches. When both contrast patches disappear or are not identified, i.e. At that time, set directly To prevent the denominator from being 0.

[0047] The boundary drift threshold is set based on the actual hydrological fluctuation intensity of the contiguous suspected areas. Because peat bogs are self-sustaining waterlogged environments, the interannual overlap of patch boundaries is typically higher than 80%, meaning the boundary drift index is less than 0.2. Therefore, for each ,like If so, it is determined to be the target peat bog area. .

[0048] In one specific implementation of this invention, the method for outputting results includes: Confirm target peat bog area Subsequently, a spatial distribution vector layer of peat bogs is generated. Combined with multi-temporal remote sensing data, the interannual variation characteristics of each patch are output for dynamic monitoring and risk early warning. Outputs include a peat bog distribution map, a map of suspected misclassified areas, a stability grading map, and thematic statistical reports, supporting visualization and spatial analysis on a GIS platform.

[0049] On the other hand, a peat bog information extraction system based on GIS remote sensing imagery is also provided. The system includes a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any of the methods described above.

[0050] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0051] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for extracting peat bog information based on GIS remote sensing imagery, characterized in that, The method includes: Acquire multi-temporal optical remote sensing images of the area to be extracted at different sampling times, digital elevation model (DEM) data, precipitation and vegetation index of different pixels in the DEM data; preprocess and semantically segment the multi-temporal optical remote sensing images to identify the region of interest in peat bogs; Elevation feature analysis was performed on the DEM data to extract cell slope and cell runoff accumulation. The precipitation uniformity of each region of interest was determined based on the temporal distribution of precipitation. Based on the precipitation uniformity, cell slope and cell runoff accumulation were weighted and fused to obtain the peat bog formation potential index. The peat bog formation potential index was used to screen the regions of interest to obtain peat formation potential areas. The coupling hysteresis index is calculated based on the time delay of the vegetation green peak and the water replenishment trend peak in the peat formation potential area. Based on the coupling hysteresis index, the peat formation potential area is screened to obtain the suspected peat swamp area. The suspected peat bog area is screened by area threshold to obtain the target peat bog area; the result of the target peat bog area is then output.

2. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The preprocessing of the multi-temporal optical remote sensing images includes: The sampling time span covers at least two complete growth periods, and radiometric calibration, atmospheric correction, geometric fine correction, and cloud masking are performed on the multi-temporal optical remote sensing images.

3. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The semantic segmentation method includes: Based on U-Net, a semantic segmentation model was trained on the labeled remote sensing image dataset to obtain the semantic segmentation model; the semantic segmentation model was then used to classify the preprocessed multi-temporal optical remote sensing images pixel by pixel to identify all swamp areas as regions of interest.

4. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The method for obtaining the uniformity of precipitation includes: At each sampling time, the average precipitation of all pixels in different regions of interest is calculated as the regional precipitation; the coefficient of variation of regional precipitation at all sampling times for each region of interest is calculated; the regional coefficient of variation is normalized by the maximum and minimum values ​​to obtain the precipitation uniformity of each region of interest.

5. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The methods for obtaining the pixel slope and pixel confluence accumulation include: Geometric registration of remote sensing images is performed based on DEM data. The maximum slope change rate is calculated for each pixel in the DEM data using the maximum slope gradient algorithm to obtain the pixel slope. The flow direction of each pixel is calculated based on the DEM data using the D8 single flow direction algorithm. The number of grids flowing into each pixel from upstream is counted through the flow direction grid to obtain the pixel confluence accumulation.

6. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The methods for obtaining the peat bog formation potential index include: The adjustment amplitude coefficient is preset based on the characteristics of different regions of interest; the slope factor weight and runoff accumulation factor weight of the region are calculated based on the precipitation uniformity and adjustment amplitude coefficient of the different regions of interest; the slope factor weight is positively correlated with the precipitation uniformity; the runoff accumulation factor weight is negatively correlated with the precipitation uniformity. The slope factor is obtained by normalizing the pixel slope to its maximum and minimum values ​​and then inverting the values. The pooling accumulation factor is obtained by normalizing the pixel pooling accumulation to its maximum and minimum values. The slope factor and its weights are multiplied together to obtain the slope index. The pooling accumulation factor and its weights are multiplied together to obtain the pooling accumulation index. The slope index and the pooling accumulation index are summed together to obtain the peat bog formation potential index.

7. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The method for obtaining the peat formation potential zone includes: Regions of interest with a peat bog formation potential index greater than or equal to a preset potential threshold are designated as peat bog formation potential zones.

8. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The method for obtaining the suspected peat bog area includes: For each pixel, the time series curve of the normalized vegetation index for each region is extracted and smoothed. The time point of the maximum value during the growing season is taken as the date of vegetation green peak. The daily precipitation time series is subjected to sliding smoothing processing with a preset time window, and the peak date of the curve is extracted to obtain the peak date of water replenishment trend. The time interval between the vegetation green peak date and the peak date of the water replenishment trend is calculated to obtain the lag days for each pixel. The lag days are normalized based on the plant growth period to obtain the lag coefficient. The lag coefficient is multiplied by the peat bog formation potential index to obtain the coupled lag index for each pixel. The average coupled lag index of each pixel in the peat bog formation potential area is taken to obtain the regional lag index of the peat bog formation potential area. The regional lag index of all peat bog formation potential areas is processed using the Otsu method to obtain the suspected threshold. By comparing the suspected threshold and the regional lag index, suspected peat bog areas are screened out.

9. The method for extracting peat bog information based on GIS remote sensing imagery according to claim 1, characterized in that, The method for obtaining the target peat bog region includes: Independent patches were extracted from suspected peat bog areas using the eight-neighbor connectivity criterion. The number of pixels in each patch was counted and the area was calculated. An area threshold was set based on the empirical value of the smallest contiguous unit in the peat bog region. The area threshold was compared with the area to obtain contiguous suspected areas. Extract the regional contours of contiguous suspected areas at the initial time during the first two growth periods, extract the internal overlapping area of ​​the two contours and the area of ​​the union region of the patch boundaries, calculate their intersection-union ratio, and obtain the boundary drift index; set the boundary drift threshold by the actual hydrological fluctuation intensity of the region, and compare it with the boundary drift index to obtain the target peat swamp region.

10. A peat bog information extraction system based on GIS remote sensing imagery, the system comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 9.

Images