Method for constructing multi-dimensional layers for urban and rural territorial space monitoring

Abstract

Embodiments of the present disclosure disclose a method for constructing a multi-dimensional layer for urban land space monitoring. A specific implementation manner of the method includes: obtaining a sequence of high-resolution remote sensing images of an urban area captured by a remote sensing satellite; performing image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area; performing multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and constructing an initial layer block group set according to the image feature set; performing layer splicing on each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer, obtaining a single-dimensional layer set; mapping each single-dimensional layer in the single-dimensional layer set to a pre-set urban area map to obtain a multi-dimensional layer of the corresponding urban area. This implementation manner can generate a multi-dimensional layer with a more accurate regional boundary.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the fields of computer technology and layer construction, and particularly to a multi-dimensional layer construction method for urban territorial space monitoring. Background Art

[0002] With the development of technology, it is becoming increasingly important to refine and supplement relevant content in urban areas. At the same time, it is necessary to monitor the changes in land types such as cultivated land, garden land, forest land, grassland, water areas (including glaciers and perennial snow cover), highways, urban residential land, and construction land in certain areas, and to master the types, areas, scopes, distributions, and changes of natural resources and human geographical elements in the whole region, so as to support the compilation and implementation supervision of urban territorial space planning, daily change surveys, ecological restoration, physical examination and assessment, and use control and other territorial space governance work. Currently, when constructing layers, the commonly adopted method is to identify all elements in the area and organize them in the same layer.

[0003] However, when constructing multi-dimensional layers in the above manner, the following technical problems often exist:

[0004] Organizing the identified different regional elements (such as road areas, vegetation areas, etc.) in the same layer will make it difficult to accurately distinguish the boundaries between different regions, thus resulting in inaccurate multi-dimensional layers. At the same time, when rendering a specified area in the layer, due to the unclear boundary division, there will be a large error in the rendered area.

[0005] The above information disclosed in this background art section is only used to enhance the understanding of the background of the inventive concept, and therefore, it may include information that does not form the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] This content part of the present disclosure is used to briefly introduce the concepts, which will be described in detail in the following detailed implementation part. This content part of the present disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0007] Some embodiments of the present disclosure propose a multi-dimensional layer construction method for urban territorial space monitoring to solve the technical problems mentioned in the above background art section.

[0008] In a first aspect, some embodiments of the present disclosure provide a method for constructing a multi-dimensional layer for urban territorial space monitoring, the method comprising: obtaining a sequence of high-resolution remote sensing images of an urban area captured by a remote sensing satellite; performing image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area according to a preset digital elevation model corresponding to the urban area to obtain a fused remote sensing image of the urban area, wherein the image fusion processing includes a registration fusion operation and a projection operation on each high-resolution remote sensing image of the urban area, and the operation sequence of the registration fusion operation and the projection operation is set according to the regional features in each high-resolution remote sensing image of the urban area, and the fused remote sensing image of the urban area is a three-dimensional image including elevation values; performing multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and constructing an initial layer block group set according to the image feature set, wherein each initial layer block group in the initial layer block group set corresponds to one dimension; performing layer splicing on each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer, obtaining a single-dimensional layer set; mapping each single-dimensional layer in the single-dimensional layer set to a preset urban area map to obtain a multi-dimensional layer of the corresponding urban area, wherein each layer in the multi-dimensional layer corresponds to one layer option in a layer operation component on a display terminal page.

[0009] Second aspect, some embodiments of the present disclosure provide a multi-dimensional layer construction device for urban land space monitoring. The device includes: an acquisition unit configured to acquire a sequence of high-resolution remote sensing images of an urban area captured by a remote sensing satellite; an image fusion unit configured to perform image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area according to a preset digital elevation model corresponding to the urban area, to obtain a fused remote sensing image of the urban area, wherein the image fusion processing includes a registration fusion operation and a projection operation on each high-resolution remote sensing image of the urban area, and the operation sequence of the registration fusion operation and the projection operation is set according to the regional features in each high-resolution remote sensing image of the urban area, and the fused remote sensing image of the urban area is a three-dimensional image including elevation values; an image recognition unit configured to perform multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and construct an initial layer block group set according to the image feature set, wherein each initial layer block group in the initial layer block group set corresponds to one dimension; a single-dimensional layer construction unit configured to perform layer splicing on each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer, and obtain a single-dimensional layer set; a multi-dimensional layer construction unit configured to map each single-dimensional layer in the single-dimensional layer set to a preset urban area map to obtain a multi-dimensional layer of the corresponding urban area, wherein each layer in the multi-dimensional layer corresponds to one layer option in the layer operation components on the display terminal page.

[0010] Third aspect, some embodiments of the present disclosure provide an electronic device, including: one or more processors; a storage device storing one or more programs thereon, when the one or more programs are executed by the one or more processors, enabling the one or more processors to implement the method described in any implementation manner of the first aspect above.

[0011] Fourth aspect, some embodiments of the present disclosure provide a computer-readable medium storing a computer program thereon, wherein when the program is executed by a processor, it implements the method described in any implementation manner of the first aspect above.

[0012] The above embodiments of the present disclosure have the following beneficial effects: Through the multi-dimensional layer construction method for urban national land space monitoring according to some embodiments of the present disclosure, a multi-dimensional layer with more accurate layer boundaries can be generated. Specifically, the reason for the inaccurate generation of the multi-dimensional layer is that when sorting different identified regional elements (such as road areas, vegetation areas, etc.) in the same layer, it is difficult to accurately distinguish the boundaries between different regions. Based on this, in the multi-dimensional layer construction method for urban national land space monitoring according to some embodiments of the present disclosure, first, a sequence of high-resolution remote sensing images of the urban area taken by a remote sensing satellite is obtained. Then, according to the preset digital elevation model corresponding to the urban area, image fusion processing is performed on each high-resolution remote sensing image of the urban area in the above sequence of high-resolution remote sensing images of the urban area to obtain a fused remote sensing image of the urban area. Among them, the image fusion processing includes registration fusion operations and projection operations on each high-resolution remote sensing image of the urban area, and the operation sequence of the registration fusion operation and the projection operation is set according to the regional characteristics in each high-resolution remote sensing image of the urban area. The above fused remote sensing image of the urban area is a three-dimensional image including elevation values. By introducing a preset elevation digital model, it is convenient to fuse remote sensing images and provide accurate elevation values for the three-dimensional image. Through fusion, multiple remote sensing images within the urban area can be fused into the same three-dimensional image. Here, by setting different operation sequences, it can be used to further improve the accuracy of image fusion. After that, multi-dimensional feature recognition is performed on the above fused remote sensing image of the urban area to obtain an image feature set, and an initial layer block group set is constructed according to the above image feature set. Among them, each initial layer block group in the above initial layer block group set corresponds to one dimension. Here, through multi-dimensional feature recognition, it can be used to determine the regional characteristics in the fused remote sensing image of the urban area, and each region is used as an initial layer block. Also because the initial layer blocks are constructed, it is not only convenient for subsequent layer construction but also can be used to accurately control the boundary positions of the identified regions. Thus, it is possible to better perform boundary division in the multi-dimensional layer. Then, layer splicing is performed on each initial layer block in each initial layer block group in the above initial layer block group set to generate a single-dimensional layer, and a set of single-dimensional layers is obtained. Here, initial layer blocks of the same type (such as those corresponding to the road area type) can be spliced into a single-dimensional layer. Thus, each type of region corresponding to an item within the urban area can be represented by each single-dimensional layer. Finally, each single-dimensional layer in the above set of single-dimensional layers is mapped to a pre-set urban area map to obtain a multi-dimensional layer corresponding to the urban area, where each layer in the above multi-dimensional layer corresponds to a layer option in the layer operation component on the display terminal page. Thus, a multi-dimensional layer with more accurate regional boundaries can be generated. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In combination with the accompanying drawings and with reference to the following specific embodiments, the above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the elements and elements are not necessarily drawn to scale.

[0014] Figure 1 is a flowchart of some embodiments of a method for constructing a multi-dimensional layer for urban land and space monitoring according to the present disclosure;

[0015] Figure 2-a is a schematic diagram of a road before fusion;

[0016] Figure 2-b is a schematic diagram of the error of the road after fusion;

[0017] Figure 3 is a schematic diagram of image edge marking;

[0018] Figure 4 is a schematic structural diagram of some embodiments of a multi-dimensional layer construction device for urban land and space monitoring according to the present disclosure;

[0019] Figure 5 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. Specific Embodiments

[0020] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not used to limit the protection scope of the present disclosure.

[0021] In addition, it should be noted that for the sake of convenience of description, only the parts related to the relevant invention are shown in the drawings. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0022] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules, or units, and are not used to limit the order or interdependence relationship of the functions performed by these devices, modules, or units.

[0023] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive. Those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0024] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are for illustrative purposes only and are not used to limit the scope of these messages or information.

[0025] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0026] Figure 1 Flow 100 of some embodiments of a multi-dimensional layer construction method for urban land and space monitoring according to the present disclosure is shown. The multi-dimensional layer construction method for urban land and space monitoring includes the following steps:

[0027] S101, obtaining a sequence of high-resolution remote sensing images of an urban area captured by a remote sensing satellite.

[0028] In some embodiments, the execution subject (e.g., a computing device) of the multi-dimensional layer construction method for urban land and space monitoring can obtain a sequence of high-resolution remote sensing images of an urban area captured by a remote sensing satellite in a wired or wireless manner. Among them, the urban area can refer to the area that needs to be monitored for space. The sequence of high-resolution remote sensing images can include images of consecutive frames of the urban area.

[0029] As an example, the remote sensing satellite can be the Gaofen-2 satellite. It should be noted that the above computing device can be hardware or software. When the computing device is hardware, it can be implemented as a distributed cluster composed of multiple servers or terminal devices, or as a single server or a single terminal device. When the computing device is embodied as software, it can be installed in the above-listed hardware devices. It can be implemented as multiple software or software modules for providing distributed services, or as a single software or software module. No specific limitation is made here.

[0030] S102, performing image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area according to a preset digital elevation model corresponding to the urban area to obtain a fused remote sensing image of the urban area.

[0031] In some embodiments, the above execution subject can perform image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area according to a preset digital elevation model corresponding to the urban area to obtain a fused remote sensing image of the urban area. Among them, the image fusion processing includes registration fusion operations and projection operations on each high-resolution remote sensing image of the urban area, and the operation sequence of the registration fusion operation and the projection operation is set according to the regional characteristics in each high-resolution remote sensing image of the urban area. The above fused remote sensing image of the urban area can be a three-dimensional image including elevation values.

[0032] In some optional implementation manners of some embodiments, the above-mentioned execution subject performs image fusion processing on each high-resolution remote sensing image of the urban area in the high-resolution remote sensing image sequence of the urban area according to the preset digital elevation model corresponding to the urban area, and obtains a fused remote sensing image of the urban area, including:

[0033] S1021. Through the above-mentioned preset digital elevation model, determine the terrain undulation degree and altitude value of each high-resolution remote sensing image of the urban area in the high-resolution remote sensing image sequence of the urban area, and obtain a terrain undulation degree sequence and an altitude value sequence. Wherein, the terrain undulation degree in the terrain undulation degree sequence and the altitude value in the altitude value sequence characterize the regional characteristics of the corresponding high-resolution remote sensing image of the urban area. Secondly, the terrain undulation degree and altitude value can be generated through the following steps: First, the satellite shooting angle and satellite position coordinates when each high-resolution remote sensing image of the urban area is taken by the remote sensing satellite can be used to determine the coordinates of the four corner points of the fixed area of the high-resolution remote sensing image of the urban area by using geometric relationships. Here, the fixed area can be a square area with the same center point as the image. The coordinates of the four corner points of the fixed area are marked in the high-resolution remote sensing image of the urban area in advance according to the scaling ratio (for example, 0.8). The four corner point coordinates are on the line connecting the image center and the four corner points. Then, the undulation degree values within each analysis window of the square area can be determined by using the elevation values in the area between the above four corner point coordinates in the digital elevation model. For example, the analysis window can be a window with a size range of 150 meters × 150 meters or 300 meters × 300 meters, and the difference between the maximum elevation value and the minimum elevation value within the analysis window is determined as the local undulation degree. Here, the average value of the local undulation degrees of each analysis window can be determined as the terrain undulation degree. In addition, the average value of the elevation values within the square area can be determined as the altitude value.

[0034] In practice, considering that the corresponding relationship between the coordinates in the image and the longitude and latitude coordinates or the digital elevation model has not been established currently, and at the same time, in order to avoid unnecessary coordinate comparison before image correction. Therefore, by introducing the satellite position and shooting direction, the high-resolution remote sensing image is quickly and vaguely located by using geometric relationships. Then, considering the positioning error, therefore, when determining the terrain undulation degree and altitude value corresponding to the image, the terrain undulation degree and altitude value within the analysis window are determined. Therefore, the error of the vague positioning can be offset. In this way, the regional characteristics of each high-resolution remote sensing image of the urban area, that is, the terrain undulation degree and altitude value, are determined.

[0035] S1022. Determine the high-resolution remote sensing images of urban areas that meet the preset mountain screening conditions in the above-mentioned high-resolution remote sensing image sequence of urban areas as high-resolution remote sensing images of mountains, and obtain a high-resolution remote sensing image sequence of mountains. Among them, the preset mountain screening conditions can be that the altitude value corresponding to the high-resolution remote sensing image of the urban area is greater than the preset altitude threshold. For example, greater than 200 meters. Or the terrain undulation degree corresponding to the high-resolution remote sensing image of the urban area is greater than 200 meters. Here, a large terrain undulation degree or a high altitude value can be used to characterize that the area is very likely to be a mountainous area. Therefore, determine the high-resolution remote sensing images of urban areas that meet the preset mountain screening conditions as high-resolution remote sensing images of mountains.

[0036] S1023. Determine the high-resolution remote sensing images of urban areas that meet the preset plain screening conditions in the above-mentioned high-resolution remote sensing image sequence of urban areas as high-resolution remote sensing images of mountains, and obtain a high-resolution remote sensing image sequence of plains. Among them, the preset plain screening conditions can be that the altitude value corresponding to the high-resolution remote sensing image of the urban area is less than the preset altitude threshold. For example, greater than 50 meters. Or the terrain undulation degree corresponding to the high-resolution remote sensing image of the urban area is less than 50 meters. Here, a small terrain undulation degree or a low altitude value can be used to characterize that the area is a plain area. Therefore, determine the high-resolution remote sensing images of urban areas that meet the preset plain screening conditions as high-resolution remote sensing images of plains.

[0037] In addition, for the high-resolution remote sensing images of urban areas where both the altitude value and the terrain undulation degree are within the range of [50 meters, 200 meters], they are also regarded as high-resolution remote sensing images of mountains.

[0038] S1024. For the above-mentioned high-resolution remote sensing image sequence of mountains, perform the following first processing steps (a - c):

[0039] a. Using the above-mentioned preset digital elevation model, project each high-resolution remote sensing image of mountains in the above-mentioned high-resolution remote sensing image sequence of mountains onto the pre-established urban three-dimensional map coordinate system to obtain a set of projected three-dimensional mountain images. The ordinate of the urban three-dimensional map coordinate system can be an initial value (for example, 0). Therefore, the elevation value corresponding to the same coordinate in the above-mentioned preset digital elevation model and the urban three-dimensional map coordinate can be set in the urban three-dimensional map to supplement the elevation value in the three-dimensional map coordinate system, and obtain an adjusted urban three-dimensional map coordinate system. Then, through the method of coordinate projection, project the high-resolution remote sensing images of mountains onto the pre-established urban three-dimensional map coordinate system to obtain a set of projected three-dimensional mountain images. Thus, each coordinate of the projected three-dimensional mountain image can correspond to an elevation value, that is, the vertical coordinate value.

[0040] As an example, the coordinate projection method may include the Gauss-Kruger projection, or the UTM (Universal Transverse Mercator Projection), etc.

[0041] b. Image registration is performed on each of the three-dimensional mountain area images after the above projection in the three-dimensional mountain area image set after projection to obtain a first image registration data set. Among them, first, feature extraction can be performed on two adjacent three-dimensional mountain area images after projection through a feature extraction algorithm to obtain two groups of image feature points. Then, feature point matching can be performed on the two groups of image feature points through a feature matching algorithm to generate a set of matching feature point pairs. After that, the translation parameters, rotation parameters, and scaling parameters corresponding to the set of feature point pairs can be determined through a preset transformation model. Finally, the set of matching feature point pairs and the corresponding translation parameters, rotation parameters, and scaling parameters are determined as the first image registration data.

[0042] As an example, the transformation model can adopt Homography.

[0043] c. Using the above first image registration data set, image fusion is performed on each of the above three-dimensional images after projection to obtain a first fused image. Among them, each first image registration data can be geometrically transformed according to the corresponding three-dimensional image after projection to obtain a transformed three-dimensional image. Secondly, through the method of weighted average fusion, the transformed three-dimensional images are fused in the shooting order to obtain a first fused image. Here, the method of weighted average fusion can perform weighted average on the pixel values in the image overlapping area to obtain the fused pixel value.

[0044] In some optional implementation manners of some embodiments, the above execution subject performs image fusion processing on each high-resolution remote sensing image of the town area in the high-resolution remote sensing image sequence of the town area according to a preset digital elevation model corresponding to the town area to obtain a fused remote sensing image of the town area, and further includes:

[0045] S1025. For each high-resolution remote sensing image of the plain area in the above high-resolution remote sensing image sequence of the plain area, the following second processing steps (d - f) are performed:

[0046] d. Image registration is performed on each high-resolution remote sensing image of the plain area in the above high-resolution remote sensing image sequence of the plain area to obtain a second image registration data set. Among them, image registration can be performed through the above transformation model.

[0047] e. Using the above second image registration data set, image fusion is performed on each of the above high-resolution remote sensing images of the plain area to obtain a fused and registered image. Among them, image fusion can be performed through the above image fusion steps to obtain a fused and registered image.

[0048] f. Project the above fused and registered image onto the above urban 3D map coordinate system to obtain a second fused image. Among them, the above fused and registered image can be projected onto the above urban 3D map coordinate system through the above coordinate projection method to obtain a second fused image.

[0049] S1026. Perform secondary image fusion on the above first fused image and the above second fused image to obtain a fused remote sensing image of the urban area. Among them, the above first fused image and the above second fused image can be subjected to secondary image fusion through the weighted average fusion method to obtain a fused remote sensing image of the urban area.

[0050] Optionally, considering that there is also an easy situation of misalignment of the stitching boundary during the secondary image fusion of the above first fused image and the above second fused image, and the boundary misalignment causes the pixel values at some positions to be zero. Therefore, the blank area of the fused remote sensing image of the urban area can be detected through a preset detection condition to obtain a blank area coordinate set. Here, the detection condition can be that the pixel value of the coordinate in the fused remote sensing image of the urban area is empty. In response to the number of blank area coordinates being greater than a preset threshold (or the area of the blank area being greater than a preset area), after determining that secondary registration is required, perform secondary image fusion again. Here, secondary registration can be performed through the above image registration method, and thus a more accurate fused remote sensing image of the urban area can be obtained.

[0051] In practice, the commonly used image fusion method can be to first perform image registration and fusion on all images to obtain a fused image, and then project it onto a three-dimensional map. The registration error of this method is small and easy to control. However, for high-altitude areas in mountainous regions, errors often occur due to the transformation during the projection process. Additionally, it can also be to first project all the images onto a three-dimensional map and then perform registration and fusion. And this method results in a higher-precision fused image due to the introduction of elevation values. At the same time, the projection errors corresponding to individual images are relatively independent, avoiding the accumulation of errors during fusion. However, this method requires more computing resources. Therefore, through step S102 and its related content, the present application sets different processing sequences according to images in different positions. That is, the operation sequences of the registration and fusion operations and the projection operation are set according to the regional characteristics in the high-resolution remote sensing images of each urban area. Here, by introducing mountain screening conditions and plain screening conditions, they can be used to distinguish whether the area where the image is located is a mountainous area or a plain. Thus, the two types of images can be respectively subjected to two different fusion processes. That is, considering the influence on high-altitude areas when projecting a two-dimensional image onto a three-dimensional image, for mountain images, projection is first performed, and then registration and fusion are carried out. Thus, the projection error is offset to a certain extent through registration. At the same time, by adopting the sequence of first performing registration and fusion on plain images and then projection, the computing efficiency can be greatly improved, and the occupation of computing resources can be reduced. Therefore, through the strategy of mixing the above two methods, not only can the consumption of computing resources be balanced, but also the boundaries of the images can be accurately located, thereby improving the accuracy of the generated fused remote sensing images of urban areas.

[0052] Optionally, before the above-mentioned execution subject performs multi-dimensional feature recognition on the above-mentioned fused remote sensing image of the urban area to obtain an image feature set, it further includes (the first step - the fourth step):

[0053] In the first step, highway recognition is performed on the above-mentioned fused remote sensing image of the urban area to obtain a sequence of highway coordinate groups. Among them, each highway coordinate in the above-mentioned sequence of highway coordinate groups represents the highway position coordinates in the urban area corresponding to a highway level greater than or equal to a preset highway level (such as a fourth-class highway), and each highway coordinate group corresponds to a highway identifier. Each highway coordinate group can correspond to a highway identifier, that is, it corresponds to a highway. Among them, a preset highway recognition algorithm can be used to perform highway recognition on the above-mentioned fused remote sensing image of the urban area to obtain a sequence of highway coordinate groups.

[0054] As an example, the highway recognition algorithm can include but is not limited to at least one of the following: mathematical morphology algorithm, Hough transform, or a pre-trained convolutional neural network.

[0055] In the second step, each highway coordinate group in the above highway coordinate group sequence is screened to obtain a target highway coordinate group sequence. Among them, each target highway coordinate group in the above target highway coordinate group sequence meets the screening condition of being in two adjacent projected mountainous area three-dimensional images or two adjacent plain high-resolution remote sensing images. The target highway coordinate group can be used to represent the highway corresponding to the two spliced images.

[0056] In practice, considering that the boundaries in image recognition are prone to be blurred, identifying the highway first can be used as the decomposition line of the recognition area, thereby further accurately determining the boundaries of different recognition areas. At the same time, considering that the images are obtained by splicing and there are likely to be large undetected splicing errors, which will affect the establishment of subsequent layers. Therefore, by screening the target highway coordinate groups, it is used to judge whether there are splicing errors between two adjacent images.

[0057] In the third step, for each target highway coordinate group in the above target highway coordinate group sequence, the following fusion and correction steps (g-j) are executed:

[0058] g. Using the two adjacent projected mountainous area three-dimensional images or two adjacent plain high-resolution remote sensing images corresponding to the above target highway coordinate group, determine the splicing points in the above target highway coordinate group. Here, each target highway coordinate group can correspond to a highway and can also correspond to two adjacent projected mountainous area three-dimensional images or two adjacent plain high-resolution remote sensing images that the highway passes through. Thus, the splicing position coordinates during the splicing process of the two adjacent projected mountainous area three-dimensional images or two adjacent plain high-resolution remote sensing images can be determined. Consequently, in the target highway coordinate group, the target highway coordinates corresponding to the splicing position coordinates can be selected as the splicing points. In practice, if the splicing area between two adjacent images is relatively wide, the positions where the boundaries of the splicing area intersect the target highway coordinate group can be used as the splicing position coordinates.

[0059] h. Using the above splicing points as the dividing points, respectively perform fitting processing on the target highway coordinates on both sides of the dividing point in the above target highway coordinate group to obtain the first highway equation and the second highway equation. Here, the curve fitting equation can be used to perform fitting processing on the target highway coordinates on both sides of the dividing point in the above target highway coordinate group to obtain the first highway equation and the second highway equation.

[0060] i. Generate a fusion error value according to the above first road equation and the above second road equation. Specifically, first, the actual coordinate sequence of the corresponding road can be obtained from the actual scene map. Then, each actual coordinate in the actual coordinate sequence can be subjected to fitting processing to obtain the actual road equation. After that, the actual road angle value of the actual road equation at the above splicing point position can be determined. In addition, the tangent angle values of the above first road equation and the above second road equation at the above splicing point position can be determined. Finally, the difference between the tangent angle value and the actual road angle value can be determined as the fusion error.

[0061] As an example, refer to Figure 2-a the schematic diagram of the road before fusion, and Figure 2-b the schematic diagram of the road error after fusion shown in Figure 2-a and Figure 2-b The splicing point position 201 in the figure is the dividing point of the road position before and after the fusion of the two images. Figure 2-a The actual road angle at the splicing point position 201 in the figure is shown as the tangent. Figure 2-b The splicing point position 201 in the figure is the image deformation caused by the image transformation after the fusion of the two images. Here, a clear comparison can be made from the aspect of the road curve. Therefore, in this application, the offset error caused by the transformation, that is, the fusion error, is determined by the difference between the angle at the splicing point position 201 and the actual angle value introduced by the road equation. Thus, it can be used to quantify the degree of the error and locate the position of the error, so as to facilitate further correction in the subsequent process.

[0062] j. In response to at least one fusion error value corresponding to two adjacent projected mountainous area three-dimensional images or two adjacent plain high-resolution remote sensing images corresponding to the target road coordinate group satisfying the preset error condition, determine that an edge of the two adjacent projected mountainous area three-dimensional images or the two adjacent plain high-resolution remote sensing images is fused incorrectly, and record the error edge identifier corresponding to each projected mountainous area three-dimensional image or each plain high-resolution remote sensing image. The error condition may be that the fusion error is greater than a preset error threshold. Here, it is also considered that each image is fused with 2-4 adjacent images during the fusion process. Therefore, an error edge identifier is set to determine the specific error position.

[0063] As an example, refer to Figure 3。The identifiers of two adjacent three-dimensional images of the mountainous areas after projection can be A and B. According to their positional relationship during fusion, the specific wrong edges can be determined. For example, if A and B are fused left and right, then the wrong edges can be the right edge of image A and the left edge of image B. Then, through a unified image edge identifier, for example, in a clockwise order, the identifiers corresponding to the four boundaries of the upper side, right side, lower side, and left side of image A can be A1, A2, A3, and A4 respectively. Thus, the final wrong edge identifiers can be: A2 and B4.

[0064] Step 4: For each of the recorded wrong edge identifiers, perform secondary registration on the corresponding two three-dimensional images of the mountainous areas after projection or two high-resolution remote sensing images of the plains, and perform secondary fusion on the two three-dimensional images of the mountainous areas after projection or two high-resolution remote sensing images of the plains after registration to obtain a corrected remote sensing image of the urban area. Among them, for each three-dimensional image of the mountainous areas after projection or each high-resolution remote sensing image of the plains, if there is a corresponding recorded wrong edge identifier, secondary fusion can be performed again through the above registration and fusion steps. Here, the secondary fusion is a local image fusion.

[0065] Here, taking Figure 3 image A and image B above as examples, only the adjacent area between image A and image B can be secondarily fused. If there are also wrong edge identifiers between image A and other images, secondary fusion can also be performed. Thus, the accuracy of the corrected remote sensing image of the urban area is further improved.

[0066] S103. Perform multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and construct an initial layer block group set according to the image feature set.

[0067] In some embodiments, the above execution subject can perform multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and construct an initial layer block group set according to the above image feature set. Among them, each initial layer block group in the above initial layer block group set can correspond to one dimension.

[0068] In some optional implementation manners of some embodiments, the above execution subject performs multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and constructs an initial layer block group set according to the above image feature set, including:

[0069] S1031. Determine the current coordinates corresponding to the above highway coordinate group sequence in the rectified remote sensing image of the urban area to obtain the current highway coordinate group sequence. Among them, for the highway coordinates in the image area of the rectified remote sensing image of the urban area that has not been re-registered, they can be directly used as the current highway coordinates. For the highway coordinates in the image area of the rectified remote sensing image of the urban area that has been re-registered, the coordinate correction can be performed through the registration data generated during the re-registration to obtain the current highway coordinates. Additionally, the highway coordinates of the image area after re-registration can also be identified through the above highway coordinate recognition algorithm to obtain the current highway coordinates.

[0070] S1032. Using the highways corresponding to each current highway coordinate in the above current highway coordinate sequence as the regional division boundaries, divide the above fused remote sensing image of the urban area to obtain a set of divided regions, and simultaneously perform regional recognition on each divided region in the above set of divided regions to generate a set of regional recognition information. Among them, each regional recognition information corresponds to a recognized region, each recognized region includes at least one divided region, and each regional recognition information in the above set of regional recognition information includes a region identifier and a regional boundary coordinate group. The difference identifier is used to characterize the layer attribute of the region, and different layer attributes correspond to different layer division dimensions. Here, the semantic segmentation model pre-trained can be used to simultaneously perform regional recognition on each divided region in the above set of divided regions to generate a set of regional recognition information.

[0071] In practice, the model structure of the above semantic segmentation model may include: an input layer, an encoder, a decoder, and an output layer. Specifically, the fused remote sensing image of the urban area can be divided into image data of three channels of RGB (Red, Green, Blue) as an input to the model. Then, a binary map can also be generated as another input to the model. Here, the binary map can be a binary map with the same size as the fused remote sensing image of the urban area and only marking the coordinates of each current road. Thus, the fused remote sensing image of the urban area (image data of three channels) and the binary map (image data of one channel) can be input into the input layer. Then, through channel splicing by the input layer, multi-modal image data of four channels is obtained. Here, the input layer can include a group of convolutional modules, including a convolutional layer, a batch normalization layer, and an activation layer. For example, the convolutional kernel of the convolutional layer can be 7×7. The stride can be 2. The padding can be 3. The input channels are 4, and the output channels are 64. This is to facilitate the input to the corresponding encoder and at the same time replace the first layer of the ResNet (Residual Network) residual network. Secondly, the encoder can be composed of a ResNet residual network as the backbone network module (Residual Block) and an Atrous Spatial Pyramid Pooling (ASPP) module. Here, the backbone network module can include at least four groups of residual blocks. The convolutional layer in the last residual block can be set as a dilated convolution to maintain the image resolution. A feature map with 512 output channels is obtained. The Atrous Spatial Pyramid Pooling module can include a convolutional layer branch, three dilated convolutions with different rates (for example, the rates can be set to 6, 12, 18 to capture multi-scale object features, such as small-scale pond areas and large-scale forest areas), one global average pooling branch (Global Average Pooling, GAP), and a fusion output layer (Fusion Output Layer). Thus, a feature map with 256 output channels is obtained. In addition, the decoder can include a first-stage upsampling module, a second-stage upsampling module, an upsampling and classification layer. The upsampling module can include an upsampling layer and a skip connection fusion layer (Skip Connection Fusion Layer). Here, the first-stage upsampling module is a skip connection from the Atrous Spatial Pyramid Pooling module to the 3rd residual block. The second-stage upsampling module is a skip connection with the 1st residual block. In this way, shallow feature fusion is performed, making the model pay more attention to the detailed features of the input road boundary. The upsampling and classification layer can restore the scale of the feature map to the same dimension as the input and output a probability map through the classification layer. Here, each region in the probability map can correspond to a discrimination identifier for characterizing the layer attributes of the region. If the layer attributes of adjacent regions are the same, they can be associated as the same region.

[0072] S1033. Classify the above regional recognition information according to the layer attributes included in each piece of regional recognition information in the above regional recognition information set to obtain a set of classified regional recognition information groups. Among them, each classified regional recognition information group can correspond to a layer attribute. Here, the layer attributes may include, but are not limited to, at least one of the following: highway attribute, water area attribute, residential land attribute, construction land attribute, forest land attribute, grassland attribute, wetland attribute, garden land attribute, special land attribute, etc.

[0073] S1034. Determine the recognition regions corresponding to the respective regional recognition information in each classified regional recognition information group in the above set of classified regional recognition information groups as initial layer block groups to obtain a set of initial layer block groups. Among them, each initial layer block includes a corresponding group of regional boundary coordinates.

[0074] S104. Perform layer splicing on each initial layer block in each initial layer block group in the set of initial layer block groups to generate a one-dimensional layer, obtaining a set of one-dimensional layers.

[0075] In some embodiments, the above execution entity may perform layer splicing on each initial layer block in each initial layer block group in the above set of initial layer block groups to generate a one-dimensional layer, obtaining a set of one-dimensional layers. Among them, each one-dimensional layer can correspond to a layer attribute.

[0076] As an example, the one-dimensional layer may be: highway layer, water area layer, residential land attribute, construction land attribute, forest land layer, grassland layer, wetland layer, garden land layer, special land layer, etc.

[0077] In some optional implementation manners of some embodiments, the above execution entity performs layer splicing on each initial layer block in each initial layer block group in the above set of initial layer block groups to generate a one-dimensional layer, obtaining a set of one-dimensional layers, including:

[0078] For each initial layer block group in the above set of initial layer block groups, perform the following processing steps:

[0079] Combine each initial layer block in the above initial layer block group to obtain a one-dimensional layer. Among them, the one-dimensional layer is bound to a preset layer option. Among them, the above layer option corresponds to the same layer attribute as each initial layer block in the above initial layer block group. Here, each layer option can correspond to a one-dimensional layer. Specifically, each initial layer block can be arranged according to the original position to obtain a one-dimensional attribute image. Secondly, other areas in the image can be set to be transparent. In addition, the path of the one-dimensional attribute image can be associated with the preset layer option to complete the binding. Thus, by displaying the layer option on the user terminal, it is convenient for the user to control the layer to perform operations such as displaying, adjusting, or rendering the one-dimensional layer.

[0080] In addition, it can also be set that one layer control binds multiple one-dimensional layers through different association paths.

[0081] S105, Map each one-dimensional layer in the one-dimensional layer set to a preset urban area map to obtain a multi-dimensional layer corresponding to the urban area.

[0082] In some embodiments, the above execution subject can map each one-dimensional layer in the one-dimensional layer set to a preset urban area map to obtain a multi-dimensional layer corresponding to the urban area. Among them, each layer in the above multi-dimensional layer corresponds to a layer option in the layer operation component on the display terminal page. Here, the urban area map is the same size as each one-dimensional layer. Thus, the mapping can be to stack each one-dimensional layer onto the urban area map in the preset attribute order to obtain a multi-dimensional layer.

[0083] Optionally, the above execution subject may further include the following steps:

[0084] S106, In response to receiving the layer operation information of the user for the layer operation component, according to the above operation information, perform a rendering processing operation on the corresponding one-dimensional layer in the above multi-dimensional layer to obtain a processed one-dimensional layer. Among them, the layer operation component includes at least one function control, and the function control is a control for performing layer rendering and / or layer order adjustment on the one-dimensional layer.

[0085] As an example, the user selects the one-dimensional layer corresponding to the forest land attribute and issues layer operation information representing layer rendering. Then the above execution subject can render the one-dimensional layer corresponding to the forest land attribute to obtain a processed one-dimensional layer. Here, the rendering can be to render the non-transparent area of the layer.

[0086] S107, Update the above processed one-dimensional layer to the above multi-dimensional layer to obtain an updated multi-dimensional layer, and send the above updated multi-dimensional layer to the above display terminal for display.

[0087] In practice, considering the situation that there is a large amount of layer information, a large layer area, and rendering requires consuming a large amount of computing resources. Therefore, layer rendering is performed through an electronic device (such as the above-mentioned execution entity). At the same time, the accuracy of image rendering can be improved.

[0088] Optionally, satellite remote sensing images, unmanned aerial vehicle (UAV) remote sensing images, oblique photography images, etc. with a resolution of 0.05 meters or higher can be collected. When comprehensively analyzing and selecting, relatively new and higher-resolution images are selected to produce orthophotos for urban and rural territorial space monitoring. Priority is given to using the initial survey orthophoto as the basis. When necessary, elevation model data is collected and utilized to provide assistance, and orthophoto processing is organized to produce orthophotos.

[0089] Optionally, the further refinement and supplementation of each layer can be as follows: Using the urban area as the base map, guided by the collected data, overlaying the latest remote sensing images, and using various technical means, combined with on-site investigations, to determine the location, scope, and attributes of the monitoring objects. For monitoring objects with independent land use, the relevant positions and scopes are vectorized directly on the base map of the change investigation results, and the relevant attributes are marked; for monitoring objects without independent land use, their vector positions are represented by point layers, and the relevant attributes are marked. For those where the actual situation on the ground changes from the original land type during monitoring, if it belongs to the land types that require refinement and supplementation of relevant information, it is refined into third-level types. If it does not belong to the land types that require refinement and supplementation of relevant information, its scope is marked. In addition, relevant elements can also be updated. For example, the update of water network data includes: updating the positions of river and lake shorelines, river structure lines, names, and codes, updating and supplementing lake water quality, reservoir and pond uses, volumes, river types, navigation properties, grades, and the grades and flow directions of water channels, etc.; collecting and updating water conservancy project facilities such as dams, gates, drainage and irrigation stations, pumping stations, and important wells that are closely related to the water network; and for another example, the update of road network data includes: updating the line positions of railways, supplementing information such as names, codes, starting and ending points, types, up and down directions, single or double tracks, etc.; updating the centerlines of intercity highways, improving attribute information such as names, codes, widths, traffic directions, number of lanes, technical grades, types, and paving materials; updating the centerlines of rural accessible roads, improving attribute information such as names, widths, and types; updating the urban road network, supplementing attribute information such as names, types, road widths, number of lanes, and elevated conditions; and at the same time, collecting and updating the positions and attributes of important transportation facilities such as stations, bridges, tunnels, highway entrances and exits, service areas, overpasses, subway stations, and transportation hubs that are closely related to road accessibility.

[0090] Optionally, land changes can also be dynamically monitored to update the multi-dimensional layers in real time. For example, suspected newly added construction patches are classified and extracted. Outside the scope of the construction land patches in the survey database, various newly added construction patches are extracted according to image features, including obvious construction land, rural settlements, villas, hydraulic facilities, roads, railways, suspected construction land, suspected facility agricultural land, earthwork (heap), photovoltaic panels, football fields, golf courses, reclamation by filling sea, etc. Another example is to extract the demolition situation of the suspected demolition patches in the layer, which are classified as construction land and facility agricultural land, and extract the building (structure) demolition patches, and subdivide the demolition patches according to types such as being under demolition, having been cleared after demolition, and having vegetation restored.

[0091] The above embodiments of the present disclosure have the following beneficial effects: Through the multi-dimensional layer construction method for urban land space monitoring according to some embodiments of the present disclosure, a multi-dimensional layer with more accurate layer boundaries can be generated. Specifically, the reason for the inaccurate generated multi-dimensional layer is that when sorting different identified regional elements (such as road areas, vegetation areas, etc.) in the same layer, it is difficult to accurately distinguish the boundaries between different regions. Based on this, in the multi-dimensional layer construction method for urban land space monitoring according to some embodiments of the present disclosure, first, a sequence of high-resolution remote sensing images of the urban area taken by a remote sensing satellite is obtained. Then, according to the preset digital elevation model corresponding to the above urban area, image fusion processing is performed on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area to obtain a fused remote sensing image of the urban area. Among them, the image fusion processing includes registration operations and fusion operations on each high-resolution remote sensing image of the urban area, and the operation sequence corresponding to the registration operation and the fusion operation is set according to the regional characteristics in each high-resolution remote sensing image of the urban area. The above fused remote sensing image of the urban area is a three-dimensional image including elevation values. By introducing a preset elevation digital model, it is convenient to fuse remote sensing images and provide accurate elevation values for three-dimensional images. Through fusion, multiple remote sensing images within the urban area can be fused into the same three-dimensional image. Here, by setting different operation sequences, it can be used to further improve the accuracy of image fusion. After that, multi-dimensional feature recognition is performed on the above fused remote sensing image of the urban area to obtain an image feature set, and an initial layer block group set is constructed according to the above image feature set. Among them, each initial layer block group in the above initial layer block group set corresponds to one dimension. Here, through multi-dimensional feature recognition, it can be used to determine the regional characteristics in the fused remote sensing image of the urban area, and each region is used as an initial layer block accordingly. Also, because the initial layer blocks are constructed, it is not only convenient for subsequent layer construction but also can be used to accurately control the boundary positions of the identified regions. Thus, it is possible to better perform boundary division in the multi-dimensional layer. Then, layer splicing is performed on each initial layer block in each initial layer block group in the above initial layer block group set to generate a single-dimensional layer, and a set of single-dimensional layers is obtained. Here, initial layer blocks of the same type (for example, all corresponding to the road area type) can be spliced into a single-dimensional layer. Thus, each type of region within the urban area can be represented by each single-dimensional layer. Finally, each single-dimensional layer in the above set of single-dimensional layers is mapped to a pre-set urban area map to obtain a multi-dimensional layer corresponding to the urban area, where each layer in the above multi-dimensional layer corresponds to a layer option in the layer operation component on the display terminal page. Thus, a multi-dimensional layer with more accurate regional boundaries can be generated.

[0092] Further reference Figure 4, as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a multi-dimensional layer construction device for urban land space monitoring. These device embodiments correspond to Figure 1 the method embodiments shown, and the multi-dimensional layer construction device for urban land space monitoring can be specifically applied to various electronic devices.

[0093] As Figure 4 shown, a multi-dimensional layer construction device 400 for urban land space monitoring in some embodiments includes: an acquisition unit 401, an image fusion unit 402, an image recognition unit 403, a single-dimensional layer construction unit 404, and a multi-dimensional layer construction unit 405. Among them, the acquisition unit 401 is configured to acquire a sequence of high-resolution remote sensing images of urban areas taken by remote sensing satellites; the image fusion unit 402 is configured to perform image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area according to a preset digital elevation model corresponding to the urban area, to obtain a fused remote sensing image of the urban area, wherein the image fusion processing includes registration fusion operations and projection operations on each high-resolution remote sensing image of the urban area, and the operation sequence of the registration fusion operation and the projection operation is set according to the regional characteristics in each high-resolution remote sensing image of the urban area, and the fused remote sensing image of the urban area is a three-dimensional image including elevation values; the image recognition unit 403 is configured to perform multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and construct an initial layer block group set according to the image feature set, wherein each initial layer block group in the initial layer block group set corresponds to one dimension; the single-dimensional layer construction unit 404 is configured to splice each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer, to obtain a single-dimensional layer set; the multi-dimensional layer construction unit 405 is configured to map each single-dimensional layer in the single-dimensional layer set to a preset urban area map, to obtain a multi-dimensional layer of the corresponding urban area, wherein each layer in the multi-dimensional layer corresponds to one layer option in the layer operation components on the display terminal page.

[0094] It can be understood that the units described in the multi-dimensional layer construction device 400 for urban land space monitoring correspond to Figure 1 the respective steps in the method described with reference to

[0095] Next, with reference to Figure 5 , which shows a schematic structural diagram of an electronic device (such as a computing device) suitable for implementing some embodiments of the present disclosure.Figure 5 The electronic device shown is merely an example and should not impose any limitations on the functions and scope of use of the embodiments of the present disclosure. As Figure 5 shown, the computer device includes a processor, a memory, and a network interface connected via a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory. The non-volatile storage medium can store an operating system and computer programs. The computer programs include program instructions that, when executed, can cause the processor to execute any of the above methods. The processor is used to provide computing and control capabilities to support the operation of the entire computer device. The internal memory provides an environment for the operation of the computer programs in the non-volatile storage medium. When the computer programs are executed by the processor, the processor can be caused to execute any of the above methods. The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 5 the structure shown in

[0096] is merely a block diagram of some structures related to the solution of the present disclosure and does not constitute a limitation on the computer device to which the solution of the present disclosure is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different component layout.

[0097] Wherein, in one embodiment, the above-mentioned processor is used to run a computer program stored in a memory to implement the following steps: obtaining a sequence of high-resolution remote sensing images of urban areas taken by a remote sensing satellite; according to a preset digital elevation model corresponding to the urban area, performing image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area to obtain a fused remote sensing image of the urban area, wherein the image fusion processing includes a registration fusion operation and a projection operation on each high-resolution remote sensing image of the urban area, and the operation sequence of the registration fusion operation and the projection operation is set according to the regional features in each high-resolution remote sensing image of the urban area, and the fused remote sensing image of the urban area is a three-dimensional image including elevation values; performing multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and constructing an initial layer block group set according to the image feature set, wherein each initial layer block group in the initial layer block group set corresponds to one dimension; performing layer splicing on each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer, and obtaining a single-dimensional layer set; mapping each single-dimensional layer in the single-dimensional layer set to a preset urban area map to obtain a multi-dimensional layer corresponding to the urban area, wherein each layer in the multi-dimensional layer corresponds to a layer option in a layer operation component on a display terminal page.

[0098] An embodiment of the present disclosure also provides a computer-readable storage medium, on which a computer program is stored, and the computer program includes program instructions, and the method implemented when the program instructions are executed may refer to each embodiment of the above method of the present disclosure.

[0099] Wherein, the above-mentioned computer-readable storage medium may be an internal storage unit of the above-mentioned computer device in the foregoing embodiment, such as a hard disk or memory of the above-mentioned computer device. The above-mentioned computer-readable storage medium may also be an external storage device of the above-mentioned computer device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the above-mentioned computer device.

[0100] It should be noted that in this article, the term "comprising", "including" or any other variation thereof is intended to cover a non-exclusive inclusion, so that a process, method, article or system including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or system. Without further limitations, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, method, article or system including the element.

[0101] The above description is only some preferred embodiments of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, the technical solutions formed by mutually replacing the above features with the technical features (but not limited to) having similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A method for constructing a multi-dimensional layer for urban territorial space monitoring, characterized in that, Including: Obtaining a sequence of high-resolution remote sensing images of urban areas captured by a remote sensing satellite; According to a preset digital elevation model corresponding to the urban area, performing image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area to obtain a fused remote sensing image of the urban area. Among them, the image fusion processing includes registration fusion operation and projection operation on each high-resolution remote sensing image of the urban area. The operation sequence of the registration fusion operation and the projection operation is set according to the regional characteristics in each high-resolution remote sensing image of the urban area. The fused remote sensing image of the urban area is a three-dimensional image including elevation values; Performing multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and constructing an initial layer block group set according to the image feature set. Among them, each initial layer block group in the initial layer block group set corresponds to one dimension; Performing layer splicing on each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer, and obtaining a single-dimensional layer set; Mapping each single-dimensional layer in the single-dimensional layer set to a preset urban area map to obtain a multi-dimensional layer corresponding to the urban area. Among them, each layer in the multi-dimensional layer corresponds to one layer option in the layer operation component on the display terminal page; 2. The method according to claim 1, wherein The method further includes: In response to receiving user layer operation information for the layer operation component, according to the operation information, performing a rendering processing operation on the corresponding single-dimensional layer in the multi-dimensional layer to obtain a processed single-dimensional layer. Among them, the layer operation component includes at least one function control, and the function control is a control for performing layer rendering and / or layer order adjustment on the single-dimensional layer; Updating the processed single-dimensional layer to the multi-dimensional layer to obtain an updated multi-dimensional layer, and sending the updated multi-dimensional layer to the display terminal for display; 3. The method according to claim 1, characterized in that, The performing image fusion processing on each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area according to a preset digital elevation model corresponding to the urban area to obtain a fused remote sensing image of the urban area includes: Determining the terrain undulation degree and altitude value of each high-resolution remote sensing image of the urban area in the sequence of high-resolution remote sensing images of the urban area through the preset digital elevation model to obtain a terrain undulation degree sequence and an altitude value sequence. Among them, the terrain undulation degree in the terrain undulation degree sequence and the altitude value in the altitude value sequence characterize the regional characteristics of the corresponding high-resolution remote sensing image of the urban area; Determining the high-resolution remote sensing images of mountainous areas that meet the preset mountainous area screening conditions in the sequence of high-resolution remote sensing images of the urban area to obtain a sequence of high-resolution remote sensing images of mountainous areas; Determining the high-resolution remote sensing images of mountainous areas that meet the preset plain screening conditions in the sequence of high-resolution remote sensing images of the urban area to obtain a sequence of high-resolution remote sensing images of plains; For the high-resolution remote sensing image sequence of the mountainous area, the following first processing steps are performed: Using the preset digital elevation model, project each high-resolution remote sensing image in the mountainous area in the high-resolution remote sensing image sequence of the mountainous area into the pre-established urban three-dimensional map coordinate system to obtain a set of projected mountain three-dimensional images; Perform image registration on each projected mountain three-dimensional image in the set of projected mountain three-dimensional images to obtain a first image registration data set; Using the first image registration data set, perform image fusion on each of the projected three-dimensional images to obtain a first fused image.

4. The method according to claim 3, wherein The method for performing image fusion processing on each high-resolution remote sensing image of the urban area in the high-resolution remote sensing image sequence of the urban area according to the preset digital elevation model corresponding to the urban area to obtain a fused urban area remote sensing image further includes: For the high-resolution remote sensing image sequence of the plain, the following second processing steps are performed: Perform image registration on each high-resolution remote sensing image in the high-resolution remote sensing image sequence of the plain to obtain a second image registration data set; Using the second image registration data set, perform image fusion on each high-resolution remote sensing image of the plain to obtain a fused and registered image; Project the fused and registered image into the urban three-dimensional map coordinate system to obtain a second fused image; Perform secondary image fusion on the first fused image and the second fused image to obtain a fused urban area remote sensing image.

5. The method according to claim 4, wherein Before performing multi-dimensional feature recognition on the fused urban area remote sensing image to obtain an image feature set, the method further includes: Perform road recognition on the fused urban area remote sensing image to obtain a sequence of road coordinate groups, where each road coordinate in the sequence of road coordinate groups represents the road position coordinates in the urban area corresponding to a road level greater than or equal to the preset road level, and each road coordinate group corresponds to a road identifier; Perform screening processing on each road coordinate group in the sequence of road coordinate groups to obtain a sequence of target road coordinate groups, where each target road coordinate group in the sequence of target road coordinate groups satisfies the screening condition of being in two adjacent projected mountain three-dimensional images or two adjacent high-resolution remote sensing images of the plain, and the target road coordinate group is used to represent the road corresponding to it passing through and splicing the two images; For each target road coordinate group in the sequence of target road coordinate groups, perform the following fusion correction steps: Using the two adjacent projected mountain three-dimensional images or two adjacent high-resolution remote sensing images of the plain corresponding to the target road coordinate group, determine the splicing point in the target road coordinate group; Taking the splicing point as the dividing point, respectively perform fitting processing on the target road coordinates on both sides of the dividing point in the target road coordinate group to obtain a first road equation and a second road equation; Generate a fusion error value according to the first road equation and the second road equation; In response to at least one fusion error value corresponding to two adjacent projected mountainous three-dimensional images corresponding to a target highway coordinate group, or two adjacent high-resolution remote sensing images of plains, satisfying a preset error condition, determine that one side of the two adjacent projected mountainous three-dimensional images, or two adjacent high-resolution remote sensing images of plains, is fused incorrectly, and record the error side identifier corresponding to each projected mountainous three-dimensional image, or each high-resolution remote sensing image of plains; For each of the recorded error side identifiers, perform secondary registration on the corresponding two projected mountainous three-dimensional images, or two high-resolution remote sensing images of plains, and perform secondary fusion on the two registered projected mountainous three-dimensional images, or two high-resolution remote sensing images of plains, to obtain a corrected remote sensing image of the urban area.

6. The method according to claim 5, characterized in that Performing multi-dimensional feature recognition on the fused remote sensing image of the urban area to obtain an image feature set, and constructing an initial layer block group set according to the image feature set, including: Determine the current coordinates corresponding to the highway coordinate group sequence in the corrected remote sensing image of the urban area to obtain the current highway coordinate group sequence; Using the highways corresponding to each current highway coordinate in the current highway coordinate sequence as regional division boundaries, divide the fused remote sensing image of the urban area to obtain a set of divided regions, and simultaneously perform regional recognition on each divided region in the set of divided regions to generate a set of regional recognition information, where each regional recognition information corresponds to a recognition region, each recognition region includes at least one divided region, and each regional recognition information in the set of regional recognition information includes a region identifier and a regional boundary coordinate group. The difference identifier is used to characterize the layer attribute of the region, and different layer attributes correspond to different layer division dimensions; Classify the regional recognition information according to the layer attributes included in each regional recognition information in the set of regional recognition information to obtain a set of classified regional recognition information groups; Determine the recognition regions corresponding to each regional recognition information in each classified regional recognition information group in the set of classified regional recognition information groups as initial layer blocks to obtain an initial layer block group set, where each initial layer block includes the corresponding regional boundary coordinate group.

7. The method according to claim 1, characterized in that Performing layer splicing on each initial layer block in each initial layer block group in the initial layer block group set to generate a single-dimensional layer to obtain a set of single-dimensional layers, including: For each initial layer block group in the initial layer block group set, perform the following processing steps: Combine the initial layer blocks in the initial layer block group to obtain a single-dimensional layer, where the single-dimensional layer is bound to a preset layer option, and the layer option has the same layer attribute as the initial layer blocks in the initial layer block group.

Images