Multi-view satellite image digital surface model reconstruction method and device based on edge constraint and light and shade recovery shape, equipment and storage medium

Abstract

The invention discloses a multi-view satellite image digital surface model reconstruction method, device and equipment based on edge constraint and light and shade recovery shape and a storage medium, and relates to the technical field of digital surface models.The method comprises the steps that an initial ground three-dimensional model is constructed based on a multi-view satellite image of a target area; performing preset edge fusion according to the multi-view satellite image and the initial ground three-dimensional model to obtain target edge information; constructing a target energy function based on a light and shade recovery shape based on the target edge information; and refining the initial ground three-dimensional model based on the target energy function to obtain a target ground three-dimensional model. According to the method, the edge information of the initial ground three-dimensional model is enriched through the edge information of the multi-view satellite image, the target energy function is determined based on the fused finer target edge information and the light and shade recovery shape technology, and then the initial ground three-dimensional model is optimized based on the target energy function. And the reconstruction precision of the final target ground three-dimensional model is improved.

Description

Technical Field

[0001] The present application relates to the technical field of digital surface models, and in particular to a method, device, equipment and storage medium for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and light and shade restoration shapes. Background Art

[0002] Given the high cost of laser scanning and the rapid development of satellite technology, multi-view satellite images have become the most popular data source for the reconstruction of Digital Surface Models (DSM). Multi-view stereo (MVS) technology, as a standard method for reconstructing DSM from multi-view satellite images, has been studied for decades, covering both traditional methods and learning-based methods.

[0003] However, whether it is the traditional method or the learning-based MVS technique, the accuracy of the generated DSM is often unsatisfactory when processing multi-view satellite images with weak texture areas and different terrain types.

[0004] Therefore, how to effectively improve the accuracy of reconstructed digital surface models has become an important issue that needs to be solved urgently. Summary of the invention

[0005] The main purpose of this application is to provide a method, device, equipment and storage medium for reconstructing digital surface models of multi-view satellite images based on edge constraints and light and shade restoration shapes, aiming to solve the technical problem of how to effectively improve the accuracy of reconstructing digital surface models.

[0006] To achieve the above objectives, the present application proposes a method for reconstructing digital surface models of multi-view satellite images based on edge constraints and light and shade restoration shapes, the method comprising: Construct an initial ground 3D model based on multi-view satellite images of the target area; Performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information; Constructing a target energy function based on shading restoration shape based on the target edge information; The initial ground three-dimensional model is refined by using the target energy function to obtain a target ground three-dimensional model.

[0007] In one embodiment, image difference analysis is performed on the initial ground three-dimensional model and the multi-view satellite image to obtain balanced geometric data items; Determining a reflectivity data item based on the reflectivity of a preset imaging model and the reflectivity of the initial ground three-dimensional model; A target energy function based on shading restoration shape is determined according to the balanced geometry data item, the reflectivity data item and the target edge information.

[0008] In one embodiment, the step of determining a target energy function based on shading restoration shape according to the balanced geometry data item, the reflectivity data item and the target edge information comprises: Determining a geometric smoothness item according to the target edge information; A shading-based shape-recovery target energy function is determined based on the balanced geometry data item, the reflectivity data item, and the geometry smoothness item.

[0009] In one embodiment, the step of determining a target energy function based on shading restoration shape based on the balanced geometric data item, the reflectivity data item and the geometric smoothness item comprises: determining a reflection smoothness term based on the multi-view satellite image; A shading-based shape-recovery target energy function is determined based on the balanced geometry data item, the reflectivity data item, the geometry smoothness item, and the reflectance smoothness item.

[0010] In one embodiment, the step of refining the initial ground three-dimensional model by using the target energy function to obtain the target ground three-dimensional model includes: Performing preset maximum value traversal on the target energy function to obtain target imaging information; The initial ground three-dimensional model is refined according to the target imaging information to obtain a target ground three-dimensional model.

[0011] In one embodiment, the step of performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information includes: Performing edge detection on the multi-viewpoint satellite image to obtain image edge information; Performing preset elevation detection on the initial ground three-dimensional model to obtain model edge information; The image edge information and the model edge information are fused to obtain target edge information.

[0012] In one embodiment, the step of performing preset elevation detection on the initial ground three-dimensional model to obtain model edge information includes: Performing elevation difference detection on adjacent pixels in the initial ground three-dimensional model to obtain elevation value change distribution information; The model edge information is determined based on the elevation value change distribution information and the preset elevation difference.

[0013] In addition, to achieve the above-mentioned purpose, the present application also proposes a multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shape, and the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shape includes: An initial ground 3D model building module is used to build an initial ground 3D model based on multi-view satellite images of the target area; An edge optimization module, used for performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information; A function construction module, used for constructing a target energy function based on shading restoration shape based on the target edge information; The target model construction module is used to refine the initial ground three-dimensional model through the target energy function to obtain a target ground three-dimensional model.

[0014] In addition, to achieve the above-mentioned purpose, the present application also proposes a multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shapes, the device comprising: a memory, a processor, and a multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes stored in the memory and executable on the processor, the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes being configured to implement the steps of the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes as described above.

[0015] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a storage medium, on which is stored a multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes. When the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes is executed by a processor, the steps of the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes as described above are implemented.

[0016] The present application provides a method, device, equipment and storage medium for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading recovery shape. The method includes: constructing an initial ground three-dimensional model based on multi-view satellite images of the target area; performing preset edge fusion according to the multi-view satellite images and the initial ground three-dimensional model to obtain target edge information; constructing a target energy function based on shading recovery shape based on the target edge information; and refining the initial ground three-dimensional model based on the target energy function to obtain the target ground three-dimensional model. The present application performs edge detection on the original multi-view satellite image and the initial ground three-dimensional model after preliminary reconstruction, respectively, so as to enrich the edge information of the initial ground three-dimensional model through the edge information of the multi-view satellite image, and realize more precise and accurate detection of the edge area, so as to construct a target energy function based on shading recovery shape technology according to the finer target edge information after fusion, and then optimize the initial ground three-dimensional model based on the target energy function. Therefore, compared with the prior art, the present application can improve the reconstruction accuracy of the target ground three-dimensional model generated in the end based on the shading recovery shape technology optimized by edge information. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0019] Figure 1 This is a flow chart of a first embodiment of a method for reconstructing a digital surface model from multi-view satellite images based on edge constraints and light-shading shape restoration of the present application; Figure 2 This is a flow chart of a second embodiment of a method for reconstructing a digital surface model from multi-view satellite images based on edge constraints and light and shade restoration shapes of the present application; Figure 3 This is a schematic diagram of the model reconstruction process of the second embodiment of the multi-viewpoint satellite image digital surface model reconstruction method based on edge constraints and light and shade recovery shape of the present application; Figure 4 This is a schematic diagram for comparing model improvements of the second embodiment of the multi-viewpoint satellite image digital surface model reconstruction method based on edge constraints and light and shade restoration shape of the present application; Figure 5 This is a schematic diagram of model contour comparison of the second embodiment of the multi-view satellite image digital surface model reconstruction method based on edge constraints and light and shade restoration shape of the present application; Figure 6 This is a schematic diagram of the module structure of a device for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and light and shade restoration shape according to an embodiment of the present application; Figure 7 Schematic diagram of the device structure of the hardware operating environment involved in the method for reconstructing a digital surface model of multi-view satellite images based on edge constraints and light and shade restoration shape in an embodiment of the present application.

[0020] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0021] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0022] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0023] The main solution of this application is: construct an initial ground three-dimensional model based on multi-view satellite images of the target area; perform preset edge fusion based on the multi-view satellite images and the initial ground three-dimensional model to obtain target edge information; construct a target energy function based on light and shade restoration shape based on the target edge information; refine the initial ground three-dimensional model through the target energy function to obtain the target ground three-dimensional model.

[0024] Currently, the accuracy of DSM generated based on multi-view satellite images is often unsatisfactory when processing satellite images with weak texture areas and different terrain types.

[0025] To solve the above problems, the present application performs edge detection on the original multi-view satellite image and the initial ground three-dimensional model after preliminary reconstruction respectively, so as to enrich the edge information of the initial ground three-dimensional model with the edge information of the multi-view satellite image, and realize more precise and accurate detection of the edge area, thereby constructing a target energy function based on the shading recovery shape technology according to the finer target edge information after fusion, and then optimizing the initial ground three-dimensional model based on the target energy function, so as to improve the reconstruction accuracy of the target ground three-dimensional model finally generated by the shading recovery shape technology based on edge constraint optimization.

[0026] It should be noted that the execution subject of this embodiment can be a multi-view satellite image digital surface model reconstruction system based on edge constraints and shading recovery shape, or a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or a multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shape that can achieve the above functions, etc. This embodiment does not specifically limit this. The following takes the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shape (referred to as reconstruction device) as an example of the execution subject to illustrate this embodiment and the following embodiments.

[0027] Based on this, the embodiment of the present application provides a method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and light and shade restoration shape, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the method for reconstructing digital surface models from multi-view satellite images based on edge constraints and light-shading shape restoration in this application.

[0028] In this embodiment, the method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and light and shade restoration shape includes steps S10 to S40: Step S10, constructing an initial ground three-dimensional model based on multi-view satellite images of the target area; It is easy to understand that the purpose of this embodiment is to establish a digital surface 3D reconstruction model corresponding to the target area, and the multi-view satellite image is a satellite image of the target area taken from multiple perspectives. Specifically, this embodiment can use an improved semi-global matching (SGM, Semi-Global Matching) method to process the multi-view satellite image of the target area to reconstruct the initial ground 3D model.

[0029] Step S20, performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information; It is understandable that the existing digital surface model reconstruction method cannot process the weak texture area information of multi-view satellite images well. Therefore, this embodiment can detect edge information of multi-view satellite images and initial ground 3D models respectively, and then fuse the edge information of the two, that is, perform the above-mentioned preset edge fusion to obtain target edge information with richer edge information. Therefore, this embodiment can subsequently perform edge constraints on the initial ground 3D model based on the richer target edge information to obtain a more accurate target ground 3D model after optimization.

[0030] Therefore, in a feasible implementation manner, in this embodiment, step S10 may include steps A1 to A3: Step A1, performing edge detection on the multi-viewpoint satellite image to obtain image edge information; Step A2, performing a preset elevation detection on the initial ground three-dimensional model to obtain model edge information; Step A3: Fusing the image edge information and the model edge information to obtain target edge information.

[0031] It should be understood that a new edge fusion module is proposed in this embodiment to perform edge detection on the structural edges of the multi-view satellite image and the discontinuous areas of the initial ground 3D model, respectively. That is, the edge detection process in this embodiment is divided into two independent stages, and different detection strategies are adopted for the multi-view satellite image and the initial ground 3D model, respectively.

[0032] Specifically, in this embodiment, the reconstruction device can first use canny edge detection to perform edge detection on the multi-view satellite image to obtain image edge information reflecting the structural edges of the multi-view satellite image; then the reconstruction device can perform preset elevation detection on the initial ground three-dimensional model to obtain model edge information representing the discontinuous area of ​​the initial ground three-dimensional model.

[0033] It should be noted that, due to the inevitable errors in the initial ground three-dimensional model, this embodiment needs to expand the edges of the image and the initial ground three-dimensional model based on the intersection of the image edge information and the model edge information to ensure that all discontinuous areas are fully covered. Specifically, the reconstruction device can fuse the discontinuous areas of the initial ground three-dimensional model and the edges in the image in the image space, and then generate an edge map, that is, the above-mentioned target edge information. The target edge information can accurately characterize the intersection of the edges detected in the multi-view satellite image and the initial ground three-dimensional model.

[0034] In a feasible implementation manner, in this embodiment, step A2 may include steps A21-A22: Step A21, performing elevation difference detection on adjacent pixels in the initial ground three-dimensional model to obtain elevation value change distribution information; Step A22, determining the model edge information based on the elevation value change distribution information and the preset elevation difference.

[0035] It should be understood that in the process of detecting the edge information of the initial ground three-dimensional model, the reconstruction device can identify the discontinuous areas of the initial ground three-dimensional model based on the changes in the elevation values ​​of adjacent pixels in the model. Specifically, in this embodiment, the reconstruction device can obtain the distribution information of the changes in elevation values ​​by traversing the elevation differences of adjacent pixels in the model. Subsequently, based on the distribution information, when the reconstruction device detects that the difference in elevation values ​​of adjacent pixels exceeds a preset threshold (for example, 0.5 meters), these adjacent pixels are determined to be discontinuously distributed pixel pairs. Ultimately, all areas marked as discontinuously distributed pixel pairs will be identified as discontinuous areas, thereby achieving the extraction of the edge information of the model.

[0036] Step S30, constructing a target energy function based on shading restoration shape based on the target edge information; Step S40: Refine the initial ground three-dimensional model by using the target energy function to obtain a target ground three-dimensional model.

[0037] It should be noted that, in this embodiment, the reconstruction device can use Shape From Shading (SFS) to further improve the accuracy of the optimized target ground 3D model. The Shape From Shading can reconstruct the detailed surface in the weak texture area, and use the designed geometric smoothness term to reconstruct the smooth surface in the weak texture area. However, when the existing SFS technology is used for model refinement, due to the use of geometric smoothness terms, discontinuous areas such as roof edges are easily flattened, which causes a significant decrease in the accuracy and fidelity of the reconstructed DSM.

[0038] In view of this, this embodiment introduces the above-mentioned target energy function based on target edge information, which takes into account the imaging process of the SFS technology, that is, the above-mentioned target energy function based on shading recovery shape can not only realize the rendering of illumination, surface albedo and surface normal for each pixel value of the multi-view satellite image, but also ensure high-precision reconstruction of discontinuous areas. Finally, this embodiment can be guided by edge information and refine the initial ground three-dimensional model by optimizing the above-mentioned target energy function.

[0039] Specifically, this embodiment can model the imaging process of the digital surface three-dimensional reconstruction model as the interaction of illumination, surface albedo and surface normal, and then design the above-mentioned target energy function to minimize the difference between the rendered image (the output result of the initial ground three-dimensional model) and the observed image (multi-viewpoint satellite image), thereby solving the surface normal (or the three directions of the surface normal vector) of the target ground three-dimensional model, and at the same time restore the fine shape details of the target ground three-dimensional model, thereby improving the model accuracy of the target ground three-dimensional model.

[0040] It should be understood that the SFS technology can use illumination, surface albedo and surface normal to model the imaging process, and then refine the initial ground 3D model by optimizing the difference between the rendered image and the observed image, so as to achieve the effect of restoring the surface details of the initial ground 3D model. Therefore, before introducing the target energy function, this embodiment needs to explain the reflectivity model corresponding to the SFS technology.

[0041] Exemplarily, the reflectivity model in this embodiment may use a Lambertian reflectivity model, and use a second-order spherical harmonic basis function to simulate the reflectivity, and is expressed as follows: ; (1)

[0042] ; (2) in, is the initial ground 3D model surface point In the The corresponding pixel values ​​on the satellite image may be different in different images. is the initial ground 3D model surface point The albedo of the surface is independent of the viewing angle and may be different at different surface points; is the initial ground 3D model surface point Normal; N is related to the order of the spherical harmonic basis function, and since the second-order spherical harmonic basis function is used in this embodiment, it is set to 9; - is the spherical harmonic basis, is the surface normal Function of - are the coefficients of the spherical harmonic basis, which are set to be dependent on the viewing angle, i.e. different images have different coefficients.

[0043] It is easy to understand that this embodiment can construct a target energy function for subsequently refining the initial ground three-dimensional model based on Formula 1. Specifically, after obtaining the target energy function, this embodiment can determine the initial values ​​of the parameters in the target energy function based on the initial ground three-dimensional model and the multi-view satellite image. Then, this embodiment can gradually solve the optimal model parameters corresponding to the initial ground three-dimensional model by optimizing the target energy function, and then obtain the above-mentioned target ground three-dimensional model based on the optimal model parameters.

[0044] In this embodiment, in order to ensure the model reconstruction accuracy in weak texture and discontinuous areas, edge detection is performed on the original multi-view satellite image and the initial ground three-dimensional model after preliminary reconstruction, so as to enrich the edge information of the initial ground three-dimensional model with the edge information of the multi-view satellite image, thereby achieving more precise and accurate detection of the edge area, and constructing a target energy function based on the shading recovery shape technology based on the more refined target edge information after fusion, and then optimizing the initial ground three-dimensional model based on the target energy function. Therefore, compared with the prior art, the present application can improve the reconstruction accuracy of the target ground three-dimensional model finally generated based on the shading recovery shape technology based on edge information optimization.

[0045] The present embodiment provides a method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading recovery shape, the method comprising: constructing an initial ground three-dimensional model based on the multi-view satellite image of the target area; performing edge detection on the multi-view satellite image to obtain image edge information; performing elevation difference detection on adjacent pixels in the initial ground three-dimensional model to obtain elevation value change distribution information; and determining the model edge information based on the elevation value change distribution information and the preset elevation difference. The image edge information and the model edge information are fused to obtain the target edge information. A target energy function based on shading recovery shape is constructed based on the target edge information; the initial ground three-dimensional model is refined by the target energy function to obtain the target ground three-dimensional model. In order to ensure the model reconstruction accuracy in weak texture and discontinuous areas, the present embodiment performs edge detection on the original multi-view satellite image and the initial ground three-dimensional model after preliminary reconstruction respectively, so as to enrich the edge information of the initial ground three-dimensional model through the edge information of the multi-view satellite image, realize more precise and accurate detection of the edge area, and construct a target energy function based on shading recovery shape technology according to the more refined target edge information after fusion, and then optimize the initial ground three-dimensional model based on the target energy function. Therefore, compared with the prior art, the present application can improve the reconstruction accuracy of the target ground three-dimensional model that is finally generated based on the shading recovery shape technology optimized by edge information.

[0046] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction and will not be repeated later.

[0047] Based on the first embodiment, refer to Figure 2 , Figure 2 This is a first flow chart of a second embodiment of a method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and light and shade restoration shape. In this embodiment, step S30 may include steps B1 to B3: Step B1, performing image difference analysis on the initial ground three-dimensional model and the multi-viewpoint satellite image to obtain balanced geometric data items; It is easy to understand that the above equilibrium geometric data items can be expressed as , the balanced geometric data item characterizes the difference between corresponding points in different satellite images. The basic assumption is that the pixel values ​​of the same surface point in different satellite images should be the same. However, due to changes in illumination, occlusion, and image noise, the pixel values ​​of the same surface point are often different. Therefore, considering only the difference in pixel values ​​is not robust. In order to obtain better performance, this embodiment can use the normalized cross correlation (NCC) within 5×5 pixels to calculate the similarity between the pixels of the same surface point in the initial ground three-dimensional model in different satellite images, and obtain the above-mentioned balanced geometric data item , Specifically expressed as: ; (3) in, is the number of visible images of the surface points of the initial ground 3D model, is the surface of the initial ground 3D model, are the coordinates of the surface point on the two-dimensional plane, and Respectively represent the mth and nth visible image surface points in the multi-view satellite image The corresponding pixel value of It is the NCC function that calculates the similarity of image pixels; It is the elevation value of the surface point after displacement along the Z axis.

[0048] Step B2, determining a reflectivity data item based on the reflectivity of a preset imaging model and the reflectivity of the initial ground three-dimensional model; It is easy to understand that since the present embodiment uses satellite images far from the ground to reconstruct a multi-view satellite image digital surface model based on edge constraints and shading recovery shape, the visibility of each surface point in the initial three-dimensional model depends on whether the surface point can be reprojected onto the image plane.

[0049] It should be understood that, in this embodiment, the above reflectivity data item can be expressed as , which can be designed based on a preset imaging model, namely the Lambertian reflectance model represented by the above formulas (1) and (2). The basic assumption is that the pixel value of the image is the interaction of illumination, surface albedo and surface normal, and the detailed surface can be recovered by solving the surface normal from the energy function. Ideally, when the illumination, surface albedo and surface normal are accurate, the image rendered with the preset imaging model should be the same as the image of the initial ground 3D model. Therefore, the reflectance data term Specifically expressed as: (4) in, is the elevation value of the surface point after displacement along the Z axis, is the surface albedo, are the coefficients of the spherical harmonic basis, The first Initial ground 3D model surface points in the visible image The rendered pixel value.

[0050] Step B3, determining a target energy function based on shading restoration shape according to the balanced geometry data item, the reflectivity data item and the target edge information.

[0051] It is easy to understand that based on the balanced geometric data items and reflectivity data items determined by the above formulas (1)-(4), combined with the target edge information, this embodiment can determine the target energy function to subsequently guide the model optimization of the initial ground three-dimensional model. It should be noted that in this process, it is also necessary to assume that the surface albedo and surface normal of the digital surface model are independent of the viewing angle. At the same time, the albedo of different surface points may be different, but the illumination in the same image, that is, the illumination coefficient, is the same.

[0052] It is understandable that in order to obtain a robust reconstruction result, this embodiment further designs a geometric smoothness term. In a feasible implementation, in this embodiment, step B3 may include steps C1-C2: Step C1, determining a geometric smoothness item according to the target edge information; Step C2, determining a target energy function based on shading restoration shape based on the balanced geometric data item, the reflectivity data item and the geometric smoothness item; It is understandable that the above geometric smoothness term can be used It is indicated that this geometric smoothness term is used to constrain the distance between a surface point and its neighboring points. It is easy to understand that in most areas, the surface of the digital surface model should be smooth and the distance between a surface point and its neighboring points should be small. However, this assumption is not very applicable to multi-terrain areas. For example, in urban scenes with many discontinuous areas, many areas will violate this assumption, which makes it difficult to accurately restore the DSM surface.

[0053] In order to solve the above problem, this embodiment can introduce the edge map of the discontinuous area, that is, the target edge information, when constructing the above geometric smoothness term, so as to apply different smoothness constraints to the discontinuous areas in the initial ground model based on the detected edge map. For example, this embodiment can set the weight of the detected discontinuous area to zero, thereby avoiding the application of geometric smoothness constraints in the discontinuous area of ​​the initial ground three-dimensional model, preventing the discontinuous area from being flattened by mistake, and thus improving the model reconstruction accuracy.

[0054] Specifically, the above geometric smoothness constraint It can be expressed as: ; (5) in, is the initial ground 3D model surface point The weight of can be determined according to the target edge information; is the initial ground 3D model surface point Neighborhood of; Belongs to Surface points of is the surface point computed using the guided filter The weight of and The surface points are and The elevation value of yes spatial resolution; is the number of neighbor points.

[0055] Specifically, the weights The calculation method of is as follows: (6) in, is a parameter to avoid division by zero when the variance is zero; yes The average value of the pixels in the yes Variance of pixel values ​​in ; and are the initial ground 3D model surface points in any satellite image. and The corresponding pixel value.

[0056] It is easy to understand that in order to better incorporate the surface albedo into the model Separated from the spherical harmonic basis coefficients, further, in a feasible implementation manner, in this embodiment, step C2 may include steps C21-C22: Step C21, determining a reflection smoothness item based on the multi-viewpoint satellite image; Step C22, determining a target energy function based on shading restoration shape according to the balanced geometry data item, the reflectivity data item, the geometry smoothness item and the reflection smoothness item.

[0057] It should be understood that the above reflection smoothness term can be used As mentioned above, this embodiment may further assume that the surface albedo of different surface points in the initial ground three-dimensional model may be different. Then the reflection smoothness term constructed by this embodiment for robustness may be expressed as follows: ; (7) in, is a constant, and is the surface point in all satellite images and The average value of the corresponding pixel values, and They are the surface points in the initial ground 3D model. and The surface albedo.

[0058] In summary, this embodiment can ultimately establish a target energy function based on the above-mentioned balanced geometric data items, reflectivity data items, geometric smoothness items and reflection smoothness items, through the above-mentioned preset imaging model and shading recovery shape technology, and using light coefficients, surface albedo and surface normals. At this time, the target energy function can be expressed as follows: ; (8) in, is the elevation value of the surface point after displacement along the Z axis, is the surface albedo, are the coefficients of the spherical harmonic basis, , and To balance the geometric data items , reflection data items , geometric smoothness term and the reflection smoothness term And set the weights separately.

[0059] It should be noted that after obtaining the target energy function represented by the above formula (8), this embodiment can determine the initial values ​​of the parameters in the target energy function according to the initial ground three-dimensional model and the multi-view satellite image, and then this embodiment can gradually solve higher-precision values ​​by optimizing the target energy function, and then obtain the optimal illumination coefficient, surface albedo and surface normal (surface normal is the second-order derivative of the surface point in the DSM), and then determine the elevation information after the initial ground three-dimensional model is refined based on the information obtained by the solution, and finally obtain a high-precision target ground three-dimensional model.

[0060] In a feasible implementation manner, in this embodiment, step S40 includes steps D1-D2: Step D1, performing a preset maximum value traversal on the target energy function to obtain target imaging information; Step D2, refining the initial ground three-dimensional model according to the target imaging information to obtain a target ground three-dimensional model.

[0061] It can be understood that the process of solving the optimal parameters in this embodiment can be to minimize the above-mentioned target energy function, that is, to traverse the target energy function values ​​under different parameters (i.e., illumination coefficient, surface albedo and surface normal), and obtain the minimum value therein, and then use the illumination coefficient, surface albedo and surface normal corresponding to the minimum value of the target energy function as the above-mentioned target imaging information to optimize the elevation information of the initial ground three-dimensional model, and then obtain an accurate target ground three-dimensional model. Exemplarily, this embodiment can use the Levenberg-Marquardt algorithm in Ceres Solver to perform the above-mentioned preset maximum value traversal.

[0062] In the specific implementation, refer to Figure 3 , Figure 3 This is a schematic diagram of the model reconstruction process of the second embodiment of the multi-view satellite image digital surface model reconstruction method based on edge constraints and light and shade recovery shape of this application. Figure 3 As shown, the reconstruction device can use the input multi-view high-resolution satellite images and apply the modified semi-global matching method (mSGM) to reconstruct the initial ground 3D model. Then, the reconstruction device can use the carefully designed edge fusion module to perform preset edge fusion to accurately detect the discontinuous area, that is, the target edge information mentioned above. Finally, the reconstruction device can use the multi-view satellite images and edge maps to optimize the initial ground 3D model based on the SFS technology to obtain a refined target ground 3D model.

[0063] For better understanding, refer to Figure 4 and Figure 5 , Figure 4 This is a schematic diagram for comparing model improvements of the second embodiment of the method for reconstructing a digital surface model from multi-view satellite images based on edge constraints and light and shade restoration of shape in this application. Figure 4 (a) shows a multi-view satellite image. Figure 4 (b) shows the digital surface model constructed based only on the improved semi-global matching method (mSGM), while Figure 4 (c) shows the digital surface model generated by the digital ground reconstruction method proposed in this application that combines improved semi-global matching, edge fusion and shading recovery shape technology. Figure 4 It can be seen that, compared with the digital surface model reconstruction method based only on mSGM, the digital surface model reconstructed by the method proposed in this embodiment has higher accuracy, especially in processing edge details.

[0064] and Figure 5This is a schematic diagram of model contour comparison of the second embodiment of the multi-viewpoint satellite image digital surface model reconstruction method based on edge constraints and light and shade recovery shape of the present application. Figure 5 (a) to (e) are schematic diagrams of house outlines in multi-view satellite images, and schematic diagrams of house outlines based on the improved semi-global matching method (mSGM), nFrames SURE, SADGE (Shading aware DSM generation from high resolution multi view satellite images), and digital surface models reconstructed based on the method of this embodiment. Figure 5 It is easy to know that the discontinuous structure contour of the DSM reconstructed by the method proposed in this embodiment has the highest matching degree with the house contour in the multi-view satellite image, and therefore has the highest accuracy.

[0065] In this implementation, a new target energy function is proposed, which minimizes the difference between the rendered image and the multi-view satellite image based on the rich target edge information, and can restore fine shape details. Specifically, this embodiment can determine the optimal spherical harmonic illumination coefficient, surface albedo and surface normal based on the designed balanced geometric data items, reflectivity data items, geometric shape smoothness constraints and reflection smoothness constraints, which can not only avoid the model discontinuous area from being wrongly flattened, but also effectively distinguish the albedo and spherical harmonic basis coefficients, so as to obtain the optimized elevation value based on the optimal target imaging information, and then restore the detailed shape of the digital surface model based on the optimized elevation value, and finally achieve the reconstruction of DSM with high precision and high fidelity in weak texture and discontinuous areas.

[0066] In this embodiment, image difference analysis is performed on the initial ground three-dimensional model and the multi-view satellite image to obtain a balanced geometric data item; a reflectivity data item is determined based on the reflectivity of the preset imaging model and the reflectivity of the initial ground three-dimensional model; a geometric smoothness item is determined according to the target edge information; a reflection smoothness item is determined based on the multi-view satellite image; and a target energy function based on light and shade recovery shape is determined according to the balanced geometric data item, the reflectivity data item, the geometric smoothness item and the reflection smoothness item. The target energy function is traversed by the preset maximum value to obtain the target imaging information; the initial ground three-dimensional model is refined according to the target imaging information to obtain the target ground three-dimensional model. This embodiment proposes a new target energy function, which minimizes the difference between the rendered image and the multi-view satellite image based on the rich target edge information, and can restore fine shape details. Specifically, this embodiment can determine the optimal spherical harmonic illumination coefficient, surface albedo and surface normal based on the designed balanced geometric data item, reflectivity data item, geometric shape smoothness constraint and reflection smoothness constraint, which can not only avoid the model discontinuous area from being flattened incorrectly, but also effectively distinguish the albedo and spherical harmonic basis coefficient. In this way, the optimized elevation value is obtained based on the best target imaging information, and the detailed shape of the digital surface model is restored based on the optimized elevation value, ultimately achieving reconstruction of the digital surface model with high accuracy and high fidelity in weak texture and discontinuous areas.

[0067] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the multi-view satellite image digital surface model reconstruction method based on edge constraints and light and shade restoration shape in the present application. More forms of simple transformations based on this technical concept are all within the scope of protection of the present application.

[0068] This application also provides a multi-view satellite image digital surface model reconstruction device based on edge constraints and light and dark recovery shape, please refer to Figure 6 , Figure 6 This is a schematic diagram of the module structure of a device for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration of shapes in an embodiment of the present application. In this embodiment, the device for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration of shapes includes: An initial ground three-dimensional model construction module 601 is used to construct an initial ground three-dimensional model based on multi-view satellite images of the target area; The edge optimization module 602 is used to perform preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information; Function construction module 603, used to construct a target energy function based on shading restoration shape based on the target edge information; The target model building module 604 is used to refine the initial ground three-dimensional model through the target energy function to obtain a target ground three-dimensional model.

[0069] Optionally, in this embodiment, the edge optimization module 602 is further used to perform edge detection on the multi-viewpoint satellite image to obtain image edge information; The edge optimization module 602 is further used to perform a preset elevation detection on the initial ground three-dimensional model to obtain model edge information; The edge optimization module 602 is further used to fuse the image edge information and the model edge information to obtain target edge information.

[0070] Optionally, in this embodiment, the edge optimization module 602 is further used to perform elevation difference detection on adjacent pixels in the initial ground three-dimensional model to obtain elevation value change distribution information; The edge optimization module 602 is further used to determine the model edge information based on the elevation value change distribution information and the preset elevation difference.

[0071] Optionally, in this embodiment, the function construction module 603 is further used to perform image difference analysis on the initial ground three-dimensional model and the multi-view satellite image to obtain a balanced geometric data item; The function construction module 603 is further used to determine the reflectivity data item based on the reflectivity of the preset imaging model and the reflectivity of the initial ground three-dimensional model; The function construction module 603 is further used to determine a target energy function based on shading restoration shape according to the balanced geometry data item, the reflectivity data item and the target edge information.

[0072] Optionally, in this embodiment, the function construction module 603 is further used to determine a geometric smoothness item according to the target edge information; The function construction module 603 is further used to determine a target energy function based on shading restoration shape based on the balanced geometric data item, the reflectivity data item and the geometric smoothness item.

[0073] Optionally, in this embodiment, the function construction module 603 is further used to determine a reflection smoothness item based on the multi-viewpoint satellite image; The function construction module 603 is further used to determine a target energy function based on shading restoration shape according to the balanced geometry data item, the reflectivity data item, the geometry smoothness item and the reflection smoothness item.

[0074] Optionally, in this embodiment, the target model building module 604 is further used to perform a preset maximum value traversal on the target energy function to obtain target imaging information; The target model building module 604 is further configured to refine the initial ground three-dimensional model according to the target imaging information to obtain a target ground three-dimensional model.

[0075] The device for reconstructing digital surface models of multi-view satellite images based on edge constraints and shading restoration shapes provided by the present application adopts the method for reconstructing digital surface models of multi-view satellite images based on edge constraints and shading restoration shapes in the above-mentioned embodiments, and can solve the technical problem of reconstructing digital surface models of multi-view satellite images based on edge constraints and shading restoration shapes. Compared with the prior art, the beneficial effects of the device for reconstructing digital surface models of multi-view satellite images based on edge constraints and shading restoration shapes provided by the present application are the same as the beneficial effects of the method for reconstructing digital surface models of multi-view satellite images based on edge constraints and shading restoration shapes provided by the above-mentioned embodiments, and other technical features in the device for reconstructing digital surface models of multi-view satellite images based on edge constraints and shading restoration shapes are the same as the features disclosed in the above-mentioned embodiment methods, which are not described in detail here.

[0076] The present application provides a device for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes. The device for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes in the above-mentioned embodiment 1.

[0077] Reference below Figure 7 , which shows a schematic diagram of the structure of a multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shape suitable for implementing the embodiment of the present application. The multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shape in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 7 The illustrated device for reconstructing a digital surface model from a multi-view satellite image based on edge constraints and light-shading restoration of shape is merely an example and should not limit the functions and scope of use of the embodiments of the present application.

[0078] like Figure 7 As shown, the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shape may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to the program stored in a read-only memory (ROM: Read Only Memory) 1002 or the program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. Various programs and data required for the operation of the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shape are also stored in RAM1004. The processing device 1001, ROM1002 and RAM1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 may allow the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shapes to communicate wirelessly or wired with other devices to exchange data. Although the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading recovery shapes with various systems is shown in the figure, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have instead.

[0079] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a multi-view satellite image digital surface model reconstruction program product based on edge constraints and shading recovery shapes, which includes a multi-view satellite image digital surface model reconstruction program based on edge constraints and shading recovery shapes carried on a computer-readable medium, and the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading recovery shapes contains program code for executing the method shown in the flowchart. In such an embodiment, the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading recovery shapes can be downloaded and installed from the network through a communication device, or installed from the storage device 1003, or installed from the ROM 1002. When the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading recovery shapes is executed by the processing device 1001, the above functions defined in the method of the embodiment disclosed in the present application are executed.

[0080] The multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shapes provided by the present application adopts the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes in the above-mentioned embodiment, which can solve the technical problem of improving the reconstruction accuracy of the existing digital surface model. Compared with the prior art, the beneficial effects of the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shapes provided by the present application are the same as the beneficial effects of the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes provided by the above-mentioned embodiment, and the other technical features of the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shapes are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0081] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0082] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0083] The present application provides a storage medium having computer-readable program instructions stored thereon (i.e., a multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes), and the computer-readable program instructions are used to execute the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes in the above-mentioned embodiment.

[0084] The storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM: RandomAccess Memory), a read-only memory (ROM: Read Only Memory), an erasable programmable read-only memory (EPROM: Erasable Programmable Read Only Memory or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM: CD-Read Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, system or device. The program code contained on the storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency: Radio Frequency), etc., or any suitable combination of the above.

[0085] The above-mentioned storage medium can be included in the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shape; or it can exist independently and not be assembled into the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shape.

[0086] The above-mentioned storage medium carries one or more programs. When the above-mentioned one or more programs are executed by the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shapes, the multi-view satellite image digital surface model reconstruction device based on edge constraints and shading restoration shapes: reconstructs the multi-view satellite image digital surface model based on edge constraints and shading restoration shapes.

[0087] The program code for reconstructing the digital surface model of multi-view satellite images based on edge constraints and shading recovery shapes for performing the operations of the present application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0088] The flowchart and block diagram in the accompanying drawings illustrate the possible implementation of the system, method and multi-view satellite image digital surface model reconstruction program product based on edge constraints and shading recovery shape according to various embodiments of the present application, The architecture, function and operation. In this regard, each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0089] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0090] The readable storage medium provided by the present application is a storage medium, which stores computer-readable program instructions for executing the above-mentioned method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes (i.e., a program for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes), which can solve the technical problem of reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes. Compared with the prior art, the beneficial effects of the storage medium provided by the present application are the same as the beneficial effects of the method for reconstructing a digital surface model of a multi-view satellite image based on edge constraints and shading restoration shapes provided by the above-mentioned embodiment, and will not be described in detail here.

[0091] The above are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for reconstructing digital surface models from multi-view satellite images based on edge constraints and shading recovery shapes, characterized in that: The method comprises: Construct an initial ground 3D model based on multi-view satellite images of the target area; Performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information; Constructing a target energy function based on shading restoration shape based on the target edge information; The initial ground three-dimensional model is refined by using the target energy function to obtain a target ground three-dimensional model.

2. The method for reconstructing digital surface models from multi-view satellite images based on edge constraints and shading recovery shapes as claimed in claim 1, characterized in that: The step of constructing a target energy function based on shading restoration shape based on the target edge information comprises: Performing image difference analysis on the initial ground three-dimensional model and the multi-viewpoint satellite image to obtain balanced geometric data items; Determining a reflectivity data item based on the reflectivity of a preset imaging model and the reflectivity of the initial ground three-dimensional model; A target energy function based on shading restoration shape is determined according to the balanced geometry data item, the reflectivity data item and the target edge information.

3. The method for reconstructing digital surface models from multi-view satellite images based on edge constraints and light-shading recovery shapes as claimed in claim 2, characterized in that: The step of determining a target energy function based on shading restoration shape according to the balanced geometry data item, the reflectivity data item and the target edge information comprises: Determining a geometric smoothness item according to the target edge information; A shading-based shape-recovery target energy function is determined based on the balanced geometry data item, the reflectivity data item, and the geometry smoothness item.

4. The method for reconstructing digital surface models from multi-view satellite images based on edge constraints and light-shading recovery shapes as claimed in claim 3, characterized in that: The step of determining a target energy function based on shading restoration shape based on the balanced geometric data item, the reflectivity data item and the geometric smoothness item comprises: determining a reflection smoothness term based on the multi-view satellite image; A shading-based shape-recovery target energy function is determined based on the balanced geometry data item, the reflectivity data item, the geometry smoothness item, and the reflectance smoothness item.

5. The method for reconstructing digital surface models from multi-view satellite images based on edge constraints and light-shading recovery shapes as claimed in claim 4, characterized in that: The step of refining the initial ground three-dimensional model by using the target energy function to obtain a target ground three-dimensional model comprises: Performing preset maximum value traversal on the target energy function to obtain target imaging information; The initial ground three-dimensional model is refined according to the target imaging information to obtain a target ground three-dimensional model.

6. The method for reconstructing digital surface models from multi-view satellite images based on edge constraints and light-shading recovery shapes as claimed in claim 1, characterized in that: The step of performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information includes: Performing edge detection on the multi-viewpoint satellite image to obtain image edge information; Performing preset elevation detection on the initial ground three-dimensional model to obtain model edge information; The image edge information and the model edge information are fused to obtain target edge information.

7. The method for reconstructing digital surface models of multi-view satellite images based on edge constraints and light-shading recovery shapes as claimed in claim 6, characterized in that: The step of performing preset elevation detection on the initial ground three-dimensional model to obtain model edge information includes: Performing elevation difference detection on adjacent pixels in the initial ground three-dimensional model to obtain elevation value change distribution information; The model edge information is determined based on the elevation value change distribution information and the preset elevation difference.

8. A device for reconstructing digital surface models of multi-view satellite images based on edge constraints and light and shade recovery shapes, characterized in that: The multi-view satellite image digital surface model reconstruction device based on edge constraints and light and shade restoration shape includes: An initial ground 3D model building module is used to build an initial ground 3D model based on multi-view satellite images of the target area; An edge optimization module, used for performing preset edge fusion according to the multi-viewpoint satellite image and the initial ground three-dimensional model to obtain target edge information; A function construction module, used for constructing a target energy function based on shading restoration shape based on the target edge information; The target model construction module is used to refine the initial ground three-dimensional model through the target energy function to obtain a target ground three-dimensional model.

9. A device for reconstructing digital surface models of multi-view satellite images based on edge constraints and light and shade recovery shapes, characterized in that: The device comprises: a memory, a processor, and a multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes stored in the memory and executable on the processor, wherein the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes is configured to implement the steps of the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores a multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes. When the multi-view satellite image digital surface model reconstruction program based on edge constraints and shading restoration shapes is executed by a processor, the steps of the multi-view satellite image digital surface model reconstruction method based on edge constraints and shading restoration shapes as described in any one of claims 1 to 7 are implemented.

Images