Geographic information acquisition method and system based on remote sensing image

Abstract

The invention discloses a geographic information collection method and system based on a remote sensing image, and relates to the technical field of geographic information extraction, and the method comprises the steps: firstly obtaining a remote sensing image sequence, carrying out the time sequence registration, and extracting pixel gray level change features to generate building edge region pixels; delimiting an edge area, and identifying disturbance points; road node positions are extracted, and a path steering angle is corrected; establishing a disturbance chain communication density value; and finally generating a structured image data unit. According to the method, pixel gray level change features are extracted through time sequence registration, a building edge area is positioned, disturbance points are recognized by combining the disturbance amplitude and gray level direction consistency, path steering deviation is corrected, and the stability of a path communication structure is enhanced; a disturbance chain communication density value is established through an extension path, a continuous response boundary is extracted by fusing a gray scale trend and structural integrity, an image data unit with a gray scale trend, spatial connectivity and disturbance characteristics is generated, and the accuracy and continuity of boundary recognition in a complex region are improved.

Description

Technical Field

[0001] This invention relates to the field of geographic information extraction technology, and more specifically to a geographic information acquisition method and system based on remote sensing imagery. Background Technology

[0002] Geographic information extraction (GIS) technology encompasses methods for extracting, analyzing, and managing geospatial information from various data sources, including remote sensing image analysis, geographic information system (GIS) data processing, spatial data fusion, and topographic modeling. GIS extraction primarily relies on remote sensing, photogrammetry, and spatial data mining, and is widely applied in resource surveys, environmental monitoring, urban planning, and disaster assessment. This technological field systematically covers the entire process from data acquisition, preprocessing, and information extraction to data analysis and storage management. It enhances the accuracy and completeness of geographic information through multi-source data fusion and improves the accuracy of GIS extraction by leveraging image recognition, morphological analysis, and statistical modeling.

[0003] Existing technologies in remote sensing image data processing often employ single-frame or few-frame static analysis methods, making it difficult to capture the dynamic changes in pixel grayscale over time. This results in weak judgment of the temporal continuity of regional responses, affecting the accurate identification of active disturbance areas. In boundary detection, existing methods mainly rely on edge operators and static image features, lacking mechanisms for judging the continuity and changing trends of disturbance directions, easily leading to broken edge contours or misjudgments. In path recognition, existing methods are mostly based on rule extraction or spatial relationship fitting, lacking dynamic correction steps during location connection and turning processes, causing path tracking to deviate or be interrupted. For the final data generation stage, although vectorized structures can form standard coordinate data, it is difficult to combine the connectivity and structural integrity of disturbance chains for composite expression, resulting in missing spatial structural relationships in image data units, reducing the completeness and response accuracy of geographic target boundary description. These problems are particularly prominent in densely populated urban areas or disturbed forest areas, easily causing problems such as blurred boundaries, classification confusion, and incoherent data structures.

[0004] Therefore, how to provide a geographic information acquisition method and system based on remote sensing imagery, generate image data units with grayscale trends, spatial connectivity and disturbance characteristics, and improve the accuracy and continuity of boundary identification in complex areas is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides a geographic information acquisition method and system based on remote sensing imagery. It extracts pixel grayscale variation features through temporal registration, locates building edge areas, identifies disturbance points by combining disturbance amplitude and grayscale direction consistency, and corrects path turning offsets to enhance the stability of path connectivity structures. Furthermore, it establishes disturbance chain connectivity density values ​​by extending the path, and extracts continuous response boundaries by fusing grayscale trends and structural integrity, generating image data units with grayscale trends, spatial connectivity, and disturbance characteristics, thereby improving the accuracy and continuity of boundary identification in complex areas.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a geographic information acquisition method based on remote sensing imagery, comprising: S100: Acquire a remote sensing image sequence, perform temporal registration on the remote sensing image sequence, extract pixel grayscale change features, and generate building edge area pixels; S200: Based on the building edge area pixels, delineate the building edge area, identify disturbance points by combining the disturbance amplitude and grayscale direction consistency, and generate the transition disturbance node distribution value; S300: Based on the distribution value of the transition disturbance nodes, extract the node positions in the road intersection area, and obtain the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames. S400: Based on the path turning correction angle value, establish the perturbation chain connectivity density value by extending the path; S500: Call the connectivity density value of the disturbance chain, calculate the difference in grayscale trend and disturbance response amplitude between the two ends of the chain, analyze the spatial extension direction and structural integrity, extract boundary pixels with continuous response characteristics, and generate structured image data units.

[0007] Preferably, in S100, the steps of acquiring a remote sensing image sequence, performing temporal registration on the remote sensing image sequence, extracting pixel grayscale change features, and generating building edge area pixels include: S101: Obtain a remote sensing image sequence, rearrange the remote sensing image sequence in chronological order, extract the gray values ​​of pixels at the same location in consecutive image frames, determine the continuity of the pixel time series based on the coordinate changes and gray value differences between frames, and generate a continuous pixel gray set. S102: Based on the continuous pixel grayscale set, calculate the grayscale change rate of multiple pixels between adjacent frames, construct a regional grayscale fluctuation map, compare the neighborhood grayscale trend direction of the region where the grayscale change rate is concentrated, mark the corresponding grayscale space segment, and generate a trend response density value. S103: Determine the location of grayscale trend direction change based on the trend response density value, extract the direction turning point to form the boundary structure, form the boundary line, compare the distance between the boundary line and the pixel line corresponding to the existing contour in the original remote sensing image, and based on the set reference value, locate the area with a close offset distance as the building edge area, and generate building edge area pixels.

[0008] Preferably, the expression for the regional grayscale fluctuation map is: ; in, Representing the The pixel grayscale fluctuation trend value in a spatial segment Representing the In the spatial segment, the first The pixel grayscale value of the frame. Representing the In the spatial segment, the first The pixel grayscale value of the frame. Representing the Within a spatial segment The average value of the frame grayscale. This represents the number of time frames involved in the computation within the sequence. This represents the frame index that participates in the grayscale calculation of this spatial segment.

[0009] Preferably, in step S200, the steps of delineating the building edge region based on the building edge region pixels, identifying disturbance points by combining the disturbance amplitude and grayscale direction consistency, and generating the transition disturbance node distribution value include: S201: Obtain the pixels of the building edge area, collect the gray value sequence of multiple rows of pixels, calculate the gray value difference between adjacent pixels, compare the gray value difference amplitude with the gray value disturbance standard value, extract the pixels whose difference amplitude exceeds the standard value, and generate gray value difference interval value. S202: Based on the grayscale difference interval value, extract pixels at the same position in consecutive frames, determine whether the grayscale change direction of the pixels in adjacent frames is consistent, mark the pixel points that meet the direction consistency, and obtain the disturbance direction preservation value. S203: Based on the disturbance direction preservation value, calibrate the location of disturbance points in consecutive frames of remote sensing image, construct a spatial distribution map of disturbance nodes, assign a distribution value corresponding to the location to each disturbance point, and generate the distribution value of transition disturbance nodes.

[0010] Preferably, in step S300, the step of extracting the node positions in the road intersection area based on the distribution values ​​of the transition disturbance nodes, and obtaining the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames includes: S301: Based on the distribution value of the transition disturbance nodes, extract the node position data in the road intersection area, match the positions of nodes with the same number in consecutive frames according to the spatial sequence of nodes in the image frame, organize the corresponding spatial change direction according to the connection order of nodes in consecutive frame images and the time order between frames, and generate a node sequence direction set. S302: Call the node sequence direction set, identify the direction change of each group of nodes, calculate the turning angle between the first two frames and the last two frames; according to the position change trend of nodes with the same number, set the angle reference value, filter out cases where the angle is offset, collect the angle change amplitude of each group of nodes, and obtain the turning angle offset value set. S303: Based on the set of turning angle offset values, select the connection direction of the previous frame in each group of nodes as a reference, adjust the position of the next frame in combination with the angle change range, calculate the position deviation angle before and after adjustment, and obtain the path turning correction angle value based on the correction value of the node.

[0011] Preferably, in S4, the step of establishing the perturbation chain connectivity density value by extending the path based on the path turning correction angle value includes: S401: Based on the path turning correction angle value, select the disturbance point in the forest disturbance zone pixel as the starting point, call the gray value of the disturbance point and the neighboring pixels, calculate the gray value change rate of the difference direction, select the direction with the largest change rate among all directions, and generate the gray value change dominant direction value according to the angle change value corresponding to the largest direction. S402: Based on the dominant direction value of grayscale change, locate the extension direction of neighboring pixels, extract the grayscale change trend of pixels in multiple directions in sequence, compare the difference value between the grayscale change angle and the dominant direction between adjacent pixels, connect the pixels with the difference value less than the direction angle difference benchmark in sequence, and obtain the extension length value with consistent direction. S403: Based on the consistent extension length value, accumulate the number of consecutive pixels in the disturbance path, and average the ratio results of all path structures according to the ratio of the number of pixel connections in each path to the length of the corresponding path direction to obtain the disturbance chain connectivity density value.

[0012] Preferably, the expression for the grayscale change rate in the difference direction is: ; in, This represents the rate of change of grayscale value in the p-th direction. This represents the grayscale value of the perturbation point at the q-th position in the p-th direction. This represents the gray value of the perturbation point at the (q-1)th position in the p-th direction. This represents the difference between the angle between the q-th perturbation point in the p-th direction and the path direction correction angle value. This represents the angle value of the grayscale change direction of the q-th perturbation point in the p-th direction across three consecutive frames. represents the angle value of the grayscale change direction of the p-th perturbation point (q-1) in three consecutive frames, and n represents the number of perturbation points involved in the calculation in the p-th direction.

[0013] Preferably, in S500, the steps of calling the perturbation chain connectivity density value, calculating the grayscale trend difference and perturbation response amplitude at both ends of the chain, analyzing the spatial extension direction and structural integrity, extracting boundary pixels with continuous response characteristics, and generating structured image data units include: S501: Call the disturbance chain connectivity density value and the gray-scale sequence of the regions at both ends of the chain, calculate the average gradient value of the gray-scale points, and use the difference between the gray-scale fluctuation amplitude and the average gradient value as a benchmark, combined with the boundary disturbance frequency parameter, to obtain the gray-scale trend disturbance response amplitude value. S502: Based on the grayscale trend disturbance response amplitude value and the chain coordinate sequence, calculate the response amplitude difference range between adjacent coordinate points, extract the segment length of continuous change in direction offset, and obtain the structural integrity value in the spatial extension direction by combining the number of structural connection points. S503: Based on the structural integrity value and grayscale trend disturbance response amplitude value in the spatial extension direction, extract the grayscale change coefficient and disturbance frequency value of the boundary pixel at the corresponding position, filter the pixels whose ratio fluctuation amplitude is in the stable range, and generate structured image data units.

[0014] Preferably, a geographic information acquisition system based on remote sensing imagery includes: The image time series module is used to acquire remote sensing image sequences, perform time series registration on the remote sensing image sequences, extract pixel grayscale change features, and generate building edge area pixels. The edge extraction module is used to delineate the building edge region based on the building edge region pixels, identify disturbance points by combining the disturbance amplitude and gray-scale direction consistency, and generate the transition disturbance node distribution value. The disturbance identification module is used to extract the node positions in the road intersection area based on the distribution value of the transition disturbance nodes, and to obtain the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames. The path construction module is used to establish the perturbation chain connectivity density value by extending the path based on the path turning angle value; The boundary generation module is used to call the connectivity density value of the perturbation chain, calculate the difference in grayscale trend and perturbation response amplitude between the two ends of the chain, analyze the spatial extension direction and structural integrity, extract boundary pixels with continuous response characteristics, and generate structured image data units.

[0015] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a geographic information acquisition method and system based on remote sensing imagery. First, a remote sensing image sequence is acquired and temporally registered; pixel grayscale change features are extracted to generate building edge area pixels. Next, edge areas are delineated and disturbance points are identified. Then, road node positions are extracted and path turning angles are corrected. Next, disturbance chain connectivity density values ​​are established. Finally, structured image data units are generated. The present invention extracts pixel grayscale change features through temporal registration, locates building edge areas, identifies disturbance points by combining disturbance amplitude and grayscale direction consistency, and corrects path turning offsets, enhancing the stability of path connectivity structures. By extending the path to establish disturbance chain connectivity density values, and fusing grayscale trends and structural integrity to extract continuous response boundaries, image data units with grayscale trends, spatial connectivity, and disturbance features are generated, improving the accuracy and continuity of boundary identification in complex areas. Attached Figure Description

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

[0017] Figure 1 This is a schematic diagram of a geographic information acquisition method based on remote sensing imagery provided by the present invention.

[0018] Figure 2 This is a schematic diagram of a geographic information acquisition system based on remote sensing imagery provided by the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] This invention discloses a method for acquiring geographic information based on remote sensing imagery, such as... Figure 1 As shown, including: S100: Acquire a remote sensing image sequence, perform temporal registration on the remote sensing image sequence, extract pixel grayscale change features, and generate building edge area pixels; Specifically, by calculating the continuous grayscale variation characteristics between frames, analyzing the active response segments within the region, and combining the grayscale trend changes with the edge contour trend, the boundary concentration area is located, and building edge area pixels are generated. S200: Based on the building edge area pixels, delineate the building edge area, identify disturbance points by combining the disturbance amplitude and grayscale direction consistency, and generate the transition disturbance node distribution value; Specifically, based on the building edge area pixels, the building edge area is delineated, grayscale sequence is extracted within the sliding window, difference operation is performed on continuous pixels, pixel points whose difference exceeds the disturbance standard are extracted, and it is determined whether the grayscale change direction of the pixel points is consistent in adjacent frames. If they are consistent, they are marked as disturbance points, and the transition disturbance node distribution value is generated. S300: Based on the distribution value of the transition disturbance nodes, extract the node positions in the road intersection area, and obtain the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames. Specifically, based on the distribution value of the transition disturbance nodes, the node positions in the road intersection area are extracted, and the change in the turning angle of the position connection line is measured in three consecutive frames. If the turning angle of the position connection line shifts between frames, it is corrected based on the extension direction of the previous position to obtain the path turning correction angle value. S400: Based on the path turning correction angle value, establish the perturbation chain connectivity density value by extending the path; Specifically, based on the path turning correction angle value, the disturbed point in the forest disturbance zone pixel is selected as the starting point, and the consistency of the gray change direction of the neighboring pixels is judged. If the direction is consistent, it is included in the path structure, and the connection is extended point by point until it is interrupted, and the disturbance chain connectivity density value is established. S500: Call the connectivity density value of the disturbance chain, calculate the difference in grayscale trend and disturbance response amplitude between the two ends of the chain, analyze the spatial extension direction and structural integrity, extract boundary pixels with continuous response characteristics, and generate structured image data units.

[0021] Specifically, in S100, the steps of acquiring a remote sensing image sequence, performing temporal registration on the remote sensing image sequence, extracting pixel grayscale change features, and generating building edge area pixels include: S101: Obtain a remote sensing image sequence, rearrange the remote sensing image sequence in chronological order, extract the gray values ​​of pixels at the same position in consecutive image frames, determine the continuity of the pixel time series based on the coordinate changes and gray value differences between frames, exclude pixel positions where coordinate drift and gray value jump coexist, and generate a continuous pixel gray set. Specifically, acquiring remote sensing image sequences requires collecting multiple periods of remote sensing data on the target area within a continuous time period. For example, five periods of images might be continuously collected for a farmland renewal area, with a seven-day interval between each period. The images are then sorted by the shooting date to construct a time series. Subsequently, the grayscale value of the corresponding pixel in each frame is extracted for each spatial location. To ensure the consistent position of the same pixel in different images, registration processing is required first, aligning the same ground feature in each frame to a unified coordinate system. By comparing the coordinate changes before and after registration, drifting pixels are identified. When the same coordinate point is offset by more than the pixel width (e.g., 10 meters) in a frame, it is considered to have incompatible coordinates. The system is stable, and the grayscale value changes are analyzed. If the grayscale abrupt change exceeds the set threshold (e.g., 30), the pixel is judged to have both spatial drift and grayscale abrupt change and should be removed. For example, if a pixel has grayscale values ​​of 124, 127, 165, 130, and 133 in 5 frames, and the abrupt change in the third frame is far beyond the range of continuous change, and the coordinate displacement in that frame exceeds one pixel, then the pixel does not meet the continuity requirement and is removed from the overall data. The remaining pixels are included in the continuous sequence set, and finally a time series pixel set composed of stable position and smooth grayscale changes is established to generate a continuous pixel grayscale set.

[0022] S102: Based on the continuous pixel grayscale set, calculate the grayscale change rate of multiple pixels between adjacent frames, construct a regional grayscale fluctuation map, compare the neighborhood grayscale trend direction of the region where the grayscale change rate is concentrated, mark the corresponding grayscale space segment, and generate a trend response density value. Specifically, the expression for the grayscale fluctuation map of the region is: ; in, Representing the The pixel grayscale fluctuation trend value in a spatial segment Representing the In the spatial segment, the first The pixel grayscale value of the frame. Representing the In the spatial segment, the first The pixel grayscale value of the frame. Representing the Within a spatial segment The average value of the frame grayscale. This represents the number of time frames involved in the computation within the sequence. The summation symbol represents the frame index participating in the grayscale calculation of this spatial segment; the range of the summation symbol represents the range within the same segment. The process involves accumulating all time frames.

[0023] Specifically, after calculating the regional grayscale fluctuation map, the following steps are performed to compare neighborhoods of areas with concentrated grayscale change rates, mark segments, and generate trend response density values: Filtering high-fluctuation areas: Set a threshold and extract V from the grayscale fluctuation map of the area. j Spatial segments exceeding the threshold are included in the candidate pool; Neighborhood orientation consistency analysis: Taking the candidate segment as the center, a 3×3 neighborhood is selected, and the direction of grayscale change (increase / decrease) of each pixel in consecutive frames is calculated. If the orientation of ≥80% of the pixels in the neighborhood is consistent with that of the central segment (e.g., all are "increase"), then the region is determined to have trend consistency. Mark grayscale space segments: For segments that meet the conditions, record their number, coordinates, and V. j value; Calculate trend response density: count the number of spatial segments marked per unit area, according to V. j The trend response density is obtained by weighting the sum and then dividing by the total area.

[0024] This process locates high-response areas at the building edge by quantifying the consistency between fluctuation trends and neighborhood directions, providing a basis for subsequent boundary extraction.

[0025] In one specific embodiment of the present invention, the grayscale data originates from the pixel grayscale sequence at the building boundary in a remote sensing image of a suburban city, spanning three frames, corresponding to frame numbers 1, 2, and 3. The pixel grayscale value in each frame is directly extracted from the remote sensing image, with the pixel number P1 and the spatial segment number set to j=5. The grayscale values ​​of the collected pixel in the three frames are as follows: frame 1 grayscale value is 102, frame 2 grayscale value is 110, and frame 3 grayscale value is 107, corresponding to... , , .

[0026] mean The average gray value of this pixel across three frames is calculated according to the rules for calculating the average gray value of remote sensing images: ; Absolute value of the difference term: ; ; In the variance denominator calculation, the root mean square error is calculated based on the stability distribution of continuous gray-level variations in the remote sensing image. ; ; Calculation of correction factor in frame 2: ; Calculation of correction factor for frame 3: ; Weighted summation of grayscale changes: ; Frame number t=3, substituting into the formula, we get: ; The results show that the grayscale fluctuation trend value of spatial segment number 5 in three consecutive frames is 6.83. This value takes into account both the grayscale jump amplitude between adjacent frames and the grayscale dispersion of the pixel in the segment. This trend value will serve as an input factor for subsequent neighborhood direction consistency judgment and will have a direct impact on the recognition results of trend response dense areas.

[0027] S103: Determine the location of grayscale trend direction change based on the trend response density value, extract the direction turning point that forms the boundary structure, connect the turning points to form a boundary line, compare the distance between the boundary line and the pixel line corresponding to the existing contour in the original remote sensing image, and determine the degree of offset between the two by calculating the spatial distance between the two lines; based on the set reference value, locate the area with similar offset distance as the building edge area, and generate building edge area pixels.

[0028] Specifically, based on the trend response density value, the location where the grayscale trend direction changes is identified, and the location where the trend direction change is significant is extracted as the turning point. The directional change in the time series is tracked in these areas. If the direction of change is found to change from southeast to northeast in consecutive time frames, and there is also a consistent change trend in adjacent areas, it is determined that the area may constitute a boundary structure. These turning points are then connected to form a boundary line. The generated boundary line is then matched with the existing building outlines in the original remote sensing image. For example, building shadow lines and roof edge lines are extracted as reference lines. By calculating the spatial distance between the two lines, areas with small offsets are identified. If the offset is within the set reference value range (e.g., no more than 5 meters), it is considered a boundary segment with high fitting degree. Taking the edge of a cluster in a built-up area of ​​a city as an example, the maximum deviation of a newly identified boundary line segment from the original outline edge line within a total length of 30 meters is 3.8 meters, which is lower than the set reference value. Therefore, this segment can be confirmed as a building edge area, and building edge area pixels are generated.

[0029] Specifically, in S200, the steps of delineating the building edge region based on the building edge region pixels, identifying disturbance points by combining the disturbance amplitude and grayscale direction consistency, and generating the transition disturbance node distribution values ​​include: S201: Obtain the pixels of the building edge area, collect the gray value sequence of multiple rows of pixels, calculate the gray value difference between adjacent pixels, compare the gray value difference amplitude with the gray value disturbance standard value, extract the pixels whose difference amplitude exceeds the standard value, and generate gray value difference interval value. Specifically, the remote sensing image is first preprocessed to remove local abnormal pixel values ​​caused by lighting and noise interference, and the overall grayscale contrast is enhanced through linear stretching to make the grayscale difference between the building edge and the background area clearer. Then, multiple consecutive rows of pixels are extracted from the building edge area, and the grayscale value sequence is recorded in row and column order to form a two-dimensional grayscale matrix. For adjacent pixels in the matrix, the grayscale difference is calculated pairwise, using the difference formula with the current pixel's grayscale value and the grayscale values ​​of its horizontally adjacent pixels as the two participating terms. The grayscale perturbation standard value is set based on the stability of grayscale changes in the background area. In the analysis of 200 Gaofen-2 remote sensing images with a resolution of 5 meters, the grayscale difference fluctuation of background area pixels usually does not exceed 15, and its 95th percentile is close to 20. Therefore, the grayscale perturbation standard value is set at 20. For example, if a pixel has a grayscale value of 120 and its neighboring pixel has a grayscale value of 160, the grayscale difference between them is 40. This value is greater than the perturbation standard value of 20, indicating that the point is located in a high-variability area at the edge of a building and is an extractable perturbation point. Following this logic, the entire row of pixels is traversed, and all pixels with grayscale differences exceeding the standard value are recorded. These locations of significant grayscale abrupt changes are extracted, and finally, grayscale difference interval values ​​are generated.

[0030] Specifically, after preprocessing the remote sensing image, a two-dimensional grayscale matrix is ​​formed by extracting multiple consecutive rows of pixels in the building edge area. When traversing the entire row of pixels, starting from the first pixel, the grayscale difference between the current pixel and its horizontally adjacent pixels is calculated sequentially (e.g., the difference between pixel grayscale value 120 and its adjacent pixel 160 is 40). Each difference is compared with the standard value of 20. If it exceeds the standard value, it is determined as a significant abrupt change point, and its row and column coordinates and the magnitude of the difference are recorded (e.g., coordinates (10,20), difference 40). The matrix is ​​scanned row by row, and all abrupt change points that meet the conditions are stored sequentially, forming an interval record containing the start and end coordinates (e.g., the abrupt change interval from (10,20) to (10,25). Finally, all abrupt change intervals are summarized row by row to generate grayscale difference interval values, which are stored in a structured data format of "row number-start coordinate-end coordinate-maximum difference", for example, "row 5: (10,20)-(10,25), maximum difference 40", for subsequent judgment of the consistency of the direction of disturbance points.

[0031] S202: Based on the grayscale difference interval value, extract pixels at the same position in consecutive frames, determine whether the grayscale change direction of the pixels in adjacent frames is consistent, mark the pixel points that meet the direction consistency, and obtain the disturbance direction preservation value. Specifically, based on the grayscale difference interval values ​​extracted in the previous step, pixels at the same location are extracted from consecutive frames of remote sensing imagery, and their grayscale changes are analyzed to identify whether the direction of pixel change remains consistent. The extraction method is based on the corresponding row and column coordinates in each frame after image registration, ensuring positional consistency. Subsequently, the direction of grayscale change is determined by horizontally comparing the grayscale values ​​of pixels with the same coordinates in consecutive frames. The direction is defined as whether the grayscale value increases or decreases relative to the previous frame. For example, if a pixel has a grayscale value of 80 in the first frame and a grayscale value of 90 in the second frame, its grayscale direction is recorded as an increase; conversely, if the grayscale value in the second frame is 70, it is a decrease. To determine directional consistency, the directions of the first two frames and the last two frames in three consecutive frames need to be compared. If the directions are consecutively the same, the pixel is marked as a pixel with consistent direction. In determining directional consistency, a fluctuation tolerance needs to be set. In the experiment, considering the typical grayscale fluctuations in building areas, the error in judging the direction of grayscale changes between consecutive frames was set to no more than ±3 grayscale units. This setting was derived from point-by-point grayscale variation calculations of 150 images from typical urban areas, industrial areas, and bare land areas. It was found that over 92% of the stable points in continuous directional changes had fluctuation ranges below this value. All pixels meeting this condition were marked as directional preservation points, ultimately forming the perturbation directional preservation value.

[0032] Specifically, based on the grayscale difference range, pixels at the same registered position are extracted from consecutive frames, and the direction of change (increase or decrease) is determined by comparing the grayscale values ​​of adjacent frames. By comparing the directions of the first two and last two frames of three consecutive frames, if the directions are continuously identical and the grayscale fluctuation is within ±3 grayscale units (covering 92% of stable points based on statistics from 150 images), it is marked as a direction-preserving point. During marking, each pixel meeting the conditions is assigned a unique identifier (e.g., "DHP_coordinates"), and its coordinates, consecutive frame direction sequence (e.g., "+", "+"), and fluctuation amplitude are recorded, forming structured data containing position markers and consistency values. During numerical processing, a confidence value of 0.8-1 is assigned based on the fluctuation amplitude (the smaller the fluctuation, the higher the value). Finally, a perturbation direction-preserving value is constructed in the form of a "position-direction-confidence" triple, which both locates the pixel and provides a quantitative basis for subsequent perturbation analysis.

[0033] S203: Based on the disturbance direction preservation value, calibrate the location of disturbance points in consecutive frames of remote sensing image, construct a spatial distribution map of disturbance nodes, assign a distribution value corresponding to the location to each disturbance point, and generate the distribution value of transition disturbance nodes.

[0034] Specifically, based on the obtained perturbation direction preservation values, the next step is to pinpoint the locations of perturbation points in the remote sensing imagery. First, based on the perturbation point locations marked in the previous step, perturbation points at the same locations within consecutive frames are identified. Then, a spatial distribution map of the perturbation nodes is constructed using these locations. Each perturbation point corresponds to a spatial coordinate, representing its position in different frames. Next, the frequency of each perturbation point's occurrence across multiple frames is calculated, and these locations are assigned corresponding distribution values. For example, if a perturbation point appears in both frames, its distribution value is 2, indicating that the perturbation point has changed in both frames. Using this method, the final transition perturbation node distribution value is generated, representing the spatial distribution of perturbation points across consecutive frames and reflecting the frequency and location of their occurrence.

[0035] Specifically, in S300, the steps of extracting the node positions in the road intersection area based on the distribution values ​​of the transition disturbance nodes, and obtaining the path turning correction angle value by analyzing and correcting the angle changes of three consecutive image frames include: S301: Based on the distribution value of the transition disturbance nodes, extract the node position data in the road intersection area, match the positions of nodes with the same number in three consecutive frames according to the spatial sequence of the nodes in the image frame, organize the corresponding spatial change direction according to the connection order of the nodes in the three consecutive frame images and the time order between frames, and generate a node sequence direction set. Specifically, based on the distribution values ​​of transition disturbance nodes, it is necessary to identify the node location data in the road intersection area. The specific operation involves extracting image segments containing transition boundaries from the labeled disturbance areas in the remote sensing image sequence, identifying areas with significant pixel density changes within these segments, and determining the node locations. Each image frame is set as an observation window, corresponding to a time point within the actual monitoring period. For example, three adjacent image frames T1, T2, and T3 are extracted to represent three consecutive observation times. Next, for the disturbance nodes around the road intersection, their corresponding spatial coordinates in frames T1, T2, and T3 are read one by one, and the node numbers are matched with the image frame numbers to ensure data comparability for the same node at different time points. Subsequently, the positions of nodes across the three frames are recorded in a sequence according to their node numbers, forming a set of connection data. Following the chronological order of the image frames, each set of connection records is arranged sequentially to deduce the spatial movement direction of the nodes. For example, the position of node N12 in image coordinates is (128, 342) in T1, (135, 351) in T2, and (148, 366) in T3. Based on this, its overall offset direction in the image from the first to the third frame can be identified, and its movement trend can be inferred through the coordinate differences. All nodes must be matched and organized according to this process, ultimately summarizing into a set of directional information records for each node across the three frames, resulting in a node sequence direction set.

[0036] S302: Call the node sequence direction set, identify the direction change of each group of nodes, calculate the turning angle between the first two frames and the last two frames; according to the position change trend of nodes with the same number, set the angle reference value, filter out cases where the angle is offset, collect the angle change amplitude of each group of nodes, and obtain the turning angle offset value set. Specifically, after calling the node sequence direction set, the direction change of each group of nodes is identified. The core operation is to extract the change value of the movement direction of each group of nodes in two consecutive time intervals and archive the difference in direction as the turning angle. Taking a real-world scenario as an example, if a node moves northeast from the first frame to the second frame, and then moves east from the second frame to the third frame, then there is a change in direction, and the degree of this change is the offset angle between the two directions. To identify which nodes have significant direction changes, an angle benchmark value needs to be set for division. This benchmark value should refer to typical direction turning situations that appear in actual road movement trajectories. Generally, a fixed interval can be set based on the angle distribution results of road bend points in the geographic layer, such as choosing 30° as the criterion for judging a sudden change in node direction. Then, the direction change values ​​of each group of nodes are traversed, and their direction differences in two time intervals are compared. Nodes that exceed the benchmark value are filtered out. For example, if the direction change of node N5 is 42° and that of node N8 is 18°, then N5 is judged to have a direction shift, while N8 is not within the range of offset judgment. For all nodes with offsets, their direction change values ​​are further collected and summarized into a set of records, ultimately forming a set of steering angle offset values.

[0037] S303: Based on the set of turning angle offset values, select the connection direction of the previous frame in each group of nodes as a reference, adjust the position of the next frame in combination with the angle change range, calculate the position deviation angle before and after adjustment, and obtain the path turning correction angle value based on the correction value of the node. Specifically, based on the set of turning angle offset values, to further correct the path direction, the movement direction of each node in the first time period needs to be used as a reference standard to correct the positional relationship in the second time period. The operation first reads the movement direction data of each node from the first frame to the second frame, and then adjusts the position of the third frame according to the magnitude of the direction change. During the adjustment process, the movement trend of the previous frame should be kept unchanged, and the position offset of the subsequent frame should be reset to approximate the first segment direction, avoiding abrupt changes in direction. For example, if the original direction of a node was east-northeast from the first frame to the second frame, and north in the third frame, then the position of the third frame is slightly adjusted eastward according to the offset angle, resetting the direction to approximate the original path. The new adjusted direction is then compared with the original uncorrected direction to evaluate the angle difference of each node before and after the position adjustment. This difference represents the correction amount for that node; the larger the correction angle, the more significant the deviation of the original path. By statistically analyzing the adjustment range of each node and recording its numerical range (e.g., the correction angle for node N3 is 22° and for N6 it is 16°), all correction data are compiled and summarized to finally generate the path turning correction angle value.

[0038] Specifically, in S4, the step of establishing the perturbation chain connectivity density value by extending the path based on the path turning correction angle value includes: S401: Based on the path turning correction angle value, select the disturbance point in the forest disturbance zone pixel as the starting point, call the gray value of the disturbance point and the neighboring pixels, calculate the gray value change rate of the difference direction, select the direction with the largest change rate among all directions, and generate the gray value change dominant direction value according to the angle change value corresponding to the largest direction. Specifically, the disturbed points in the forest disturbance zone are selected as the starting point because the forest disturbance zone is a relatively sensitive area in the geographical environment, and its pixel grayscale changes are usually more significant. The grayscale change rate of the disturbed points is greater than that of ordinary ground features, which makes it easier to capture continuous grayscale change trends. Moreover, the disturbances in this area have spatial continuity, and the dominant direction of grayscale change of the disturbed points is more likely to form a long-distance consistent extension, which is conducive to improving the accuracy of the connectivity density value of the disturbance chain. At the same time, such disturbances are mostly caused by human or natural factors, which are real geographical changes. The directionality and continuity of grayscale changes are strong, which can reduce noise interference.

[0039] Specifically, the expression for the grayscale change rate in the difference direction is: ; in, This represents the rate of change of grayscale value in the p-th direction. This represents the grayscale value of the perturbation point at the q-th position in the p-th direction. This represents the gray value of the perturbation point at the (q-1)th position in the p-th direction. This represents the difference between the angle between the q-th perturbation point in the p-th direction and the path direction correction angle value. This represents the angle value of the grayscale change direction of the q-th perturbation point in the p-th direction across three consecutive frames. represents the angle value of the grayscale change direction of the p-th perturbation point (q-1) in three consecutive frames, and n represents the number of perturbation points involved in the calculation in the p-th direction.

[0040] In one specific embodiment of the present invention, firstly, the gray values ​​of two adjacent pixels are obtained. and These values ​​are obtained directly from remote sensing image data. For example, the gray values ​​of the q-th and q-1-th pixels in direction p are 150 and 145, respectively. Calculate the gray value difference between these two pixels: ; Next, adjust the angle value according to the path direction. and These two angles represent the angular values ​​of the grayscale change direction of a specific perturbation point in consecutive frames, and can be obtained from orientation data sensors or calculated. Assuming... and Calculate the absolute value of the difference between the two angles and substitute it into the formula: ; Next, calculate the denominator of the angle correction factor and substitute the above result into it: ; Now processing path direction correction values Assuming This is the angle correction factor, whose value is obtained from the orientation sensing data. It takes into account the effect of a 1-degree offset. ,but: ; Substitute all values ​​into the main formula to perform the calculation: ; If n=10, it means there are 10 such computational units in direction p. The summation is then divided by n to calculate... : ; The results indicate that the average gray-level change rate in direction p is 5.41. This value indicates that the gray-level change is most significant in this direction, which helps to identify the dominant direction of gray-level change and thus determine the key change trends in the image.

[0041] S402: Based on the dominant direction value of grayscale change, locate the extension direction of neighboring pixels, extract the grayscale change trend of pixels in multiple directions in sequence, compare the difference value between the grayscale change angle and the dominant direction between adjacent pixels, connect the pixels with the difference value less than the direction angle difference benchmark in sequence, and obtain the extension length value with consistent direction. Specifically, based on the dominant direction value of grayscale change, the direction angle information corresponding to the current perturbation point is called, such as the 45° direction. Starting from this direction, the process extends outward, extracting the grayscale values ​​of multiple pixels adjacent to the perturbation point in that direction, recording their grayscale change trends, and comparing the spatial change direction of grayscale changes of adjacent pixels with the dominant direction value. If the direction difference is within the allowable error range (e.g., not exceeding 10°), the pixel is retained as a path extension node. The path continues to extend in the original direction by sequentially connecting this node with the previous node. If the grayscale change direction of a pixel deviates too much from the dominant direction, the extension in that direction is terminated. During this process, the connection length of all pixels that meet the direction consistency condition is accumulated to form a continuous and consistent perturbation path in the direction. In a typical example, if one path extends continuously by 6 pixels in the dominant direction with a total length of 90 meters, and another path extends by 4 pixels with a total length of 60 meters, the direction consistency connection length of each path is recorded, and finally, the direction consistency extension length value is obtained.

[0042] S403: Based on the consistent extension length value, accumulate the number of consecutive pixels in the disturbance path, and average the ratio results of all path structures according to the ratio of the number of pixel connections in each path to the length of the corresponding path direction to obtain the disturbance chain connectivity density value.

[0043] Specifically, based on the consistent extension length value, the number of connected continuous cells on each disturbance path is called, and the numerical relationship between the number of continuous cells and the corresponding path length is compared. The ratio of the number of continuous cells to the path extension length in each disturbance path is calculated. In the same area, if path A contains 8 continuous cells and has a total length of 100 meters, and path B contains 5 continuous cells and has a total length of 60 meters, the ratio values ​​are 0.08 and 0.083 respectively. The ratio values ​​of all paths are averaged, and the connectivity of all disturbance paths in the area is recorded. The connection strength of the disturbance paths in the gray-scale dominant direction in the area is obtained, and finally the disturbance chain connectivity density value is generated.

[0044] Specifically, in S500, the steps of calling the perturbation chain connectivity density value, calculating the grayscale trend difference and perturbation response amplitude at both ends of the chain, analyzing the spatial extension direction and structural integrity, extracting boundary pixels with continuous response characteristics, and generating structured image data units include: S501: Call the disturbance chain connectivity density value and the gray-scale sequence of the regions at both ends of the chain, calculate the average gradient value of the gray-scale points, and use the difference between the gray-scale fluctuation amplitude and the average gradient value as a benchmark, combined with the boundary disturbance frequency parameter, to obtain the gray-scale trend disturbance response amplitude value. Specifically, by calling the connectivity density value of the perturbation chain and the grayscale sequence of the regions at both ends of the chain, in forest development boundaries or suburban expansion areas, multiple pixel points at the start and end points of the chain are first extracted, and their grayscale values ​​are obtained frame by frame to form a complete temporal grayscale sequence. Then, the continuous grayscale value change amplitude of each pixel point is calculated point by point to obtain the average rate of grayscale change of the point in multiple frames, and the amplitude difference of its grayscale fluctuation range is statistically analyzed. Then, by comparing the degree of difference between the average rate of change and the grayscale fluctuation difference, combined with the set grayscale perturbation benchmark, the region with a difference value higher than the benchmark is extracted as a sensitive perturbation segment. On this basis, the frequency of change of pixels in the boundary region in multiple frames is statistically analyzed to identify the number of pixels with large grayscale fluctuations in multiple time frames, and the ratio is formed with the total number of pixels participating in the statistics as the input of the perturbation frequency parameter. Taking into account the degree of grayscale difference and the frequency of perturbation, the grayscale trend perturbation response amplitude value is obtained by weighted comparison.

[0045] S502: Based on the grayscale trend disturbance response amplitude value and the chain coordinate sequence, calculate the response amplitude difference range between adjacent coordinate points, extract the segment length of continuous change in direction offset, and obtain the structural integrity value in the spatial extension direction by combining the number of structural connection points. Specifically, based on the grayscale trend disturbance response amplitude value and the chain coordinate sequence, in the rural road intersection area or artificial intervention diffusion area, the grayscale response values ​​corresponding to each coordinate point inside the chain are extracted sequentially. The grayscale response difference between any two adjacent points is extracted, and the response difference of all adjacent points in the entire chain is analyzed point by point to determine which consecutive position points have differences within a set stable change range. The length of the consecutive points that meet the conditions is recorded as the length of the continuous segment with spatial direction change. Then, the number of effective connection points in each continuous segment is counted, that is, the number of pixels with consistent grayscale response direction and change amplitude exceeding the disturbance reference value in the segment, and a corresponding relationship is established with the total number of pixels in the segment to represent the structural connection integrity. Using the degree of structural connection as the main quantity, combined with the length of the continuous segment to form a joint parameter, the structural extension degree of each segment in space is finally determined, and the structural integrity value in the spatial extension direction is obtained.

[0046] S503: Based on the structural integrity value and grayscale trend disturbance response amplitude value in the spatial extension direction, extract the grayscale change coefficient and disturbance frequency value of the boundary pixel at the corresponding position, screen the pixels whose ratio fluctuation amplitude is in a stable range, and generate structured image data units. Specifically, based on the structural integrity value and grayscale trend disturbance response amplitude value in the spatial extension direction, boundary pixel points under corresponding coordinates are extracted in the urban building edge zone or urban renewal fault zone area. The grayscale change range of each point in continuous time frames is counted, and the corresponding grayscale disturbance frequency is recorded, that is, the number of jumps in the continuous image. The grayscale change amplitude of a single point is compared with the disturbance frequency, and the ratio between the change frequency and amplitude is calculated. Pixel points with a stable ratio within the preset stable range are selected according to the preset stable range value, and pixels with abrupt changes or too low response frequency are excluded. Finally, all the selected pixel points are uniformly encapsulated according to their coordinate position, grayscale change value, response amplitude, disturbance frequency and structural extension strength to form standard data units and generate structured image data units.

[0047] In a specific embodiment of the present invention, the building edge area pixel includes the boundary concentration area number, grayscale trend direction, and contour gradient value; Specifically, the location of grayscale trend direction change is determined by calculating the trend response density value, the direction turning point is extracted to form the boundary line, the boundary line is compared with the original image outline, and the area with the offset distance close to the reference value is located. Each area that meets the conditions is assigned a unique number to obtain the boundary concentration area number. The boundary concentration area number is used to distinguish different building edge areas, which facilitates the subsequent management, tracking and analysis of data in specific areas (such as accurately locating specific areas when identifying disturbance points).

[0048] The grayscale change rate is calculated by the regional grayscale fluctuation map, and the trend direction is marked by the comparison of the neighborhood direction. The specific directional change position is further determined according to the trend response density to obtain the grayscale trend direction. The grayscale trend direction reflects the change trend of pixel grayscale in the time series (such as from southeast to northeast), providing a basis for judging the spatial orientation of the building edge.

[0049] By calculating the distance difference between the boundary line and the original image outline, the steepness of the edge is quantified by pixel-level distance changes to obtain the outline gradient value. The outline gradient value is used to measure the clarity and gradient change of the building edge, and to help judge the authenticity and stability of the edge (the higher the gradient value, the more obvious the edge).

[0050] The distribution values ​​of transition disturbance nodes include the sequence of disturbance node locations, grayscale change direction indicators, and disturbance amplitude classification labels; Specifically, by extracting the grayscale sequence of pixels in the building edge area, calculating the grayscale difference between adjacent pixels, filtering pixels whose difference exceeds the standard value, marking the location of disturbance points based on directional consistency, and arranging them in chronological order to form a coordinate sequence, a sequence of disturbance node locations is obtained; this is used to record the spatial movement trajectory of disturbance points in the time dimension, and to analyze the dynamic change pattern of disturbances.

[0051] By determining the direction of grayscale change (increase / decrease) of pixels in consecutive frames, if the direction is consistent in adjacent frames (such as continuous increase), it is marked as "+" or "-" (with an error of no more than ±3 grayscale units), thus obtaining the grayscale change direction identifier. This identifier is used to mark the grayscale change trend of disturbance points, helping to identify real disturbances (such as continuous changes at the edge of a building) and noise interference.

[0052] By comparing the grayscale difference with the standard disturbance value (e.g., 20), and classifying the disturbance amplitude according to the size of the difference (e.g., difference > 30 is "high", 15-30 is "medium"), a disturbance amplitude classification label is obtained. This label is used to quantify the intensity of the disturbance and facilitates prioritizing the treatment of high-amplitude disturbance areas (e.g., significant change points at the edge of buildings).

[0053] The path steering correction angle value includes the starting path direction identifier, the angle change value of consecutive frames, and the angle offset correction value; Specifically, by extracting the positions of road intersection nodes in consecutive frames and connecting them in chronological order, the direction of the connection from the first frame to the second frame is calculated using coordinate differences to obtain the starting path direction identifier. The starting path direction identifier serves as the reference direction for path turning and is used to determine the angle offset of subsequent frames.

[0054] By calculating the turning angle of the line connecting the first two frames and the last two frames, and using the vector angle formula to calculate the angle difference between the two connecting lines, the angle change value of the continuous frames is obtained. The angle change value of the continuous frames is used to reflect the turning magnitude of the path in the time series and to identify abnormal offsets (such as offsets exceeding the baseline value of 30°).

[0055] By using the direction of the previous frame as a reference, the position of the next frame is adjusted, and the angle deviation before and after the adjustment is calculated to obtain the angle offset correction value. The angle offset correction value is used to correct the offset of the path turning to ensure the accuracy of path tracking (such as avoiding abrupt changes in road node connections due to noise).

[0056] The perturbation chain connectivity density value includes path extension continuity, node connection frequency, and chain structure cell number; Specifically, by extending the path along the dominant direction of grayscale change, the ratio of the number of continuous pixels to the actual length of the path is counted to obtain the path extension continuity. The path extension continuity is used to measure the continuity of the path in space. The higher the continuity, the better the connectivity of the disturbance chain (such as the continuous change area of ​​the forest disturbance zone).

[0057] The node connection frequency is obtained by counting the number of connections between the disturbed nodes in each path. The node connection frequency reflects the correlation strength between the disturbed nodes. The higher the frequency, the stronger the transitivity and stability of the disturbance.

[0058] By recording the unique cell number from the starting point in an extended sequence, the chain structure cell number is obtained; the chain structure cell number is used to identify the spatial structure of the perturbation chain, making it easy to trace the position and connection relationship of each cell in the chain.

[0059] The structured image data unit includes boundary response pixel number, grayscale trend difference value, and structural integrity judgment value.

[0060] Specifically, by screening pixels with stable grayscale variation coefficients and perturbation frequency ratios, their numbers are recorded (screened from chain structure pixels) to obtain boundary response pixel numbers. Boundary response pixel numbers are used to locate true boundary pixels and eliminate noise interference (such as edge points with stable responses in urban renewal areas).

[0061] By calculating the average gradient value of the grayscale sequences at both ends of the chain, and using the difference between the grayscale fluctuation amplitude and the average gradient as the difference value, the grayscale trend difference value is obtained. The grayscale trend difference value is used to reflect the difference in grayscale changes at both ends of the chain and to help judge the integrity of the boundary (the smaller the difference value, the more continuous the boundary).

[0062] By analyzing the difference in response amplitude between adjacent coordinate points, the length of continuous segments is extracted, and the structural integrity judgment value is obtained by combining the number of connection points. The structural integrity judgment value is used to quantify the spatial structural stability of the boundary. The higher the judgment value, the better the continuity and integrity of the boundary (such as the complete boundary of the edge of urban buildings).

[0063] In one specific embodiment of the present invention, a geographic information acquisition system based on remote sensing imagery, such as... Figure 2 As shown, including: The image time series module is used to acquire remote sensing image sequences, perform time series registration on the remote sensing image sequences, extract pixel grayscale change features, and generate building edge area pixels. The edge extraction module is used to delineate the building edge region based on the building edge region pixels, identify disturbance points by combining the disturbance amplitude and gray-scale direction consistency, and generate the transition disturbance node distribution value. The disturbance identification module is used to extract the node positions in the road intersection area based on the distribution value of the transition disturbance nodes, and to obtain the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames. The path construction module is used to establish the perturbation chain connectivity density value by extending the path based on the path turning angle value; The boundary generation module is used to call the connectivity density value of the perturbation chain, calculate the difference in grayscale trend and perturbation response amplitude between the two ends of the chain, analyze the spatial extension direction and structural integrity, extract boundary pixels with continuous response characteristics, and generate structured image data units.

[0064] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0065] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for acquiring geographic information based on remote sensing imagery, characterized in that, include: S100: Acquire a remote sensing image sequence, perform temporal registration on the remote sensing image sequence, extract pixel grayscale change features, and generate building edge area pixels; S200: Based on the building edge area pixels, delineate the building edge area, identify disturbance points by combining the disturbance amplitude and grayscale direction consistency, and generate the transition disturbance node distribution value; S300: Based on the distribution value of the transition disturbance nodes, extract the node positions in the road intersection area, and obtain the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames. S400: Based on the path turning correction angle value, establish the perturbation chain connectivity density value by extending the path; S500: Call the connectivity density value of the disturbance chain, calculate the difference in grayscale trend and disturbance response amplitude between the two ends of the chain, analyze the spatial extension direction and structural integrity, extract boundary pixels with continuous response characteristics, and generate structured image data units.

2. The geographic information acquisition method based on remote sensing imagery according to claim 1, characterized in that, In step S100, the steps of acquiring a remote sensing image sequence, performing temporal registration on the remote sensing image sequence, extracting pixel grayscale change features, and generating building edge area pixels include: S101: Obtain a remote sensing image sequence, rearrange the remote sensing image sequence in chronological order, extract the gray values ​​of pixels at the same location in consecutive image frames, determine the continuity of the pixel time series based on the coordinate changes and gray value differences between frames, and generate a continuous pixel gray set. S102: Based on the continuous pixel grayscale set, calculate the grayscale change rate of multiple pixels between adjacent frames, construct a regional grayscale fluctuation map, compare the neighborhood grayscale trend direction of the region where the grayscale change rate is concentrated, mark the corresponding grayscale space segment, and generate a trend response density value. S103: Determine the location of grayscale trend direction change based on the trend response density value, extract the direction turning point to form the boundary structure, form the boundary line, compare the distance between the boundary line and the pixel line corresponding to the existing contour in the original remote sensing image, and based on the set reference value, locate the area with a close offset distance as the building edge area, and generate building edge area pixels.

3. The geographic information acquisition method based on remote sensing imagery according to claim 2, characterized in that, The expression for the grayscale fluctuation map of the region is: ； in, Representing the The pixel grayscale fluctuation trend value in a spatial segment Representing the In the spatial segment, the first The pixel grayscale value of the frame. Representing the In the spatial segment, the first The pixel grayscale value of the frame. Representing the Within a spatial segment The average value of the frame grayscale. This represents the number of time frames involved in the computation within the sequence. This represents the frame index that participates in the grayscale calculation of this spatial segment.

4. The geographic information acquisition method based on remote sensing imagery according to claim 1, characterized in that, In S200, the steps of delineating the building edge region based on the building edge region pixels, identifying disturbance points by combining the disturbance amplitude and gray-scale direction consistency, and generating the transition disturbance node distribution values ​​include: S201: Obtain the pixels of the building edge area, collect the gray value sequence of multiple rows of pixels, calculate the gray value difference between adjacent pixels, compare the gray value difference amplitude with the gray value disturbance standard value, extract the pixels whose difference amplitude exceeds the standard value, and generate gray value difference interval value. S202: Based on the grayscale difference interval value, extract pixels at the same position in consecutive frames, determine whether the grayscale change direction of the pixels in adjacent frames is consistent, mark the pixel points that meet the direction consistency, and obtain the disturbance direction preservation value. S203: Based on the disturbance direction preservation value, calibrate the location of disturbance points in consecutive frames of remote sensing image, construct a spatial distribution map of disturbance nodes, assign a distribution value corresponding to the location to each disturbance point, and generate the distribution value of transition disturbance nodes.

5. The geographic information acquisition method based on remote sensing imagery according to claim 1, characterized in that, In S300, the steps of extracting the node positions in the road intersection area based on the distribution values ​​of the transition disturbance nodes, and obtaining the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames include: S301: Based on the distribution value of the transition disturbance nodes, extract the node position data in the road intersection area, match the positions of nodes with the same number in consecutive frames according to the spatial sequence of nodes in the image frame, organize the corresponding spatial change direction according to the connection order of nodes in consecutive frame images and the time order between frames, and generate a node sequence direction set. S302: Call the node sequence direction set, identify the direction change of each group of nodes, calculate the turning angle between the first two frames and the last two frames; according to the position change trend of nodes with the same number, set the angle reference value, filter out cases where the angle is offset, collect the angle change amplitude of each group of nodes, and obtain the turning angle offset value set. S303: Based on the set of turning angle offset values, select the connection direction of the previous frame in each group of nodes as a reference, adjust the position of the next frame in combination with the angle change range, calculate the position deviation angle before and after adjustment, and obtain the path turning correction angle value based on the correction value of the node.

6. The geographic information acquisition method based on remote sensing imagery according to claim 1, characterized in that, In S4, the step of establishing the perturbation chain connectivity density value by extending the path based on the path turning correction angle value includes: S401: Based on the path turning correction angle value, select the disturbance point in the forest disturbance zone pixel as the starting point, call the gray value of the disturbance point and the neighboring pixels, calculate the gray value change rate of the difference direction, select the direction with the largest change rate among all directions, and generate the gray value change dominant direction value according to the angle change value corresponding to the largest direction. S402: Based on the dominant direction value of grayscale change, locate the extension direction of neighboring pixels, extract the grayscale change trend of pixels in multiple directions in sequence, compare the difference value between the grayscale change angle and the dominant direction between adjacent pixels, connect the pixels with the difference value less than the direction angle difference benchmark in sequence, and obtain the extension length value with consistent direction. S403: Based on the consistent extension length value, accumulate the number of consecutive pixels in the disturbance path, and average the ratio results of all path structures according to the ratio of the number of pixel connections in each path to the length of the corresponding path direction to obtain the disturbance chain connectivity density value.

7. A geographic information acquisition method based on remote sensing imagery according to claim 6, characterized in that, The expression for the grayscale change rate in the difference direction is: ； in, This represents the rate of change of grayscale value in the p-th direction. This represents the grayscale value of the perturbation point at the q-th position in the p-th direction. This represents the gray value of the perturbation point at the (q-1)th position in the p-th direction. This represents the difference between the angle between the q-th perturbation point in the p-th direction and the path direction correction angle value. This represents the angle value of the grayscale change direction of the q-th perturbation point in the p-th direction across three consecutive frames. represents the angle value of the grayscale change direction of the p-th perturbation point (q-1) in three consecutive frames, and n represents the number of perturbation points involved in the calculation in the p-th direction.

8. The geographic information acquisition method based on remote sensing imagery according to claim 1, characterized in that, In S500, the steps of calling the perturbation chain connectivity density value, calculating the grayscale trend difference and perturbation response amplitude of the regions at both ends of the chain, analyzing the spatial extension direction and structural integrity, extracting boundary pixels with continuous response characteristics, and generating structured image data units include: S501: Call the disturbance chain connectivity density value and the gray-scale sequence of the regions at both ends of the chain, calculate the average gradient value of the gray-scale points, and use the difference between the gray-scale fluctuation amplitude and the average gradient value as a benchmark, combined with the boundary disturbance frequency parameter, to obtain the gray-scale trend disturbance response amplitude value. S502: Based on the grayscale trend disturbance response amplitude value and the chain coordinate sequence, calculate the response amplitude difference range between adjacent coordinate points, extract the segment length of continuous change in direction offset, and obtain the structural integrity value in the spatial extension direction by combining the number of structural connection points. S503: Based on the structural integrity value and grayscale trend disturbance response amplitude value in the spatial extension direction, extract the grayscale change coefficient and disturbance frequency value of the boundary pixel at the corresponding position, filter the pixels whose ratio fluctuation amplitude is in the stable range, and generate structured image data units.

9. A geographic information acquisition system based on remote sensing imagery, employing the geographic information acquisition method based on remote sensing imagery as described in any one of claims 1-8, characterized in that, include: The image time series module is used to acquire remote sensing image sequences, perform time series registration on the remote sensing image sequences, extract pixel grayscale change features, and generate building edge area pixels. The edge extraction module is used to delineate the building edge region based on the building edge region pixels, identify disturbance points by combining the disturbance amplitude and gray-scale direction consistency, and generate the transition disturbance node distribution value. The disturbance identification module is used to extract the node positions in the road intersection area based on the distribution value of the transition disturbance nodes, and to obtain the path turning correction angle value by analyzing and correcting the angle changes of continuous image frames. The path construction module is used to establish the perturbation chain connectivity density value by extending the path based on the path turning angle value; The boundary generation module is used to call the connectivity density value of the perturbation chain, calculate the difference in grayscale trend and perturbation response amplitude between the two ends of the chain, analyze the spatial extension direction and structural integrity, extract boundary pixels with continuous response characteristics, and generate structured image data units.

Images