A road distortion correction method based on DEM interactive editing and local orthographic updating
By constructing a road target elevation surface and performing local DEM correction and orthophoto updates, the geometric correspondence problem in the road distortion area was solved, achieving spatial correspondence consistency between the road and surrounding features and efficient surveying and mapping production.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHENGDU TECH UNIV
- Filing Date
- 2026-04-27
- Publication Date
- 2026-07-10
Smart Images
Figure CN122130046B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of remote sensing image processing, digital photogrammetry and mapping production technology, and in particular to a rapid correction method for road distortion in orthophoto production through interactive editing of local digital elevation models (DEMs) and driving local orthophoto updates. Background Technology
[0002] Digital orthophotos (DOM) are an important output in land surveys, resource monitoring, national geographic censuses, and basic surveying and mapping production. Current DOM production processes typically involve orthorectification based on the original image, image positioning parameters, and a reference DEM. Essentially, this involves using the ground elevation provided by the DEM to participate in image point backprojection and ground point intersection calculations to establish the correspondence between image pixels and ground locations.
[0003] When urban expansion, new or expanded roads, bridge renovations, and site reshaping cause changes in the landform, time-related discrepancies will appear between the existing DEM and the actual land surface corresponding to the current image. This is especially true for linear or strip-shaped areas such as roads, ramps, connecting roads, and site edges. If the DEM elevation does not match the actual ground level, the projection correction of the same road feature at different locations during orthorectification will show inconsistent changes, which will then manifest in the DOM as road distortion, jagged edges, misaligned boundaries, local blurring, abnormal width, or discontinuity.
[0004] In current surveying and mapping production, a common method for manual retouching is to crop local segments from the original image and directly paste them into the DOM. While this method can visually cover anomalous areas, it does not change the DEM upon which orthorectification relies, nor does it re-establish the correct geometric correspondence between pixels within the anomalous area and ground targets. Therefore, the patched road section may still have relative inconsistencies with surrounding buildings, green spaces, plot boundaries, and other features. Furthermore, due to differences in observation angles, projection differences, radiometric characteristics, and resampling relationships between different image segments, it is easy to further generate edge-joining marks, texture jumps, tone differences, and secondary misalignment problems.
[0005] Therefore, using methods such as local updates, flattening, or smoothing in isolation can easily be misinterpreted as a simple patchwork of routine processing steps. There is an urgent need for a technical solution that collaboratively organizes the entire process—"road target elevation surface construction—local DEM gradual correction—local orthorectification recalculation—real-time result verification"—into a closed-loop workflow. This solution aims to restore the correct spatial correspondence between the road area and surrounding features at the three-dimensional geometric level, while simultaneously considering processing accuracy, boundary continuity, and surveying production efficiency. Summary of the Invention
[0006] The present invention provides a road distortion correction method based on DEM interactive editing and local orthophoto update, in order to solve the problems of failure to restore geometric correspondence, poor boundary connection, low efficiency and difficulty in large-scale reuse in surveying and mapping production when using the original image patch method to process road distortion in the prior art.
[0007] To solve the above problems, the technical solution adopted by the present invention is as follows:
[0008] This invention provides a road distortion correction method based on DEM interactive editing and local orthophoto update, comprising the following steps:
[0009] S1, construct the job project, and import the original images, initial DOM, DEM and image positioning parameters of the same job area;
[0010] S2, identify the road distortion region in the initial DOM, delineate the region to be processed, and expand outward along the region to be processed to form a local update window and edge transition zone;
[0011] S3. Select stable ground sample points around the area to be processed, and construct the target elevation surface of the road based on the elevation of the stable ground sample points, combined with the road lateral smoothness constraint and the road longitudinal continuity constraint.
[0012] S4. Based on the road target elevation surface, the DEM in the area to be processed is corrected to obtain a locally corrected DEM; wherein, a gradient method is used in the edge transition zone to make the locally corrected DEM continuously connected with the uncorrected DEM.
[0013] S5, based on the local modified DEM, the original image and the image positioning parameters, perform orthorectification and resampling only on the pixels within the local update window to generate a local orthorectification result;
[0014] S6, Generate an update mask based on the edge transition zone, and write the local orthophoto result back to the initial DOM based on the update mask to obtain the updated DOM;
[0015] S7. The updated DOM is checked based on the misalignment between the road boundary and adjacent features, the continuity of the road centerline, and the consistency of the road width. If the check result does not meet the preset requirements, the process returns to step S3 to reconstruct the road target elevation surface and executes subsequent steps to continue to correct the DEM in the area to be processed until the preset requirements are met.
[0016] S8. When the inspection results meet the preset requirements, save the local modified DEM, the local update record, and the updated DOM.
[0017] Furthermore, in step S2, the method for delineating the area to be processed includes: directly delineating the road distortion area on the initial DOM; or, delineating the road distortion area after assisting in locating it by combining road vector data, existing ground feature results, or road centerline.
[0018] Based on the defined road distortion area, a buffer zone of a preset width is extended outward to form the local update window and edge transition zone used for DEM correction and orthophoto inverse kinematics calculation.
[0019] Further, in step S3, the stable ground sample points are selected from the ground on both sides of the road, the normal road sections connecting the front and rear of the road, or the stable areas around the area to be processed that have not undergone deformation; the method for constructing the road target elevation surface includes any one of the following: average value calculation, weighted average calculation, local plane fitting, piecewise linear fitting, and surface fitting.
[0020] Further, in step S3, the road target elevation surface function within the area to be processed is denoted as... , , Let the coordinates be planar coordinates; the stable ground sample points are denoted as... ; This represents the total number of stable ground sample points; , For the first The planar coordinates of a stable ground sample point For the first The elevation values of a stable ground sample point; then the target road elevation surface is determined by solving the following objective function:
[0021]
[0022] In the formula, J is the objective function value. The fitting error term for the road target elevation surface to the stable ground sample points is denoted as . This is a lateral smoothness constraint for the road. This refers to the longitudinal continuity constraint term for the road. The elevation surface function of the road target in plane coordinates Elevation value at the location; For the first The weights corresponding to each stable ground sample point The constraint weighting coefficients for the lateral smoothness constraints of the road; The constraint weight coefficient is the longitudinal continuity constraint of the road.
[0023] Furthermore, in step S4, the operation of correcting the DEM in the area to be processed includes one or more combinations of flattening, interpolation, local stretching, local pressure drop, and smooth transition.
[0024] Let the original DEM be The corrected local modified DEM is as follows: The core correction area within the region to be processed is The edge transition zone is The locally corrected DEM is then calculated using the following formula:
[0025]
[0026] In the formula, The elevation surface of the road target within the area to be processed. The gradual weighting function is derived from the point... The normalized distance to the outer boundary of the edge transition zone is determined.
[0027] Further, in step S6, the calculation formula for writing the local orthophoto result back to the initial DOM is:
[0028]
[0029] In the formula, , For planar coordinates, For the initial DOM, This is a local orthophoto result. For the updated DOM; To update the mask, it is defined as:
[0030]
[0031] in, As the core correction area, It is an edge transition zone. This is a gradual weighting function.
[0032] Furthermore, in step S5, the pixels of the initial DOM outside the local update window remain unchanged to achieve local fast updates.
[0033] Further, in step S7, the misalignment index is calculated. Continuity indicators Width Consistency Index The formulas are as follows:
[0034]
[0035]
[0036]
[0037] when , and If the result meets the preset requirements, the system determines that the inspection result does not meet the preset requirements and continues iterating.
[0038] In the formula, For the first Discrete points along the road centerline; For the first Discrete points along the road centerline; For the first One corrected road boundary sampling point, For the first Corresponding points of nearby reference features This represents the total number of boundary sampling points; For the first Discrete points along the road centerline This represents the total number of discrete points along the road centerline. The width value is obtained by sampling along the road. K represents the average width, and K represents the total number of width sampling points. , and All are preset thresholds.
[0039] Furthermore, the image positioning parameters include one or more combinations of rational polynomial coefficients (RPC), exterior orientation elements, or correction parameters obtained from regional network adjustment.
[0040] Compared with the prior art, the beneficial effects of the present invention are:
[0041] (1) It abandons the traditional patch repair method that directly covers anomalies at the image level. By constructing the road target elevation surface and driving local orthorectification recalculation, it restores the correct spatial correspondence between the road area pixels and the real ground target from the orthorectification geometry source. It effectively solves the problems of inconsistent relative positions of roads and surrounding features, texture jumps and secondary misalignment that are easily caused by traditional patching methods.
[0042] (2) By jointly solving the objective function that includes the fitting error term of the surrounding stable ground sample points and the road lateral smoothness and road longitudinal continuity constraint term, the target elevation surface of the road is constructed. This avoids the deformation caused by empirical simple leveling, so that the target elevation surface reflects the current real ground state and avoids the sudden change of the elevation surface.
[0043] (3) The “edge transition zone gradient correction” and “update mask weighted rewrite” are organized into an integrated process. The target elevation surface is directly used in the core area, and the gradient weight function of normalized distance is used in the edge transition zone to perform dual gradient fusion of elevation and image. This can effectively eliminate road distortion and scraggly patterns, and greatly reduce the risk of boundary elevation jumps and image seams.
[0044] (4) By adopting a local recalculation strategy, only the pixels within the local update window are resampled by orthophoto recalculation, while the DOM outside the local window remains unchanged. This avoids the repeated production of the entire scene image while ensuring geometric consistency, significantly shortens the processing time, and improves the efficiency of surveying and mapping production.
[0045] (5) Quantitative evaluation indicators such as boundary misalignment, centerline continuity and road width consistency are introduced, which upgrades the traditional empirical visual judgment to a quality closed-loop judgment with configurable thresholds, forming an intelligent iterative closed loop of "correction-recalculation-verification", which greatly improves the quality and controllability of the map.
[0046] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, embodiments of the present invention are described below in detail with reference to the accompanying drawings. Attached Figure Description
[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0048] Figure 1 This is a flowchart of the method of the present invention;
[0049] Figure 2 The diagram shows the process of local road distortion correction on Gaofen-2 image using the method of the present invention. (a) is the DOM before correction, (b) is a schematic diagram of local DEM correction based on the road target elevation surface, and (c) is the result after local orthorectification update.
[0050] Figure 3 The diagram shows the process of using the method of the present invention to perform local road distortion correction on a No. 2 image of a certain area. (a) is the DOM before correction, (b) is a schematic diagram of local DEM correction based on the road target elevation surface, and (c) is the result diagram after local orthorectification update.
[0051] Figure 4 This study compares the results of local road distortion correction before and after two sets of cases: one from Gaofen-2 imagery and the other from Gaofen-2 imagery of a certain location. Detailed Implementation
[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, but not all embodiments.
[0053] When the digital elevation model (DEM) is inconsistent with the actual road elevation, the road area will exhibit phenomena such as bending, streaking, boundary misalignment, and width anomalies in the initial digital orthophoto (DOM). These anomalies are not simply texture issues, but rather result from inconsistent geometric corrections caused by the involvement of elevation in projection calculations during orthorectification. This invention, based on the fundamental idea of modifying the DEM rather than pasting the image, provides a road distortion correction method based on interactive DEM editing and local orthorectification updates.
[0054] like Figure 1 As shown, this embodiment of the invention provides a road distortion correction method based on DEM interactive editing and local orthophoto updating, specifically including the following steps S1 to S8:
[0055] S1, Build the job project and import the raw data.
[0056] The project involves constructing a work plan, importing the original imagery, initial digital orthophoto (DOM), digital elevation model (DEM), and image positioning parameters for the same work area, and establishing the relationships between image layers, DOM layers, DEM layers, update mask layers, and editing log layers. The image positioning parameters include one or more combinations of rational polynomial coefficients (RPC), exterior orientation elements, or correction parameters obtained from regional network adjustment. If necessary, road vectors, existing traffic line results, or other auxiliary data can also be imported for reference.
[0057] S2 identifies the road distortion region in the initial DOM, delineates the region to be processed, and expands outward along the region to be processed to form a local update window and edge transition zone.
[0058] Workers browse and identify areas of road distortion (such as bends, stripes, misaligned boundaries, local jumps, or abnormal widths) in a multi-layer linked display interface (especially the initial DOM layer).
[0059] Methods for delineating the area to be processed include: directly delineating the road distortion area on the initial DOM; or, combining road vector data, existing ground features, or road centerlines to assist in locating the road distortion area before delineation.
[0060] To avoid unnatural stitching caused by the update boundary being too close to the abnormal area, a buffer zone of preset width is extended outward from the defined road distortion area to form a local update window for subsequent DEM correction and local orthophoto inverse kinematics calculation, and an edge transition zone for gradual connection is set in it.
[0061] S3, Select stable ground sample points and construct the road target elevation surface.
[0062] Switch to the DEM layer and select stable ground sample points around the area to be processed. Specifically, stable ground sample points are selected from the ground on both sides of the road, normal road sections connecting the front and rear of the road, or stable areas around the area to be processed that have not undergone deformation.
[0063] like Figure 2 (b) and Figure 3 As shown in (b), the target elevation surface of the road is then constructed based on the elevation of the stable ground sample points, combined with the road's lateral smoothness constraints and longitudinal continuity constraints. A gradual transition zone is set at the edge, i.e., the aforementioned edge transition zone, to avoid simply interpreting it as empirical leveling. This ensures that the locally corrected DEM can reflect the current road conditions and naturally connect with the surrounding uncorrected areas. When the road is relatively smooth overall, the target elevation can be obtained by calculating the average value or weighted average. When the road has gentle slopes or longitudinal slopes, the target elevation surface can be constructed using any of the following methods: local plane fitting, piecewise linear fitting, and surface fitting.
[0064] Specifically, the road target elevation surface function within the area to be processed is denoted as... , , Let be the plane coordinates; denote the stable ground sample points as . ; To stabilize the total number of ground sample points; , For the first The planar coordinates of a stable ground sample point For the first The elevation values of a stable ground sample point are used to determine the target road elevation surface by solving the following objective function:
[0065]
[0066] In the formula, J is the objective function value. This is the fitting error term for the road target elevation to stable ground sample points. This is a lateral smoothness constraint for the road. This refers to the longitudinal continuity constraint term for the road. The elevation surface function of the road target in plane coordinates Elevation value at the location; For the first The weights corresponding to each stable ground sample point The constraint weighting coefficients for the lateral smoothness constraints of the road; The constraint weight coefficient is the longitudinal continuity constraint of the road.
[0067] By jointly solving the sample fitting term and the road smoothness constraint term, the target elevation surface can reflect the current ground condition of the road while avoiding abrupt changes in the locally corrected elevation surface.
[0068] S4, based on the road target elevation surface, corrects the DEM in the area to be processed to obtain a locally corrected DEM.
[0069] Based on the calculated target road elevation surface, one or more combinations of operations are performed on the DEM within the area to be processed, including leveling, interpolation, local stretching, local compression, and smooth transition. To avoid the formation of steep slopes, a gradual transition method is used within the edge transition zone to ensure a continuous connection between the locally corrected DEM and the uncorrected DEM.
[0070] In the specific calculation, let the original DEM be... The corrected local modified DEM is The core correction area within the region to be processed is The edge transition zone is The locally corrected DEM is then calculated using the following formula:
[0071]
[0072] The terms and characters in the formula are explained below: The original DEM; This is the corrected local DEM; This is the core correction area within the region to be processed, used for direct correction. This serves as the outer edge transition zone surrounding the core correction area; The gradual weighting function is derived from the point... transition zone to edge The normalized distance of the outer boundary is determined.
[0073] By directly using the road target elevation surface in the core correction area and using gradient fusion in the edge transition zone, the corrected DEM can be continuously connected with the uncorrected DEM, thereby reducing the risk of boundary jumps and seams after local orthophoto updates.
[0074] S5, perform local orthorectified update.
[0075] The system retrieves the corrected local DEM, the original image, and image positioning parameters. It then performs orthorectification and resampling based on the image positioning parameters only on the pixels within the local update window to generate a local orthorectified result. Meanwhile, the pixels of the initial DOM outside the local update window remain unchanged. This avoids repetitive generation of the entire scene while maintaining geometric consistency, enabling rapid local updates and significantly reducing processing time.
[0076] S6, update mask writeback.
[0077] An updated mask is generated based on the edge transition zone. The local orthophoto result is then smoothly written back to the initial DOM based on the updated mask to obtain the updated DOM, which is then displayed in real time on the interface.
[0078] The formula for writing back is:
[0079]
[0080] Update mask Defined as:
[0081]
[0082] The terms and characters in the formula are explained below: , In planar coordinates; This is the initial DOM; This is a local orthophoto result; For the updated DOM; This is the corresponding update mask function, used to control the boundary stitching range and fusion weights.
[0083] As explained above, this invention does not regenerate the entire DOM, but only performs orthogonal inverse resolution and resampling on the pixels within the local update window, and smoothly writes the local results back to the current DOM through a mask weighting method, thereby ensuring geometric consistency while taking into account production efficiency.
[0084] Combination Figure 2 (c) Figure 3 (c) and Figure 4 It can be seen that after adopting the method of the present invention, the continuity of the road boundary in the two sets of cases is significantly improved, the original distortion and striping phenomena are eliminated or significantly reduced, and the connection between the road and the surrounding features is more natural. This shows that the present invention can effectively solve the problem of the difficulty in restoring the spatial correspondence of the patched image in the background technology by correcting the DEM at the three-dimensional geometric level and triggering local orthorectification recalculation.
[0085] S7, Quality Inspection and Closed-Loop Verification.
[0086] The system displays the updated road area in real time and supports split-screen comparison, flickering comparison, or transparent overlay comparison with the previous results. The updated DOM is quantitatively checked based on the misalignment between road boundaries and adjacent features, the continuity of the road centerline, and the consistency of road width. If the check results do not meet the preset requirements, the system returns to step S3 to reconstruct the road target elevation surface and executes subsequent steps to continue correcting the DEM within the area to be processed until the preset requirements are met.
[0087] Specific quantitative evaluation is achieved by calculating the misalignment index. Continuity indicators Width Consistency Index The formula is as follows:
[0088]
[0089]
[0090]
[0091] If and only if , and If the result meets the preset requirements, it is determined that the inspection result does not meet the preset requirements; otherwise, it is determined that the result does not meet the preset requirements and the iteration is continued.
[0092] The terms and characters in the formula are explained below: The misalignment index is used to characterize the average degree of misalignment between the road boundary and adjacent stable features; It is a continuity indicator used to characterize the smoothness and continuity of the road centerline; It is a width consistency index used to characterize the consistency of road width along the route; For the first One corrected road boundary sampling point; For the first One nearby reference feature corresponding point; This represents the total number of boundary sampling points; For the first Discrete points along the road centerline For the first Discrete points along the road centerline For the first Discrete points along the road centerline; This represents the total number of discrete points along the road centerline. The width value is obtained by sampling along the road; K represents the average width of all sampled width values; K is the total number of width sample points. , and These are the preset thresholds for the corresponding indicators configured in the system. S8, Results Preservation and Traceability.
[0093] When the inspection results meet the above preset requirements, the system saves the locally corrected DEM, the locally updated mask, the editing log (including tool type, editing range, timestamp, and operator information), and the final updated DOM results, and can then proceed to the next road anomaly area for processing. The system can also include a caching unit to support fast rollback, version rollback, and multi-solution comparison.
[0094] It should be noted that, through the collaborative organization method of this invention, "patching and repairing on images" is transformed into "orthorectification geometric source control." This invention's method is applicable not only to standard road distortion areas caused by road construction, widening, or rerouting, but also to local anomalies caused by insufficient reference DEM timeliness in bridge approach roads, ramps, connecting lines, and site edges. Furthermore, the projection deformation patterns caused by elevation mismatch during orthorectification are not limited to roads; features with specific geometric topological relationships, such as buildings and power lines, will also deform. When the target to be processed expands from roads to other linear, strip-shaped, or block-shaped features, only the stable sample point selection strategy and the continuous / smooth constraint term in S3 need to be appropriately adjusted according to the geometric characteristics of the target object itself. The entire mechanism and interactive iterative process of "engineering organization—target elevation surface driving—local recalculation—mask rewriting—quantitative closed-loop verification" proposed in this invention remains applicable, greatly improving the DOM production and correction efficiency for various types of local deformation.
[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A road distortion correction method based on DEM interactive editing and local orthophoto update, characterized in that, Includes the following steps: S1, construct the job project, and import the original images, initial DOM, DEM and image positioning parameters of the same job area; S2, identify the road distortion region in the initial DOM, delineate the region to be processed, and expand outward along the region to be processed to form a local update window and edge transition zone; S3. Select stable ground sample points around the area to be processed, and construct the target elevation surface of the road based on the elevation of the stable ground sample points, combined with the road lateral smoothness constraint and the road longitudinal continuity constraint. S4. Based on the road target elevation surface, the DEM in the area to be processed is corrected to obtain a locally corrected DEM; wherein, a gradient method is used in the edge transition zone to make the locally corrected DEM continuously connected with the uncorrected DEM. S5, based on the local modified DEM, the original image and the image positioning parameters, perform orthorectification and resampling only on the pixels within the local update window to generate a local orthorectification result; S6. Generate an update mask based on the edge transition zone, and write the local orthophoto result back to the initial DOM based on the update mask to obtain the updated DOM.
2. The method according to claim 1, characterized in that, In step S2, the method for delineating the area to be processed includes: directly delineating the road distortion area on the initial DOM; or, delineating the road distortion area after assisting in locating it by combining road vector data, existing ground feature results, or road centerline. Based on the defined road distortion area, a buffer zone of a preset width is extended outward to form the local update window and edge transition zone used for DEM correction and orthophoto inverse kinematics calculation.
3. The method according to claim 1, characterized in that, In step S3, the stable ground sample points are selected from the ground on both sides of the road, the normal road sections connecting the front and rear of the road, or the stable areas around the area to be processed that have not undergone deformation; the method for constructing the road target elevation surface includes any one of the following: average value calculation, weighted average calculation, local plane fitting, piecewise linear fitting, and surface fitting.
4. The method according to claim 3, characterized in that, In step S3, the road target elevation surface function within the area to be processed is denoted as... , , Let the coordinates be planar coordinates; the stable ground sample points are denoted as... ; This represents the total number of stable ground sample points; , For the first The planar coordinates of a stable ground sample point For the first The elevation values of a stable ground sample point; then the target road elevation surface is determined by solving the following objective function: In the formula, J is the objective function value. The fitting error term for the road target elevation surface to the stable ground sample points is denoted as . This is a lateral smoothness constraint for the road. This refers to the longitudinal continuity constraint term for the road. The elevation surface function of the road target in plane coordinates Elevation value at the location; For the first The weights corresponding to each stable ground sample point The constraint weighting coefficients for the lateral smoothness constraints of the road; The constraint weight coefficient is the longitudinal continuity constraint of the road.
5. The method according to claim 1, characterized in that, In step S4, the operation of correcting the DEM in the area to be processed includes one or more combinations of flattening, interpolation, local stretching, local pressure drop, and smooth transition. Let the original DEM be The corrected local modified DEM is as follows: The core correction area within the region to be processed is The edge transition zone is The locally corrected DEM is then calculated using the following formula: In the formula, , In planar coordinates; The elevation surface of the road target within the area to be processed is in coordinates The elevation value at that location, The gradual weighting function is derived from the point... The normalized distance to the outer boundary of the edge transition zone is determined.
6. The method according to claim 1, characterized in that, In step S6, the calculation formula for writing the local orthophoto result back to the initial DOM is as follows: In the formula, , For planar coordinates, For the initial DOM, This is a local orthophoto result. For the updated DOM; To update the mask, it is defined as: in, As the core correction area, It is an edge transition zone. This is a gradual weighting function.
7. The method according to claim 1, characterized in that, In step S5, the pixels of the initial DOM outside the local update window remain unchanged to achieve local fast update.
8. The method according to claim 7, characterized in that, Following step S6, the following steps are also included: S7. The updated DOM is checked based on the misalignment between the road boundary and adjacent features, the continuity of the road centerline, and the consistency of the road width. If the check result does not meet the preset requirements, the process returns to step S3 to reconstruct the road target elevation surface and executes subsequent steps to continue to correct the DEM in the area to be processed until the preset requirements are met. S8. When the inspection results meet the preset requirements, save the local modified DEM, the local update record, and the updated DOM.
9. The method according to claim 8, characterized in that, In step S7, the misalignment index is calculated. Continuity indicators Width Consistency Index The formulas are as follows: when , and If the result meets the preset requirements, the system determines that the inspection result does not meet the preset requirements and continues iterating. In the formula, For the first Discrete points along the road centerline; For the first Discrete points along the road centerline; For the first One corrected road boundary sampling point, For the first Corresponding points of nearby reference features This represents the total number of boundary sampling points; For the first Discrete points along the road centerline This represents the total number of discrete points along the road centerline. The width value is obtained by sampling along the road. K represents the average width, and K represents the total number of width sampling points. , and All are preset thresholds.
10. The method according to claim 1, characterized in that, The image positioning parameters include one or more combinations of rational polynomial coefficients (RPC), exterior orientation elements, or correction parameters obtained from regional network adjustment.